U.S. ENVIRONMENTAL PROTECTION AGENCY

SCIENCE ADVISORY BOARD STAFF OFFICE

CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE (CASAC) OXIDES OF NITROGEN
(NOx)PRIMARY REVIEW PANEL

PUBLIC MEETING

MARRIOTT AT RESEARCH TRIANGLE PARK

4700 Guardian Drive

Durham, North Carolina 27703

MAY 1, 2008

EPA CASAC MEETING 05/01/08 CCR#15905-1	2

2	4

1	U.S. ENVIRONMENTAL PROTECTION AGENCY

2	SCIENCE ADVISORY BOARD STAFF OFFICE

3		CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE (CASAC)

4		OXIDES OF NITROGEN (NOx)PRIMARY REVIEW PANEL

5	May 1, 2008

6	DR. NUGENT:    Good morning, everyone.  I

7  think it's time to start the formal time of the

8  meeting.  And, I know we have people on the telephone

9  line and I -- this is Angela Nugent, and I'd like to

10   make a brief statement in my role as Designated Federal

11   Officer for the Clean Air Scientific Advisory Committee

12   Oxides of Nitrogen Primary NOx Review Panel.  So

13   welcome, everyone.

14	This is a public meeting of the panel,

15   which is a subcommittee of the Clean Air Scientific

16   Advisory Committee.  I am Angela Nugent from EPA

17   Science Advisory Board Staff Office, and my job is to

18   support the committee.  I'd like to make a few brief

19   remarks in my capacity as DFO before I introduce our

20   Office Director, Dr. Vanessa Vu, and the Chair of the

21   Committee, Dr. Rogene Henderson.

22	This subcommittee, this panel, is a

23   panel of the Clean Air Scientific Advisory Committee,

24   and by law and EPA policy it operates under the

1  Nitrogen or Nitrogen NO2.

2	I'd like to take this opportunity on

3  behalf of the Agency to thank Dr. Rogene Henderson and

4  the distinguished members of this -- the CASAC -- the

5  Clean Air Scientific Advisory Committee, as well as the

6  distinguished members of this panel, which will be

7  providing the independent advice to the Administrator

8  of EPA regarding the health standards for the National

9  Ambient Air Quality Standards for Nitrogen Oxides, or

10   Oxides of  Nitrogen, which include all the other

11   nitrogen species.

12	I would like to thank Dr. Ila Cote and

13   Dr. Mary Ross  from the Office of Research and

14   Development, National Center for Environmental

15   Assessment and RTP for their interaction presentation

16   today -- the next two days -- regarding the second

17   draft ISA.  And also, the team from the Office of Air

18   Radiation, head by the -- represented today, including

19   Dr. Stephen Graham, Harvey Richmond, and Scott Jenkins

20   for their presentation today -- interaction with all of

21   you regarding the second draft -- first draft of the

22   Risk and Exposure Assessment.

23	As Angela indicated that public

24   involvement is a critical component of the deliberation

25   of the advisory process of the Clean Air Scientific



3	5

1  means the deliberations of this committee, this panel,

2  operate in public -- in a public forum. that notice is

3  given in advance of the meeting, that records are kept

4  of the meeting and that minutes are provided to the

5  public within ninety days.

6	There have been four requests for oral

7  comment today.  We have provided the public with

8  opportunities to provide written and oral comments, so

9  we've had four requests for oral comment.  There will

10   be three this morning, one this afternoon.  And there

11   have been sets of written public comments provided that

12   are in the members' packets and that are on the table.

13	I thank all the members of the committee

14   for being here and for their work for the panel, and

15   members of the public and the agency for the

16   preparations, and I'd like to turn the meeting over to

17   Dr. Vu for her welcoming remarks, and then to Dr.

18   Henderson.

19	DR. VU:    Thank you, Angela.  Good

20   morning.  I'd like to welcome everyone to the public

21   meeting of the Clean Air Scientific Advisory Committee.

22   The topic of today's meeting involves the review of the

23   Agency's second draft/Integrated Science Assessment,

24   the ISA, and the first draft of -- EPA's first draft on

25   the Risk and Exposure Assessment for the Oxides of

1  Advisory Committee, and we greatly appreciate receiving

2  comments from interested members of the public, as well

3  as those who will be presenting their public comments

4  today -- I guess this morning and this afternoon, for

5  your consideration.  It's important that the CASAC

6  will have opportunity to review the comments of the

7  public and incorporate into their deliberations for

8  your advice to the Administrator.

9	With that I'd like to, you know, give

10   the floor to Dr. Rogene Henderson.  But one final

11   comment is I'd like to thank Dr. Angela Nugent --

12   excellent DFO for this panel, and appreciate Angela and

13   all the personnel for contractors who support the

14   meeting logistics.

15	Dr. Rogene Henderson.

16	DR. HENDERSON:	Thank you, Vanessa.

17   And welcome again to this meeting for the NOx health

18   panel.  I really appreciate all the hard work each of

19   you are doing on this, and I think it is -- it shows

20   the hard work you're doing in the documents that we're

21   reviewing today, in that I see in the ISA that a

22   product that I think has been much improved by the EPA

23   staff using the advice of CASAC.  So, we'll look

24   forward to looking at that document later on.

25	I'd like, now, for everyone to go around

EPA CASAC MEETING 05/01/08 CCR#15905-1	3

6	8

1  the table and introduce yourself; saying just your name

2  and where you're from.  And then we'll be sure and

3  include those who are on the telephone to make sure

4  we're aware that they're there.

5	Terry, would you like to start and go

6  around that way?

7	DR. GORDON:    Terry Gordon from NYU

8  School of Medicine.

9	DR. AVOL:    Ed Avol from University of

10   Southern California School of Medicine.

11	DR. HATTIS:    Dale Hattis, Clarke

12   University.

13	DR. PINKERTON:    Kent Pinkerton,

14   University of California Davis.

15	DR. SHEPPARD:    I'm Lianne Sheppard,

16   University of Washington.

17	DR. POSTLETHWAIT:    Ed Postlethwait,

18   University of Alabama, Birmingham.

19	DR. THURSTON:    George Thurston, NYU

20   School of Medicine.

21	DR. SAMET:    Jon Samet, Johns Hopkins

22   Bloomberg School of Public Health.

23	DR. HENDERSON:    Okay.  And, if the EPA

24   staff could introduce themselves.

25	DR. LUBEN:    Tom Luben with NCEA.

1	DR. HENDERSON:	Welcome, John.

2	DR. SCHLESINGER:    Rich Schlesinger.  We

3  can hardly hear you.

4	DR. LARSON:    Tim Larson here.

5	SPEAKER:    No problem here on our end.

6	DR. HENDERSON:    You're coming through

7  loud and clear.

8	DR. NUGENT:	Rich, this is Angela.  I

9  thank you and Tim

10   for sending me an email saying you're having trouble

11   hearing.  Please keep on sending those emails, and

12   we'll work through the problems.

13	I asked everyone here in the room to

14   speak into their mics.  That will help the people on

15   the phone to hear, and also the people on the phone to

16   turn off their speaker phones.  Or if they're going to

17   use their speaker phones, put it on mute.  That will

18   help -- the engineers tell us that will help.

19	DR. BALMES:    So, Angela, you just came

20   across fairly clearly.  You must be talking right into

21   the microphone, but most of the people introducing

22   themselves did not.

23	DR. HENDERSON:	Okay.  So that's our

24   lesson.

25	DR. KINNEY:    This is Pat Kinney at



7	9

1	DR. MENG:    Qingyu Meng, EPA NCEA.

2	DR. KOTCHMAR:    Dennis Kotchmar, EPA.

3	DR. HENDERSON:	And then just come up.

4	DR. ULTMAN:    Jim Ultman, Penn State

5  University.

6	DR. SEIGNEUR:    Christian Seigneur,

7  Atmospheric and Environmental Research.

8	DR. RUSSELL:    Ted Russell, Georgia Tech.

9	DR. KENSKI:    Donna Kenski, Lake Michigan

10   Air Directors Consortium.

11	DR. CRAPO:    James Crapo from the

12   National Jewish Medical and Research Center.

13	DR. WYZGA:    Ron Wyzga, Electric Power

14   Research Institute.

15	DR. KLEEBERGER:    Steve Kleeberger,

16   National Institute of Environmental Health Sciences.

17	DR. HENDERSON:	Thank you, all, very

18   much.  And, now, we'll turn to an introduction of this

19   draft ISA.

20	SPEAKER:	We haven't introduced the

21   people on the phone.

22	DR. HENDERSON:	We forgot the people on

23   the telephone.  Sorry.

24	Okay, who's on the phone?

25	DR. BALMES:    John Balmes.

1  Columbia.

2	DR. HENDERSON:	Thanks, Pat.

3	DR. SCHLESINGER:    Yeah, Rogene, we can

4  hear you very well -- that's about it.

5	DR. HENDERSON:	Who can you hear?

6	DR. BALMES:    You and Angela came across

7  very well.

8	DR. HENDERSON:	Well, okay.  We'll just

9  all speak up then.

10	DR. NUGENT:	So, on the phone there's

11   John and Pat and

12   Rich Schlesinger.  Anyone else on the phone?

13	DR. LARSON:    Tim Larson.

14	DR. HENDERSON:	Okay, now we can go and

15   hear from the -- our EPA folk.  Ila, are you starting?

16	DR. COTE:	Yeah.

17	DR. HENDERSON:	There you are.  You

18   were standing right there.

19	DR. COTE:	How about --

20	DR. HENDERSON:	It's for this

21   presentation.  And I would like to say,  I was quite

22   impressed with your -- the improvement in this.

23	DR. COTE:	Thank you.

24	How's the sound level for people on the

25   phone?



EPA CASAC MEETING 05/01/08 CCR#15905-1	4

10	12

1	SPEAKER:	Good.

2	DR. COTE:	Okay.  We're going to spare

3  you the overview slide that you've seen 37,000 times at

4  this point, and we're going to jump right into the

5  charge questions.  The way we've structured this talk

6  is to put up the charge questions and then talk a bit

7  about each one.

8	The first charge question was focused on

9  the characterizing the search strategy for identifying

10   literature criteria for study selection and the

11   framework for evaluation studies and causal

12   determination.  Appendix -- or Annex A lays out the

13   search strategy for the literature and criteria for

14   study selection, as well as talks about the framework,

15   but I think the framework is the most important part of

16   the charge question, and so I'm going to focus on that

17   particular aspect.

18	As we had talked about at the last time

19   we all met, we had a framework and some guidelines that

20   we were using internally for making decisions within

21   the document and to ensure we had consistency in the

22   document.  We hadn't articulated those in the document

23   itself and made those explicit.  So the committee I

24   think quite strongly felt that that needed to be done,

25   and we agreed.  So we've made a pretty significant

1  written on causal determination.

2	Next slide.

3	Here are the factors for judging

4  causality .  These are essentially modified Hill

5  criteria that we're using.  As all of you know, the

6  Hill criteria are used widely throughout science and is

7  a common feature of all the documents that I've just

8  mentioned, as well as the Agency's guidelines on

9  cancer.

10	Next slide.

11	There's a two-step process that these

12   documents use.  One can think of it as a hazard I.D.

13   dose response, although that's a little simplistic, but

14   it's not -- it's a useful construct.  The first step is

15   to determine causality.  Can this chemical cause this

16   effect.  In the most simplistic form you think of it --

17   can it cause this effect irrespective of the

18   concentration of exposure.  I think that concept has

19   been modified in the scientific community by

20   essentially a consensus that we're not using extremely

21   high concentrations, but focusing on the lower end for

22   the dose response rate when we talk about causality.

23	We're using a five category descriptor

24   for causality.  Sufficient to infer a causal

25   relationship.  Sufficient to infer a likely causal



11	13

1  effort to incorporate more explicitly this.

2	We -- you know, the other thing I want

3  to mention is if for those of you that are familiar

4  with the smoking and health report, as well as the

5  recent Institute of Medicine report, you will see a

6  great deal of similarity in the language.  We tried to

7  either copy or plagiarize language so we had a

8  consistency across these documents rather than

9  reinventing another description of how to do this.

10   Both of these efforts a lot of thought and effort went

11   into them and so we thought it would be -- I thought it

12   would be a mistake to kind of invent new language.  So,

13   if you see a similarity in language, you're on your

14   toes.

15	But, anyway, the intent of this was to

16   establish a more uniform language concerning causality

17   and bring a greater specificity and I would say rigor

18   to our findings.  This helps it to -- we feel this

19   would help us to assess both the separate and combined

20   lines of evidence, as well as classify and characterize

21   the data to evaluate causality.  And as I noted in this

22   last slide, we've drawn heavily on the Surgeon

23   General's smoking reports, as well as the IOM document.

24   The IOM document is very new.  It's probably the most

25   sophisticated of the various documents that have been

1  relationship.  Suggestive, but not sufficient,

2  inadequate, or suggestive of no causal relationship.

3	The smoking and health document, as well

4  as the IRM document, use a four category, can we have a

5  pointer, okay, we use a four category scheme with this

6  -- and they don't include the sufficient to infer a

7  likely causal relationship. So that is not included in

8  the smoking and health and the IRM documents.  There

9  are two reasons that we put it in.  One, we liked it.

10   It felt better to us in the sense of the gap between

11   suggestive and sufficient to infer a causal

12   relationship seemed very broad to us, and we wanted a

13   more nuanced kind of description.

14	This five category scheme is also

15   consistent with what the Agency has articulated in the

16   guidelines for cancer risk assessment.  Thank you.  And

17   so at least by using the five category scheme, we are

18   -- EPA is internally consistent and it is consistent

19   across cancer and non-cancer in terms of the descriptor

20   -- the five category set of descriptors.

21	The second sub -- once you've made some

22   decision about what you think the nature of the causal

23   relationship is -- is the dose response part.  And so

24   if evaluating the population dose response, the shape

25   of the dose response, susceptibility, those factors



EPA CASAC MEETING 05/01/08 CCR#15905-1	5

14	16

1  come into consideration.

2	For those people who are

3  epidemiologists, you will instantly recognize that

4  epidemiologists often are sort of  doing both of these

5  things at the same time.  Where the experimental

6  community is often kind of doing them as a clearer one,

7  two step process.

8	Next slide.

9	And so that's essentially what we've

10   done on the framework, and we've articulated it in the

11   document.  And I would really love to have -- this is

12   going to be the framework that we use for every one of

13   our documents.  So it's really important and we want to

14   get it right, so I'm really looking forward to quite a

15   bit of discussion on the framework.

16	So what we're going to do next is this-

17   is the team of people that's worked on the NOx tox

18   document, and Tom Luben who -- one of our crack

19   epidemiologists, is now going to talk about if you take

20   the framework and you apply it to the available

21   evidence, what are your conclusions.

22	Tom?

23	DR. KOTCHMAR:	Actually, we're going to

24   have Dr. Meng talk about the --

25	DR. COTE:	Oh, I'm sorry.

1  here are related to the key question, how well can we

2  use the ambient concentration as a surrogate for

3  population exposure?  Also, these issues can help us

4  better understand health sciences findings.

5	Next, please.

6	Here comes the first issue, ambient

7  measurement methods and ambient concentrations, there

8  are two key points associated with this issue.  The

9  first one is ambient level of  N02, and the second one

10   is the errors associated with the ambient measurement

11   method.

12	In response to CASAC comments, we

13   provided the information on the distribution -- the

14   trends -- the NO2, Ambient NO2, across the United

15   States as shown in the figure on the left.  The solid

16   white line is the mean of the concentration, and the

17   bottom upper borders of the blue area are the tenth and

18   90's percentile of the distribution.

19	In the past 26 years, the Ambient NO2

20   concentrations have been declining.  The figure on the

21   right shows the distribution of Ambient NO2 measured in

22   all the metropolitan statistical areas with NO2

23   monitors from 2003 to 2005.  The current Ambient

24   concentration, annual concentration is about 15 ppb,

25   which is well below the current National Ambient Air



15	17

1	Tom will talk about applying the

2  framework, but Dr. Meng is going to talk next on the

3  atmospheric/chemistry exposure.

4	DR. KOTCHMAR:	And, also, this

5  represents the scientific staff from the office that's

6  worked on the document, but obviously we have support

7  staff producing the documents and various expert

8  contractors.  But at this time, let's move right along

9  with Dr. Meng, and then Dr. Luben are going to present

10   some of the background information on the revised

11   document.

12	DR. COTE:	Do you want to stand or do

13   you want to do it

14   from there?

15	DR. NUGENT:	And a reminder, please to

16   -- pull the mic

17   fairly close and speak into the mic, please.

18	SPEAKER:	We can't hear.

19	DR. MENG:    I haven't spoken yet.

20	Good morning.  This is Qingyu Meng from

21   NCEA.  I'm going over Chapter 2, which is covered by --

22   next please -- which is covered by Charge Question 2.

23   Next, please.

24	I'm not trying to give a comprehensive

25   review of Chapter 2, but three issues I'm talking about

1  Quality Standard, which is .053 ppm or 53 ppb, but the

2  one hour maximum concentration could be a three or even

3  four times higher than the annual concentration.

4	We should always keep in mind that these

5  numbers are associated with positive interferences

6  caused by nitric acid and PANs.  So in response to

7  CASAC comments, we extended our discussion on this

8  positive interferences issues.  Since the interferences

9  -- the positive interferences are larger during the

10   summer than winter, and also larger in remote sites

11   than the sites close to NO2 sources on a routed basis.

12	Next, please.

13	The second issue is spatial and temporal

14   variation in Ambient NO2 concentrations.  The questions

15   here are can we capture the spatial and temporal

16   variation with a current NO2 Ambient monitoring

17   network?  If the answer is yes, how does the spatial

18   variation of NO2 look like in the urban scale, and how

19   does spatial variation of NO2 look like in a near road

20   small scale.

21	The table here shows the, in response to

22   CASAC comments, actually we provided the information on

23   the distribution of NO2 monitoring network across the

24   United States -- with the declining of NO2

25   concentrations, the number of NO2 monitoring size has



EPA CASAC MEETING 05/01/08 CCR#15905-1	6

18	20

1  been decreasing.  So, very few cities have more than

2  five NO2 monitoring monitors.  So the table shows here

3  are the spatial variation of NO2 in urban scale.

4  Cities listed here are the cities with more than five

5  NO2 monitors.  We can see from the table that the

6  spatial variation of NO2 can be quite large in an urban

7  scale.

8	"R" is the -- the column "R" shows the

9  range of Pierson  correlation coefficient between NO2

10   concentration between each individual part --

11   individual site pairs across all possible pairs of

12   sites.  And the mean concentration of that column shows

13   the overall mean of NO2 in an urban site, and also, in

14   the parentheses, shows the range of NO2 concentration

15   at each individual site across all sites.  We can see

16   the mean concentration at each individual site varies

17   by a factor of 1.5 to 6.

18	In response to CASAC comments, you also

19   provided the, and illustrated with new figures the

20   spatial -- the large scale spatial NO2 concentration

21   across the United States and also a seasonal and

22   diurnal variations of NO2.  Also for the small scale

23   near-road variations, we just illustrated with new

24   figures that people near-road or on a road can be

25   exposed to elevated NO2.

1  that's what I just said -- infiltration factor and

2  Alpha are very important to understand Ambient

3  contribution to indoor and personal exposure.  The

4  infiltration factor ranges from .4 to .7 and Alpha

5  ranges from .3 to .6.

6	The last issue in the personal exposure

7  assessment is the multi pollutant personal exposure

8  issues.  Personal exposure to NO2 can be associated

9  with personal exposure to other co-pollutants due to

10   the common sources like traffic, and also due to

11   chemical reactions in the Ambient air and in the indoor

12   air.

13	With that, I'll turn it to Dr. Luben.

14	DR. LUBEN:	Thank you.

15	My name is Tom Luben, and I'm going to

16   talk about Charge Question 3 through 5, which cover

17   Chapters 3, 4, and 5.  So in chapters three, four, and

18   five, we really tried to address CASAC comments by

19   tightening up the language, making the format or the

20   style of the chapters more consistent, adding summary

21   sections to the end of each section, and really

22   focusing on the integration.

23	But one of the large overarching topics

24   that we wanted to address was this issue of NO2 as a

25   component of a mixture.  We know that NO2 does not



19	21

1	Next, please.

2	The third issue is the association

3  between the personal exposure and Ambient

4  concentrations.  There are two key points in this

5  issue.  One is the Ambient contribution to personal

6  exposure, the other is the correlation between the

7  personal exposure and Ambient concentration.  During

8  the last review CASAC suggested us to expand our

9  discussion on the relationship between the

10   indoor/outdoor and personal NO2 concentrations with an

11   emphasis on two key parameters.  One, is the

12   infiltration factor, the other is Alpha or Ambient

13   exposure factor.  And also CASAC suggested us to

14   reorganize the discussion of the correlation

15   coefficient between personal exposure and Ambient

16   concentration to reflect the exposure study features in

17   the context of different epidemiological studies.

18	So we did that, and the first -- shown

19   here are the figures 2.5 for A and 2.5 for B on page

20   241.  Basically, we calculate -- recalculated the

21   correlation coefficient and the confidence interval for

22   each study based on the feature of the exposure study

23   -- different exposure study designs.  And also we

24   interpreted this correlation coefficient in the context

25   of different types of epi studies.  Two parameters --

1  occur by itself in the atmosphere, and when we're

2  looking at epidemiologic studies we're not looking at

3  NO2 by itself.  We're looking at NO2 and people are

4  being exposed to NO2 and this combustion related

5  pollutant mixture, as we wanted to discuss this

6  throughout the document.

7	We believe that there are effects,

8  health effects, adverse health effects associated with

9  NO2, and it's really difficult to quantify these

10   effects and also to disentangle the effects that are

11   attributable directly to NO2 by itself compared to

12   those that are attributable to the pollutant mixture.

13   Still, we believe that there is enough evidence to

14   support an independent effect of NO2, and the reason we

15   believe this is because we believe the results for the

16   epidemiologic studies are consistent and coherent and

17   that the animal toxicological studies and human

18   clinical studies provide biological plausibility for

19   these respiratory effects.  We also believe that the

20   addition of other pollutants and multi-pollutant models

21   provide further evidence of an independent effect of

22   NO2.

23	If I could have the next slide, please.

24	So this is a summary figure that we've

25   been working on to kind of address the range of



EPA CASAC MEETING 05/01/08 CCR#15905-1	7

22	24

1  respiratory effects that we see.  And if you start on

2  the left, we were looking at respiratory symptoms here.

3  Let me mention that the studies that are in black are

4  U.S. and Canadian studies.  The other colors are

5  studies outside the U.S. or Canada.  All the red

6  studies are the respiratory symptoms, and if we look at

7  all the epidemiologic studies that looked at

8  respiratory symptoms, we see that all of the central

9  estimates are above one.  So we think that this shows

10   consistency in the studies.  Some of them are

11   statistically significant, some are not, but overall,

12   they are showing a positive effect with respiratory

13   symptoms and exposure to NO2.

14	As we kind of progress along and move

15   over towards the right, we look at the emergency

16   department visits and hospital admissions. Again, the

17   majority of the studies here have a positive central

18   estimate.  Again, some are statistically significant,

19   some are not, but by-and-large, we believe that this

20   shows consistency in the evidence that epidemiologic

21   studies are showing an effect of hospital admissions or

22   ED visits among adults, children, older adults, for all

23   respiratory diseases and for asthma.

24	DR. HENDERSON:	For those people on the

25   phone, can you tell them what page in the document this

1  before, that looking at co-pollutant or multi-pollutant

2  models are not the best tool that we have to try to

3  disentangle the independent effects of the criteria

4  pollutants from one another, but right now, they're the

5  best tool that we have.  They're available in much of

6  the published literature.

7	This slide looks at the addition of PM

8  species, different size fractions of PM into the model

9  with NO2, and we see that the central estimates do not

10   move very much at all.  There's one case, the Yang

11   study that was in Taiwan where the addition of PM 10

12   increased the central estimate.  That was only when --

13   during days when the temperature was above 25 degrees

14   Celsius.  Other than that, the studies that looked at

15   the addition of PM to the model, including NO2, did not

16   show any significant change leading us to believe that

17   there is an independent effect of NO2.

18	Next slide, please.

19	So with that I'm going to kind of show

20   you the key conclusions that we draw from the document.

21   The key conclusions are the same as what we had in the

22   first draft.  We have changed the language to kind of

23   bring it back to the framework that Dr. Cote was

24   talking about at the beginning.  For respiratory

25   morbidity with short term exposure, we believe that the



23	25

1  graph is on?

2	DR. LUBEN:	It's figure 5.1, so it

3  would be the first figure in Chapter 5.

4	DR. HENDERSON:	Okay.

5	DR. LUBEN:	I can have a page number in

6  just a minute.  I'm sorry.  Page 5.9, thank you.

7	As we progress all the way to the right,

8  we look at the studies that looked at respiratory

9  mortality.  So these are studies that looked at

10   mortality that could be attributable to respiratory

11   disease.  Again, we find that the majority of these

12   studies have a positive central estimate.  Not as many

13   are statistically significant, but we still believe

14   that this shows a consistent positive effect with NO2

15   and respiratory mortality.

16	When we look at all of these together,

17   we believe that this shows a coherence across the

18   severity of respiratory effects, starting with

19   respiratory symptoms, progressing through hospital

20   admissions or ED visits and then ending unfortunately

21   in respiratory mortality.

22	Next slide, please.

23	We also -- excuse me -- we also looked

24   at the addition of co-pollutants into the models.  We

25   know that, and we've talked about it here at CASAC

1  evidence is sufficient to infer a likely causal

2  relationship.  For cardiovascular morbidity and short

3  term exposure, we believe that the evidence is

4  inadequate to infer the presence or absence of a causal

5  relationship.  And for short term mortality, we find

6  that the evidence is suggestive, but not sufficient to

7  infer a causal relationship.

8	When we move on to long term exposure,

9  we found that for respiratory morbidity the evidence is

10   suggestive, but not sufficient to infer a causal

11   relationship.  For other morbidity, including pre-natal

12   and birth outcomes, as well as cancer, we found that

13   the evidence was inadequate to infer the presence or

14   absence of a causal relationship.  And finally, for

15   long term mortality, we also found that this evidence

16   was inadequate to infer the presence or absence of a

17   causal relationship.

18	I hope that we have sufficiently

19   addressed CASAC's comments and we look forward to

20   discussions this afternoon.  Thank you.

21	DR. HENDERSON:	Thank you.  Is that

22   everything on the --

23	DR. KOTCHMAR:	This concludes our

24   presentation.

25	DR. HENDERSON:	This concludes.  Okay.



EPA CASAC MEETING 05/01/08 CCR#15905-1	8

26	28

1  Thank you, very much.  That was very helpful.

2	So we can now move to the public

3  comments; correct?  I'm sorry, we'll turn the mic over

4  to Angela, who will facilitate the public comment

5  period.

6	DR. NUGENT:	I have not greeted

7  personally our public commentors.  Is Dr. Long in the

8  room?  Welcome, and Dr. Vu is offering you her seat, so

9  thank you, Vanessa.

10	I'd like to introduce Dr. Christopher

11   Long from Gradient Corporation.  And you have

12   presentation slides?

13	DR. CHRISTOPHER LONG:    I do.  Yeah, I

14   think.  Yeah, there they are.

15	My name is Chris Long.  I'm a senior

16   scientist at Gradient Corporation in Cambridge,

17   Massachusetts.  I'm presenting comments on behalf of

18   the utility air regulatory group.

19	Next slide, please.

20	Our comments deal principally with

21   Chapter 5 in the NOx ISA.  And the overall theme of our

22   comments is that Chapter 5 does not provide sufficient

23   integration and analysis of the different lines of NOx

24   health effects evidence.  And, you know, in our written

25   comments, we support this theme with several different

1  subject specific correlations are negative and they

2  highly -- you know, they highly vary and overall the

3  correlations are very weak.  These studies show that

4  there are really no improvements in these correlations

5  when subjects were stratified by the presence of indoor

6  sources and by home ventilation.  Generally, these

7  studies showed much stronger personal to Ambient

8  correlations for other pollutants, such as fine

9  particles.  And interestingly they observed a good

10   correlation between Ambient NO2 levels and personal PM

11   2.5 concentrations.

12	You know, based on study findings such

13   as these, it can be asked how the epidemiological

14   evidence can be judged to be consistent and coherent

15   when there remains large uncertainty regarding the

16   validity of the exposure metric.

17	Next slide, please.

18	In addition, recent studies provide

19   compelling evidence for Ambient NO2 acting as a

20   surrogate.  These are our figures from the Brook, et

21   al. publication, and, you know, the two figures show

22   strong correlations between Ambient NO2 concentrations

23   and concentrations of a number of gaseous and

24   particulate phase species.  Including in the left

25   panel, it shows strong correlations between Ambient NO2



27	29

1  lines of analysis, including number one, a figure, you

2  know, by disregarding the NO2 concentrations at which

3  health effects associations have been observed, figure

4  5.3-1 gives an incomplete and misleading picture of the

5  epidemiological evidence for short term exposure NO2

6  health effects.

7	Secondly, the association between

8  Ambient NO2 concentrations and personal NO2 exposures

9  is complex and remains poorly understood raising

10   questions regarding the proper interpretation of the

11   reported NO2 epidemiologic associations.

12	Number three, USEPA still does not

13   sufficiently consider the fact that NO2 may be acting

14   as a surrogate for other pollutants.  And lastly, USEPA

15   should quantitatively contrast the dose levels typical

16   of Ambient NO2 epidemiological studies versus those

17   used in human controlled exposure studies.  In the next

18   few minutes I'll develop a few of these ideas.

19	Next slide, please.

20	This figure is from a recent publication

21   by Sarnett, et al., and it illustrates the idea that

22   Ambient NO2 and personal NO2 correlations are poor and

23   vary widely.  This figure shows subject specific

24   personal to Ambient NO2 correlations for four U.S.

25   cities.  And generally, you can see that a lot of the

1  and acid-aldehyde, as well as for b-tex compounds.  And

2  in the right panel, it shows very strong correlations

3  between Ambient NO2 and Ambient PAH concentrations;

4  including specifically benzo(e)pyrene in the figure.

5	You know, based on correlations such as

6  these, Brook, et al. concluded that the strong effect

7  of NO2 that they observed in Canadian Time Series

8  Mortality Studies could be due to its serving as a good

9  surrogate for other toxic air pollutants.

10	Next slide, please.

11	You know, as discussed a few minutes

12   ago, the ISA seems to dismiss the NO2 surrogate idea

13   based on the conclusion that NO2 associations generally

14   remained robust in two pollutant models.  But in

15   general, you know, the multi-pollutant model results

16   are very limited and conflicting. You know, of the two

17   pollutant model results provided in the two figures in

18   the ISA, only one of the models adjusting for particle

19   concentrations adjusted for fine particles with most

20   adjusting for PM10.  Only two of the studies included

21   adjustments for a gaseous pollutant other than ozone or

22   SO2.  None of them adjusted for aldehydes, PAHs, or

23   particle-bound organics.  And none of the cited studies

24   are for U.S. locations.

25	You know, recently, several published



EPA CASAC MEETING 05/01/08 CCR#15905-1	9

30	32

1  multi-pollutant model results, including those of

2  Tolbert, et al., McKreaner, et al., and Delphino, et

3  al., they contradict the EPA conclusion that Ambient

4  NO2 is robust in multi-pollutant models.

5	Next slide, please.

6	And lastly, you know, as shown in this

7  figure, epidemiologic associations are generally

8  reported for NO2 doses far below human clinical

9  toxicology no effect levels.  In this figure the

10   colored symbols represent different types of

11   epidemiology studies that used one hour maximum NO2

12   concentrations as exposure metrics.  The dotted lines

13   represent no effect levels from clinical toxicology

14   studies, and in general they show a discordance, you

15   know, that still remains between the human clinical

16   toxicology and the epidemiological  associations.

17	And last slide, please.

18	A few recommendations for EPA.  Merely

19   acknowledging uncertainties is not sufficient.

20   Uncertainties must be quantified and affect the weight

21   that is placed in particular study findings or

22   particular lines of evidence.  Secondly, the supportive

23   or non-supportive role of clinical and experimental

24   studies at the specific Ambient concentrations in

25   question should be more directly addressed.  And

1	I thought the slide you presented showed

2  a grouping of clinic, of epidemiological studies where

3  the concentrations were below a no effect level in the

4  clinical study.  So how did you determine when there

5  was no effect in the clinical studies -- that's what

6  I'm asking?

7	DR. CHRISTOPHER LONG:    Yeah, actually, I

8  guess I should rephrase.  Those aren't -- those are the

9  doses at which effects were observed, so those would be

10   sort of the lowest.

11	DR. ULTMAN:    Oh, I see.  Okay.

12	DR. CHRISTOPHER LONG:    I'm sorry, if

13   that was confusing.

14	DR. ULTMAN:    So, it wasn't a specific

15   NOAEL that was found?  Okay.

16	DR. CHRISTOPHER LONG:    No, no.  I'm

17   sorry.

18	DR. HENDERSON:	Okay. If there's no

19   more questions.

20	DR. NUGENT:    Our next public speaker is

21   on the teleconference line.  It's Mr. John Heuss.  I

22   hope I'm saying that right.  From the Air Improvement

23   Resources -- Resource Inc., on behalf of the Alliance

24   of Automobile Manufacturers.

25	Are you on the line, please?



31	33

1  lastly, Chapter 5 needs to be less of an introduction

2  of ideas and recitation of selected study findings and

3  more of an integrated synthesis that can inform policy

4  makers.

5	Thank you for your attention.

6	DR. HENDERSON:	Are there questions

7  from the panel for Dr. Long?

8	DR. ULTMAN:	The dotted line in the no

9  effects level; where did that come from?

10	DR. CHRISTOPHER LONG:    That came from --

11   actually, from chapter 5 where they discuss, you know,

12   findings from different human clinical toxicology

13   studies.  And, you know, basically, I took the

14   concentrations, the exposure durations and calculated

15   the NO2 dose.  Those come directly from chapter 5.

16	DR. ULTMAN:	When did you determine the

17   zero effect?

18	DR. CHRISTOPHER LONG:    I don't believe

19   there was a zero effect dotted line.

20	DR. ULTMAN:    I thought you concluded

21   that there was no effect, the epidemiological studies

22   there was no effect

23	DR. HENDERSON:	Jim, you need to speak

24   in the mic if you want the

25	DR. ULTMAN:    Oh, I'm sorry.

1	MR. HEUSS:	Yes, I am.

2	DR. NUGENT:	Thanks.  Speak up.  And I

3  think we can hear you in the room.

4	MR. HEUSS:	Thank you.

5	My name is John Heuss with Air

6  Improvement Resource.  We provided detailed written

7  comments earlier this week.  We noted a number of

8  changes and improvements in the second draft,

9  particularly commending the introduction of the

10   application of the framework for making causal

11   determinations.  However, we believe the following

12   areas can be improved through the continued attention

13   of staff and CASAC.

14	First, although the framework is

15   generally applied throughout the second draft, its

16   application is not as rigorous or complete as it should

17   be.  Particularly in the way consistency is evaluated

18   in the acute respiratory epidemiology is less than

19   scientifically rigorous.  The ISA notes that there are

20   about 50 publications in this area.  Only selected

21   single pollutant results are included in the main data

22   presentations in the figure.  However, the ISA also

23   notes there are additional studies that show negative

24   or null results, and notes still others that are

25   characterized as studies that could not inform the



EPA CASAC MEETING 05/01/08 CCR#15905-1	10

34	36

1  associations.  The rationale for why some studies are

2  highlighted in the text and included in figure 5.3-1;

3  others are relegated to tables in the annex is not

4  clear.

5	All these studies implicate pollutants

6  other than or in addition to NO2, but this fact is not

7  discussed adequately in the ISA.  Since NO2 occurs in

8  conjunction with other common air pollutants, issues of

9  confounding and surrogacy plague the interpretation of

10   this literature, therefore it's particularly important

11   to carefully consider the results of the controlled

12   studies throughout the ISA.

13	Second, the draft still focuses on

14   single-pollutant model results rather than evaluating

15   the results in the context of the full suite of air

16   pollutants.  This can lead to double or triple counting

17   of health effects as different pollutants are reviewed.

18   In the recent PM and Ozone reviews, single-pollutant

19   model results were used to show the strength and

20   consistency of the association and used in the risk

21   assessment.  However, single-pollutant models are known

22   to be biased high.  If selected single-pollutant model

23   results are also used to claim the same health effects

24   caused by NO2, it will be a clear case of double or

25   triple counting.

1  approach is insufficient to establish consistency or

2  coherence.  A more holistic and rigorous evaluation of

3  this observational literature is needed if double and

4  triple counting of health effects is to be avoided.

5	Biological plausibility involves two

6  considerations, both the effects the pollutant can

7  cause and then the concentrations that can cause the

8  effects.  All these effects are non-specific.

9  Questions of NO2 acting as a surrogate are prevalent,

10   therefore the ISA should address the plausibility for

11   NO2 along with the plausibility for other anthropogenic

12   and natural materials in the air causing these various

13   potential health effects.

14	Fourth, the ISA mischaracterizes the epi

15   findings in the prior review when it indicated, quote,

16   "The main conclusion was that there was insufficient

17   epidemiologic evidence for an association between short

18   term exposure and health effects," unquote.  A 1995

19   staff paper did not indicate there was insufficient

20   evidence for an association, rather it noted issues

21   that limit the use of the available associations in

22   evolving a basis for the NAAQS..  Our written comments

23   document the errors.  By mischaracterizing the extent

24   and interpretation of epi studies in the previous

25   review, the draft now sets up a false comparison



35	37

1	Third, the ISA should further address

2  the issues of publication bias and model selection

3  uncertainty that hinder the interpretation of these

4  studies.  During the ozone review, CASAC pointed out

5  where systematic analysis have been carried out as in

6  NNMAPS, Steve, et. al, 2002 and 3 and Ito, 2003.

7  Similar patterns of associations are reported for many

8  pollutants. While there are many more observational

9  studies and available in the prior review, there's an

10   implausibly wide-range of results from positive to

11   negative in systematic analysis.  This is not just

12   heterogeneity since there's substantial portion of

13   negative associations.  The ISA needs to acknowledge

14   and consider the wide-range of associations with regard

15   to both biological plausibility and the limitations on

16   the use of time series studies to set Ambient

17   standards.

18	One implication of the variability

19   documented in our written comments is it's not

20   surprising to find some positive NO2 associations in

21   the literature for any health endpoint that's

22   evaluated.  Even for endpoints where there is no

23   underlying effect.  This raises a serious question

24   about the approach in the ISA of documenting any and

25   all NO2 associations in the literature.  Such an

1  between the state of knowledge in the previous review

2  and that in the current review.

3	Fifth, the conclusions in section 5.4

4  for acute respiratory effects are overly broad.  While

5  there is evidence for acute respiratory effects from

6  NO2, the evidence for which there was strong causal

7  support is similar to that in the last review.  The

8  only potential respiratory health effect for which

9  evidence is markedly different in the current review is

10   emergency department visits and hospital admissions for

11   respiratory causes.  However, as detailed in our

12   written comments, the pattern of results is implausibly

13   wide, similar to that for other pollutants, the results

14   are inconsistent for specific respiratory diseases, and

15   the authors of these studies themselves do not focus on

16   NO2.  All this makes the assumption of likely NO2

17   causality highly suspect.

18	Thank you.

19	DR. HENDERSON:	Are there any questions

20   for Dr. Heuss?

21	Yes, Frank Speizer.

22	DR. SPEIZER:    Dr. Heuss, you -- this

23   table -- this figure, 5.3-1 is quite full as it is, and

24   I think staff would be greatly helped if you could

25   indicate specific articles that you think have been



EPA CASAC MEETING 05/01/08 CCR#15905-1	11

38	40

1  left out that would contribute to the, a better view of

2  this picture.

3	Now, I don't -- I tried to look through

4  your comments, but I couldn't quickly figure out

5  whether you had indicated the specific articles that

6  you think have been left out.

7	MR. HEUSS:    Well, I think there's a

8  whole list of articles in the ISA where it says could

9  not inform for both asthma and all respiratory, so

10   those obviously are there.  And I guess one of the

11   things I did was, for example, used Ross Anderson's

12   1998 article where he specifically looked at 15 earlier

13   articles and studies, and concluded they weren't

14   consistent.  Some of them showed ozone effects and some

15   did not, some showed NO2 effects, some did not, some

16   showed PM effects, some did not, et cetera.  He looked

17   at it more holistically at that point in time with 15

18   studies and concluded that, you know, there were

19   associations because there's different age categories

20   and different respiratory categories.  So I used those

21   sorts of things.  But I think, you know, if there are

22   indeed 50 studies, then, you know, they really ought to

23   look at all of them and try to figure out not only are

24   there NO2 associations, but what those authors actually

25   concluded relative to air pollution or PM or Ozone or

1  the first external draft.  At that time, API provided

2  several recommendations to improve the ISA.  Subsequent

3  API followed up with more detailed written comments

4  filed to the docket.  API's objective was to have the

5  NOx ISA accurately reflect the current state of the

6  science.

7	As Dr. Henderson and others have already

8  mentioned today, the second external draft is much

9  improved.  While we find that there are many of the

10   concerns identified by API that are now included, the

11   connections -- the corrections are mentioned but are

12   not necessarily in our opinion given the proper weight

13   as conclusions are developed.

14	Specifically, the ISA relies extensively

15   on epidemiological studies that suffer from a number of

16   significant limitations making them an insufficient

17   basis for ISA's conclusions regarding association

18   between NOx exposure and health effects.  I think

19   you've already heard that comment today.  Let me save

20   you time -- so I'm not repeating things.

21	Yeah, I guess, the other thing we wanted

22   to talk about is to -- I guess first of all, we will

23   follow up with more detailed written comments to the

24   docket on the issues that we found in the secondary

25   external review.



39

1  NO2 or whatever.

2	DR. HENDERSON:	Are there any responses

3  to that from EPA that you want to make?

4	DR. LUBEN:	In the ISA when it says an

5  article could not inform the decision, that usually

6  means that the author said there were or were not

7  statistically significant effects but did not present

8  quantitative results.  And with no quantitative results

9  there was no way that we could plot them on the graph.

10	DR. HENDERSON:	Thank you.  Any more

11   questions for Dr. Heuss?  Okay.

12	DR. NUGENT:    Our third and last

13   commentor for this morning is Mr. Ted Steichen from the

14   American Petroleum Institute.  I understand he's here

15   to join us and provide comments at the meeting.  Please

16   take a seat.

17	MR. STEICHEN:    Thank you.

18	I'll get close to the microphone here.

19   Good morning.  I'm Ted Steichen, and I'm with the

20   American Petroleum Institute.  API represents about 400

21   members related to all aspects of the petroleum

22   industry.  We appreciate the opportunity to comment on

23   the second external draft of Oxides of  Nitrogen ISA.

24	On October 24 of last year, Howard

25   Feldman of API provided comments before CASAC regarding

41

1	But I did want to talk about our concern

2  with the process for the review of the Oxides of

3  Nitrogen NAAQS.  As many of you know or may know, that

4  API's participated in reviews of the NAAQS for several

5  such reviews from the beginning of each review.  As we

6  all know, EPA has over the years unfortunately not been

7  able to meet its five year deadlines for completing the

8  NAAQS reviews.  API is therefore very supportive when

9  EPA decided, or it was announced by the Deputy

10   Administrator, Marcus Peacock, in December of '06 and

11   modified in April of '07, to the extent that the

12   revised process increases the likelihood that the

13   agency will be able to complete the reviews in a timely

14   manner while maintaining the scientific integrity of

15   the reviews.

16	We are concerned, however, about the

17   application of the process in the present review of the

18   primary Oxides of  Nitrogen.  While the new process

19   sets forth an ambitious time line to complete the NOx

20   review within five years after a workshop is held on

21   the science policy issues, this review for the Oxides

22   of  Nitrogen is being conducted on an even more

23   abbreviated schedule.  EPA's schedule, which we would

24   definitely acknowledge, is driven at least by a consent

25   decree has compressed that review basically to three



EPA CASAC MEETING 05/01/08 CCR#15905-1	12

42	44

1  years.  And most of the time that's being saved is

2  being taken from time that others would be -- that

3  otherwise would be allocated to the critical science

4  and risk exposure assessment processes.

5	The generic NAAQS review process

6  announced by the Deputy Administrator provides 33

7  months from the start of a review until the issuance of

8  a final integrated science assessment.  In this case,

9  however, the agency plans to complete that process in

10   just 17 months according to the integrated review plan.

11   Similarly, while the generic NAAQS review process

12   contemplates 18 months between the release of the

13   Agency's plan for conducting a Risk and Exposure

14   Assessment and the issuance of that final Risk and

15   Exposure Assessment, that process has been reduced to

16   14 months for the current review.

17	The shorter time for the science review

18   in this case is not necessarily cause for concern.

19   However, when the shorter time-line leads to a less

20   than careful presentation of the scientific record,

21   this can cause alarm.  And again, as I think has been

22   indicated, there have been substantial progress made in

23   this second external draft, but I think, you know, in

24   the short term it's like -- I think the first draft --

25   we'd like to see the first draft better, so we can

1  schedule, it was first identifying 17 days for public

2  comment, and we asked for an extension and it was

3  granted for 45 days.  Again, you know, we think 60 is

4  where we'd like to be, and the model schedule would

5  typically allow that.  So, again, recognizing that

6  there are constraints both related to the, you know,

7  consent decree issues, and also, just resources in

8  general, I want to identify that, you know, the better

9  we can see it in  the first round hopefully the more

10   useful our public comments can be, and we can get it in

11   in a time that does not lead us to have certain

12   decisions already identified in future documents.

13	So essentially, that's my statement, and

14   I appreciate the opportunity to speak today and if

15   there were any questions about that I'd certainly try

16   to answer them.

17	DR. HENDERSON:	Are there any

18   questions?

19	Thank you.

20	MR. STEICHEN:    Thank you.

21	DR. COTE:	I'd actually like to make

22   one correction.  It is incorrect that the regulatory

23   evaluation will be done on the first draft.  It will be

24   done on the final draft.  And obviously -- or there

25   would be no point in doing these second and final



43	45

1  provide more useful comment at that time because we

2  recognize that, again, the way the schedule is set up,

3  that the first draft is really informing on the Risk

4  and Exposure Assessment that we'll be talking about in

5  the rest of your meeting, or you'll be talking about

6  the rest of your meeting.

7	The CASAC panel, in fact, concluded in

8  its November 29th letter that the first draft was

9  inadequately described the NOx review process including

10   EPA's approach to literature identification and

11   evidence evaluation, and further noted that the first

12   draft did not appear to have developed by a process

13   that approaches the current state-of-the-art around the

14   development of systemic reviews for decision making

15   purposes.  And, again, the point is not to belabor

16   something that's already been resolved, but it's simply

17   to make the point that with a compressed schedule, it

18   becomes even more critical that the first draft really

19   be of highest quality possible.  And that's really the

20   comment that I'm trying to make here.

21	And finally, we'd just like to identify

22   that we did have a little concern about the opportunity

23   for public comment on the Risk and Exposure Assessment

24   document, which when it was released and that's really

25   the first document that was released somewhat off

1  drafts.  But we communicate this information to OAQPS

2  on a very rapidly and efficiently, and so that their

3  evaluations are done on the, really on the final draft

4  -- the final rulemaking will be done.

5	MR. STEICHEN:    Well, I'm looking at the

6  generic NAAQS schedule.

7	DR. COTE:	You know that -- that

8  schedule when we put that up is a little bit -- as you

9  can see the Risk and Exposure Assessment lags the --

10   lags the ISA.  And, you know, if there's -- if the

11   schedule -- the generic NAAQS schedule doesn't reflect

12   that, please point that out, because we would like to

13   correct that.  But, in fact, the rulemaking will be

14   done on the final document.

15	MR. STEICHEN:    Yeah, and I don't think

16   it's so much of the process is that it's simply that

17   what we're going to talk about in terms of the Risk and

18   Exposure Assessment can only really be identified from

19   the first draft of the ISA, and that's why we're simply

20   making the point that having the ISA a little stronger

21   the first draft --

22	DR. COTE:	That's not correct.  We --

23   there's kind of this continuous flow of information, so

24   they will be using the information the second draft.

25	DR. HENDERSON:	Okay. That is our final



EPA CASAC MEETING 05/01/08 CCR#15905-1	13

46	48

1  question.

2	DR. ROSS:	Rogene, can I also clarify

3  the schedule?

4	DR. HENDERSON:	Yes.

5	DR. ROSS:	A part of the -- this is

6  Mary Ross.  This is Mary Ross with NCEA, and part of

7  the generic process involves approximately sixty days

8  of review for the ISA documents and approximately

9  thirty days of review for the Risk and Exposure

10   Assessments.  And we've been pressed with the schedule

11   and will acknowledge it's a court ordered schedule,

12   it's been difficult, but we have been keeping to

13   approximately that schedule and so that's consistent

14   with a generic process that we have set-out initially.

15	DR. HENDERSON:	Okay. Thank you.

16	MR. STEICHEN:	With the extension, yes.

17   Yeah, we're there.  Thank you.

18	DR. HENDERSON:	Thank you.

19	MR. STEICHEN:	Okay.

20	DR. HENDERSON:	And thank you.

21	Okay.  I appreciate those good public

22   comments.  I heard a lot of comments about the problem

23   of addressing multi-pollutant atmospheres and that's a

24   problem that we all are concerned with, and it's not an

25   easy problem to solve.  It's something that I think

1  of the peer review.  So just a comment, and I think it

2  would be useful to see if others feel the same way.

3	Turning to Charge Question 1, we saw

4  substantial changes in response to comments made with

5  regard to the first review.  The annex provides

6  extensive coverage of other documents.  It probably

7  would be useful to know just how much of that is

8  actually drawn straight from those documents with

9  proper quotation and citation.  I did have a sense of

10   deja vu, in fact, of authorship as I read some of that

11   material.  And I think you probably should be careful

12   just so you don't -- if you will get called-out and I

13   think that should be --

14	DR. COTE:	We tried to do that, but

15   we'll go back and --

16	DR. SAMET:	Yeah, I think I'd just be

17   really -- really careful.

18	DR. COTE:	Often, we change a few words

19   in there, so we don't have that problem, but we'll go

20   back and look at that.

21	DR. SAMET:	And I certainly agree with

22   the approach of saying that, you know, a number of

23   groups have set-out these procedures and protocols for

24   evidence evaluation.  You don't need to reinvent that.

25	Now, so that said, so having set-out the



47	49

1  will be a focus in the future as we try to improve our

2  management -- assessment and management of our air

3  quality.

4	But let's move on now to a discussion by

5  the committee on the Agency Charge Question 1 related

6  to -- that's going to be led by Jon Samet.  Jon, are

7  you ready to go?

8	DR. SAMET:	Sure.

9	So, in reference to Charge Question 1,

10   first, I just made some general comments that I

11   thought, as others have already said, that the second

12   draft is improved over the first.  I'll just tuck-away

13   the comment that it certainly would be easier to review

14   these documents if there was some way to know what

15   changes had actually been made from document to

16   document.  And I don't know -- that's a generic

17   comment, but on spending hours plowing through this and

18   trying to understand what had changed in this

19   electronic age, it's certainly possible to indicate

20   what has been made and to provide some sort of covering

21   memo or something indicating the major changes.  And I

22   think that would be useful I think for the committee's

23   efficiency, and also perhaps to document what has

24   actually happened from version one to version two, and

25   it would be consistent with I think the usual practice

1  framework in Chapter 1 there's I think several issues;

2  is it the right framework and is it on track, and I

3  guess there's a few issues there.  One is the, sort of

4  the use of the so-called Hill Criteria.  My other

5  comment is please don't rename these as decisive

6  factors.  We definitely do not need EPA renaming what

7  has been around for decades.  So, if you could, please,

8  please, take that out.

9	But the question is whether the Hill

10   with its nine categories and criteria are needed versus

11   some of the simplifications of the Hill that have taken

12   place.  And that's probably worth consideration.  I

13   mean, I think if I really pushed you on analogy

14   experimentation, plausibility coherence and said what's

15   the difference, we would all probably flunk that --

16   flunk that test.

17	The -- another related issue,  and I

18   didn't comment on this specifically, but it's in part

19   triggered by, again, seeing Ila's presentation, it's

20   the question of the five categories and what exactly

21   you mean.  And the second one I think is important,

22   because these categories are going to be defined in the

23   sense by the way you use them in applications, sort of

24   by their cases.  If you will, I think if I were to ask

25   you what is a, you know, sufficient to infer with the



EPA CASAC MEETING 05/01/08 CCR#15905-1	14

50	52

1  top level without likely, and the distinction in sort

2  of where you would place your massive certainty.  I

3  think it would be useful.  You know, there are those

4  pictures in the IOM report that are sort of I think a

5  useful guide to talking about classification of

6  evidence when you are less than certain what you have

7  -- that you're not a hundred percent certain let's say

8  of a causal relationship as with let's say active

9  smoking and lung cancer, and then you sort of become

10   less and less certain.  I'm not sure that you right now

11   know where the distinction lies.  It might be useful to

12   talk about that a little bit and even provide some case

13   examples.

14	Then, the other issue is having set out

15   this framework, is it well-applied throughout the

16   document?  And there I think I do see the document as

17   coming up short.  I think some of my other colleagues

18   made the same comments that they do, in fact, set this

19   out.  But then when you come back in these sort of

20   integrated discussions, I don't know whether it's just

21   that it was not carried through, that there were

22   different teams doing the writing of these chapters

23   from those who set-out Chapter 1, but I did not see the

24   kind of I think more rigorous application of these that

25   should be made.  And if you look at the use, for

1  more difficult.

2	I don't know on this question -- I

3  should know -- but whether the St. George's database,

4  which I think is a useful compilation of least of time

5  series studies and others -- whether -- what the group

6  has put in for Nitrogen Oxide.  But that would be one

7  potentially useful way to try and understand what's

8  gotten into the literature.

9	I think that's it, Rogene.

10	DR. HENDERSON:	Thank you.  Is there

11   any need for clarification from NCEA?  You understood

12   that?  Okay.

13	DR. COTE:	I wanted to -- the only

14   reason we kind of were renaming the Hill Criteria

15	DR. HENDERSON:	Can you talk in the

16   mic, because I'm sure the phone

17	DR. COTE:	This is Ila Cote.  I was

18   just looking -- trying to pull up the Hill paper as we

19   were speaking. And I'm not sure he used criteria in the

20   original paper.  But what we were trying to avoid was

21   having criteria for the criteria.  So that we thought

22   was a little awkward.  So I don't know if you want to

23   have discussion about if we could --

24	DR. SAMET:    Well,  I don't know.

25   Decisive factor sounds a little too decisive I think.



51	53

1  example, of these criteria and the Surgeon General's

2  reports where there are very comprehensive sections at

3  the end of evidence reviews on evidence synthesis, I

4  think you would find useful models.  Now, I comment on

5  that later.

6	I do think one issue that's come up

7  already and that I do comment on that's not adequately

8  addressed and I think is potentially problematic for

9  the literature on Nitrogen Dioxide is publication bias.

10   And I think of -- concern here is that -- that there

11   have been many multi-pollutant studies, many time

12   series studies, where, you know, sets of pollutants

13   have been used.  Maybe very often there's been a

14   primary emphasis on particulate matter or ozone.  NO2

15   has been available as an indicator.  It's been put in

16   the models.  And I think you have to be concerned that

17   those studies in which something was found that was

18   either high or statistically significant there was more

19   likely to be mention of a coefficient or something for

20   NO2.  And, in fact, I think just looking at the figure

21   5.1, I think the high estimates, not surprisingly,

22   have very wide confidence bounds -- very much as you

23   would anticipate.  So, I think that issue needs to be

24   certainly mentioned.  The extent to which you want to

25   explore the possibility of publication bias is I think

1  That's all.

2	DR. HENDERSON:	Okay.

3	DR. COTE:    Point taken.

4	DR. HENDERSON:	Okay, Dale.

5	DR. SPEIZER:    Can I comment on this --

6	DR. HENDERSON:	Oh, sure.  Frank.

7	DR. SPEIZER:    -- Hill criteria business?

8	DR. HENDERSON:	Yeah.

9	DR. SPEIZER:    I haven't seen the IOM

10   report and maybe it's done a better job.  But I recall,

11   and I'm using my memory, in reading the Hill criteria

12   in 1965 and subsequently, that there's another

13   paragraph underneath that says these are not rigid

14   criteria. These are suggestions of how to look at data.

15   And, certainly, judgment still needs to be there in

16   doing it.

17	Now, I don't know whether the IOM  has

18   come up with a different set of criteria.

19	DR. SAMET:	So just a comment, actually

20   Hill has.  It's a wonderful paper if you haven't read

21   it.  And he actually concludes at the end, in fact,

22   that you need judgment evidence -- it's always

23   uncertain, and you still have to make decisions and go

24   ahead.  And this same comment is actually made -- the

25   Surgeon General's report in '64 I think has five or six



EPA CASAC MEETING 05/01/08 CCR#15905-1	15

54	56

1  criteria -- makes the same.  These are not, by any

2  means, a check list.  And I think everyone who has

3  written about these has said that.  And I think that's

4  one more reason that you don't want to call them

5  decisive factors.  They're just points to line-up how

6  much you know.  And I think that Frank's point has been

7  made repeatedly and everybody has proposed these kinds

8  of systems.  And, in fact, those calls of philosophers

9  -- think about this -- argued that you must have

10   temporality, but all else is not necessary for

11   causation.

12	DR. HENDERSON:	Okay. We're going to

13   move on to Dale.  Where is Dale?

14	DR. HATTIS:    Yeah.  Here.

15	DR. HENDERSON:	Okay.

16	DR. HATTIS:    So, I focused on the

17   literature search rather than this --

18	SPEAKER:	Dale, could you speak into

19   the mic?

20	DR. HATTIS:    Yes.

21	I've focused on the literature search,

22   and I basically started by reading the document from

23   the reference section.  And what I did note was that

24   there are a large number of 2007 references and a

25   smattering of 2008.  So that really suggests that this

1  the framework was on target.  I felt like the document

2  read very much like a first draft.  And there were a

3  lot of concepts mentioned but they were very brief, and

4  a lot of them were incomplete or weren't well justified

5  as to why you made the modifications.  I learned a

6  little bit more today about, you know, wanting a little

7  bit of nuance that I would have liked to have seen

8  directly in the document.

9	I went back to the annex and the

10   transitions weren't there either.  I thought it was

11   great that the annex had all of the source information

12   and you could look and see what all the documents were

13   that were pulled from, but I didn't see the transition

14   I wanted to -- the conclusions that were stated in

15   Chapter 1.

16	Let's see.  Then, there were a couple

17   places where I wanted to see some changes.  For

18   instance, there's a brief discussion of features of

19   study design and later there's details about the study

20   design, so there's some reorganization that's needed,

21   but then I thought there were some details missing.

22   For instance, in experimental studies, one of the

23   problems with them is they are experiments and you

24   can't do five million experiments to replicate natural

25   conditions, and that piece wasn't mentioned as a



55	57

1  is, in fact, quite a recent literature search and that

2  the results are there.

3	Then I selected what I thought were a

4  dozen or so of the most, you know, facially important

5  sounding titles and went and retrieved the abstracts

6  from Library of Medicine Sources, and then I checked

7  the way in which the authors described their results

8  compared to the way the -- they were summarized in the

9  document.  And in every case I found that there was a

10   good correspondence between what the ISA document had

11   said about the references and what the abstracts seemed

12   to say.  Although I didn't review the articles in

13   depth, it appeared that the literature search had done

14   -- had been thorough and reasonably accurate in

15   reporting what the authors had reported about their

16   findings.

17	DR. HENDERSON:	Thank you, Dale.  When

18   I read your comments, I said I was very glad that you

19   did such a thorough review.  So, thank you.  That's a

20   help to the committee.

21	And, Lianne, you have a comment?

22	DR. SHEPPARD:    Yeah, I focused

23   predominantly on the framework side of the question, a

24   little bit less on the literature review.  And I

25   thought -- I thought it was a great improvement; that

1  limitation of the experimental literature.

2	The measurement error discussion was

3  derived from the Zieger, et al. paper, but it wasn't

4  cited, and it really only applies to the time series

5  design.  It doesn't apply to all observational studies

6  of air pollution.  And literature, at this point, isn't

7  really complete on that area.

8	And, I wasn't really clear on how much

9  the section should be a general review versus a

10   specific application to the air pollution context. For

11   instance, Table 1.6-1, which talks about the Hill

12   criteria and has modifications, I would say it would be

13   better to directly summarize the sources and then

14   comment on how they should be different in this context

15   in the text.  And then I would also agree with Jon that

16   we need to bring out more clearly the issue of

17   publication bias.

18	So those are my main comments.

19	DR. HENDERSON:	Thank you, Lianne.

20   George?

21	DR. THURSTON:    Yes.  My comments really

22   focused on the last part of the question, which had to

23   do with the framework for scientific evaluations of

24   studies and causality determination.  And, you know, I

25   felt that they did a good job looking at the



EPA CASAC MEETING 05/01/08 CCR#15905-1	16

58	60

1  epidemiology, and I don't think really publication bias

2  is the problem that others seem to feel it is.  I think

3  that they've done a good job on the document of

4  summarizing what's there, and the evidence is pretty

5  strong and the epidemiology.  The place where I really

6  felt needed work in the criteria in the document, which

7  I think is a very good document, but the thing -- the

8  gap that I see is between the epidemiology and the

9  toxicology.  And looking at the epidemiology, you see

10   very strong evidence at ambient levels, and then the

11   toxicology finds effects only at much higher levels.

12	And, so, I think that the thing that's

13   missing is something that we do have a fair amount of

14   understanding of and needs to be put into the document

15   is the interaction of particles with gases.  And this

16   is something that I did bring up at our last meeting

17   and I really was disappointed that it was not looked at

18   more.  And I even provided in my comments a paper by

19   Boren from 1964, I believe it was.  It's actually a

20   paper that talks about carbon particles as a carrier

21   for gases and their effects.  And I found this not in

22   Pub Med, because if you look in Pub Med it's not there.

23   There was this young investigator who wrote a summary

24   document of the evidence, Frank Speizer, and he

25   referred to it.  So, when I looked at that I said,

1  them aspects that should be considered.  He asks, what

2  aspects of that association should we especially

3  consider before deciding that the most-likely

4  interpretation of its causation.  And at the end of it

5  I think it's maybe worth looking -- if I can flip to

6  the end of the paper -- at what he concludes at the end

7  -- "All scientific work is incomplete whether it be

8  observational or experimental.  All scientific work is

9  liable to be upset or modified by advancing knowledge.

10   That does not confer upon us the freedom to ignore the

11   knowledge we already have or to postpone the action

12   that it appears to demand at a given time. Who knows,

13   asks Robert Browning, but the world may end tonight,

14   true, but on available evidence most of us make ready

15   to commute on 8:30 the next day."

16	So, I guess what he's saying is use

17   these criteria or aspects -- consider these aspects and

18   then move forward with a decision.

19	DR. HENDERSON:	That's a great quote.

20   I'll have to remember that.  Thank you.  But is there

21   any response from EPA?  I know there's a section -- I

22   would be hard pressed to find exactly where it is --

23   where they did discuss the difference between the tox

24   and the epi, but I don't recall whether they had the PM

25   in there.



59	61

1  well, let me go find that.  And, so, I went through our

2  old library and found this paper.  And just because

3  something has not been published since the beginning of

4  the Internet does not mean it does not bear relevance

5  to this.

6	And, so, I think we really have to look

7  far and wide for -- but I think this is something that,

8  you know, I -- we know.  And it's not an indictment of

9  NO2.  You know, there seems to be this thing, oh, well,

10   you know, there's co-pollutants and these are

11   confounders.  Well, no, actually, they may not be

12   confounders, they may be co-conspirators in this

13   effect.  And that particles are known to be carriers of

14   gases and can make -- be a vector for them reaching

15   deeper in the lung.  And I think that that connection

16   needs to be considered in the document and is not.

17	So -- and then to make that connection

18   between the epidemiology and the toxicology.  And also,

19   I think to identify where, you know, okay, we do have

20   these studies that have been done, but where do we need

21   more information.  So this is an opportunity here for

22   us to see where there are gaps in knowledge.  What we

23   do know and what we don't know.

24	Now, with regard to Hill, I did pull up

25   his paper, and he doesn't call them criteria.  He calls

1	Do you all -- I'm sure you all have it

2  memorized.

3	DR. ROSS:	We certainly did look for

4  the toxicological studies that looked at the

5  interactions between the two, and there aren't many.

6  But we do take your heart -- your concern to heart and

7  we'll try to expand our discussion of NO2 and PM --

8  potential interactions in the upcoming document.

9	DR. THURSTON:    Because I think too often

10   this statistical finding that they're correlated is

11   seen as a negative, you know, that oh, well, then we

12   can dismiss this.  But, actually, it makes a lot of

13   sense when you understand the biological action of

14   these gases and particles.

15	So, I think that the document needs to

16   bring that information to bear and into the

17   understanding of the what seems to be a disparity

18   between the toxicology and the epidemiology, which I

19   see as not -- you know, I'm not surprised that looking

20   at just pure NOx in a toxicology study you don't see

21   effects down to the same level you see it in the real

22   world where there are particles present.  And that

23   makes all perfect sense, and, you know, in my written

24   comments, I talk about the example of ozone, where we

25   had toxicology studies of pure ozone in chambers and



EPA CASAC MEETING 05/01/08 CCR#15905-1	17

62	64

1  they said, well, you can't see lung function decrements

2  below 120.  So when we came out, and Frank I'm sure

3  remembers this, we did our studies of children in the

4  '80's, and we showed lung function decrements down

5  well-below a hundred, and they said, well, you must

6  have a confounder or something.  And then when they did

7  the -- redid the experiments to be more like the real

8  world with multi-hour exposures and people exercising,

9  they were able to replicate the epidemiology.

10	So, I think there's a real example there

11   of how by thinking about the disparities between the

12   toxicology and epidemiology you can bring it together

13   and understand better what's going on.

14	DR. HENDERSON:	Thank you, very much,

15   George.

16	I would -- before I open it up for

17   everybody else who may have some comments on Charge

18   Question 1, I want to bring up something Ellis Cowling

19   asked me to say, because Ellis is a charter member of

20   CASAC, and he was just feeling very, very guilty that

21   for the first time in all of these years on service at

22   CASAC, he didn't get his written comments in.  So, he

23   asked me to verbally mention one thing.

24	He liked the way you put the policy

25   relevant questions in in the introduction and in the

1  atmospheric chemistry measurement and the like.  I

2  think it is effective.  I thought the first draft was

3  effective, and we provided some points that could be

4  improved.  And I think you've addressed most of those.

5  I think there's still a few points that, at least on my

6  end, could -- a next draft, or actually, I guess it

7  would no longer be a draft -- a final document might

8  improve -- that I'd still like to see the -- at least a

9  brief list of sources somewhere in the actual document.

10   I just think that would be helpful.  It would highlight

11   what are the major sources such that someone who is

12   reading this document without reading the annexes would

13   have a real feel for where things are and, you know,

14   that it's not just mobile, on-road traffic, which a lot

15   of people think.  And, in fact, when you start looking

16   at the next document, the REA, you see that one of the

17   issues that comes up is things like the airports, and

18   you can also start thinking about railroads, et cetera;

19   in terms of transportation corridors.  And I think that

20   would be good to show that up front.

21	Likewise, and this also comes up from

22   now reading the REA, and I didn't put -- I don't think

23   I put this in my initial review of the -- the first

24   draft of the ISA, but you start seeing that in the REA

25   that models are playing a bigger and bigger role in



63	65

1  Chapter 5, and I also like that.  And he suggested that

2  at some point in the process that they be answered

3  directly.  In other words, you have a list of the

4  question and then you list directly the answers.  And

5  this is not from Ellis, this is from me, I was thinking

6  that might go in the policy assessment document or

7  something, you know, at a later stage.  But his comment

8  was that those were great questions and we need a

9  specific spot in part of the processes that gives the

10   answers.

11	And now I will open up the discussion.

12   Does anyone else have comments on Charge Question 1?

13	Okay.   Jon, you know you're the one

14   that's going to pull all this together and develop

15   eventually a short paragraph or two for the letter to

16   the Administrator?

17	That's what the lead discussant is

18   doing.

19	Okay, let's move on then to Charge

20   Question 2.  And, Ted Russell is the lead discussant.

21	DR RUSSELL:    Thank you.  And, I do like

22   to give my congratulations to what I think is a

23   substantially improved document.

24	Looking at Charge Question 2, which

25   deals primarily with how the document covers

1  terms of what they're addressing their task.  And when

2  I started thinking about that, I really think that

3  those models that they're mentioned in the annexes, but

4  should come up in the document itself, albeit short,

5  but still a good synopsis of what models are available

6  and what are their likely limitations, uncertainties,

7  and how they might be applied in a scientific realm in

8  the risk exposure -- risk and exposure aspects of the

9  next document.

10	I think this is particularly true when,

11   again, this is sort of looking forward to the next

12   document, when you see how they essentially figured out

13   what concentrations might be near the roadway.  And

14   they have -- they come up with a formula that, you

15   know,  it would have been nice to have that sort of

16   issue more vetted in the ISA.  What sort of approaches

17   one might actually take to come up with what are the

18   peak on-road or near-road concentrations based upon

19   ambient monitors and a greater evaluation of such,

20   which I don't think that -- I think it's somewhat not

21   fully vetted in the REA document.

22	Another comment I'd still make is that

23   it doesn't -- this draft, the issue of the monitor

24   interferences, it's, I think, better characterized, but

25   I'm still worried that one gets the wrong impression



EPA CASAC MEETING 05/01/08 CCR#15905-1	18

66	68

1  how important those interferences might be, in that if

2  you start looking at the endpoints of concern here,

3  that being both the average, but more likely the peak

4  concentrations of NO2, it's unlike -- you know, one

5  gets the feeling, okay, that the interference could be

6  as much as 50 percent, but that 50 percent first is

7  generally taken from -- well, in this case you referred

8  to a Mexican study, and I'll tell you Mexico City is

9  very different than the U.S., and then a Swiss study,

10   and you correctly state that that's typically -- you

11   get the maximum interferences during the summer, in the

12   afternoon, when you have the high levels of the

13   interference, as in nitric acid and the PANs, but

14   that's also not when you typically have the higher

15   levels of NO2 that are going to weigh-in on either the

16   maximum concentrations in an area which are really

17   going to be not during the middle of the afternoon,

18   they're going to be more in the morning or the later

19   afternoon typically.  And then they also probably would

20   not be during the middle of the summer.  And so I think

21   you get this wrong impression there.  And, also, even

22   on the averages is probably going to be somewhat

23   decreased, so I would recommend that that be considered

24   further -- is that the interferences in terms of how

25   they might be -- how they might come up in terms of

1  worth.

2	And also, like Ted, we had requested in

3  the first round of reviews that this document

4  incorporate some additional quantitative information on

5  sources and, you know, oddly, there was a nice

6  paragraph or two in the risk assessment document that

7  was more quantitative than what was in the ISA, you

8  know, not counting the annex -- the table in the annex.

9  So, something along those lines may be a little -- with

10   a little more detail would be very helpful.

11	And another sort of carry-over from the

12   previous review was that the summaries -- I thought

13   Chapter 3 incorporated a lot of nice summary statements

14   at the ends of sections and sort of made your

15   discussions, and that would have been helpful in this

16   Chapter 2, as well.

17	Just one item that wasn't in my written

18   comments, but became more clear as I was reading

19   through the Risk Assessment, in light of the emphasis

20   on on-road concentrations.  I think that Section 2.5.4

21   could be -- or should be -- sort of additionally

22   supported by some analysis of the distribution of

23   on-road concentrations, or near-road concentrations.  I

24   guess I was somewhat uncomfortable with the model used

25   in the Risk Assessment documents.  So, you know, going



67	69

1  assessing a standard -- probably just not as great.

2	Other than that, I've got some other

3  individual -- oh, one other thing I'd note is again in

4  the REA, it notes that very few of the monitors that

5  are used are at 15 meters, though in the ISA there's

6  still the issue of this monitor -- monitors being

7  between 15 meters and more like four meters, is an

8  issue.  And, so I think there's a couple of things that

9  come up here that would suggest, you know, maybe

10   belatedly, that if you start looking at the REA how you

11   might -- how that might reflect on this draft of the

12   ISA, and to make them more consistent.

13	Thank you.

14	DR. HENDERSON:	Thank you, Ted.  Donna?

15	DR. KENSKI:	Yeah.  Thanks.

16	Like Ted, I thought this was a much

17   improved document, and I guess I'd like to thank the

18   EPA folks for incorporating so many of our comments

19   from the first round.  So, just a handful of comments

20   for me, too.

21	I thought the section 2.4.5 on

22   concentrations of the NOz species was still pretty

23   limited and could benefit from a little additional

24   information on those other related species.  So, I

25   suggested a couple of references there for what it's

1  back to this section, you know, I understand that the

2  data are scarce, but what are there I think could be

3  incorporated in this document to help -- I don't know

4  -- help me believe the models that are used in the Risk

5  Assessment portions.  So I would like to see data

6  specifically on the distribution of those on-road and

7  near-road concentrations to the extent that it exists.

8	DR. HENDERSON:	Thank you, Donna.

9	Is Tim Larson on the phone?

10	DR. LARSON:    Yes.

11	DR. HENDERSON:	Okay, we can hear you

12   fine.

13	DR. LARSON:    Great.  Thank you.

14	I  second the comments so far that this

15   is a much-improved draft over the first one.  I

16   sympathize with EPA's attempt here to do this risk

17   assessment given the sort of sparsity of the monitoring

18   network, and not only is it fairly sparse, but it's

19   dwindling, which is a challenge.  Add to that the fact

20   that the focus or the interest seems to be moving more

21   from a long-term average to a short-term exposure and

22   the network and its spatial intensity becomes even more

23   of a challenge because, as we know, spatial variation

24   is not only greater over these shorter time scales.

25	I really liked Ted's suggestion about



EPA CASAC MEETING 05/01/08 CCR#15905-1	19

70	72

1  adding a section to the ISA to summarize the models

2  that are used in the risk assessment, especially

3  extrapolating from the long-term average to monitor

4  values to the shorter term peak exposures.  Because

5  there's a number of issues that are difficult to

6  untangle not only for the average reader but for the,

7  for those who are more familiar with this area.  And,

8  so, I think that's a great idea because it lays out

9  better what the options are.

10	To that end I made a comment in my -- in

11   my written comments I provided some more recent

12   references on some of the integral models that are used

13   in confined urban areas because there was a statement,

14   albeit a passing one, in the ISA that said that the

15   only way you can do this is with some complicated

16   computational fluid dynamics model, and it was sort of

17   intractable.  But if you look at that literature, it --

18   there's a number of papers -- I think I had four in

19   there from the last year -- three or four of them in

20   2008 even that are pretty encouraging, and if this has

21   its own literature, but I think it would be certainly

22   something that would be worth discussing in the ISA in

23   a more summary form as Ted suggested.

24	One idea -- I like the idea in this

25   memorandum that was sent around of looking at the

1  that aren't, and you made the classification, which

2  seemed reasonable, a hundred meters.  And I'm wondering

3  if there's any information -- if that information sort

4  of carried through the interpretation of any of those

5  epi studies?

6	DR. ROSS:    The memorandum you're

7  referring to is about the Risk and Exposure Assessment;

8  is that right?

9	DR. LARSON:    Well, it is in its detail,

10   but I mean there was also an analysis made of the

11   monitors and which ones, you know, information is

12   available on which ones are right next to roads.

13	DR. HENDERSON:	Maybe we can wait 'til

14   we get to that.

15	DR. ROSS:    So what you're suggesting

16   though is interpreting the epi studies?

17	DR. LARSON:    Right.

18	DR. ROSS:    Looking at the monitors used

19   in those

20	DR. LARSON:    I mean, I don't know to

21   what extent the epi studies that are summarized there

22   had monitors that were right next to roads versus were

23   away from roads, and I'm just wondering since you had

24   that information in another analysis, in another part

25   of the document, whether or not any of that was



71	73

1  monitors near the roads versus those that are further

2  from the roads as sort of a classification scheme.  And

3  I'm wondering whether or not any of that information

4  was translated into the epidemiological

5  interpretations, because it seems like for either the

6  long or short term numbers there's sort of a

7  qualitative difference in those monitors.  And I don't

8  really have a sense of which of the epi  studies may or

9  may not have used some of those monitors that are

10   seemingly different.  I don't know if that's possible,

11   but it might be something that's illuminating.

12	DR. HENDERSON:	Would you like for us

13   to -- for EPA to respond to that now?  I mean, you were

14   asking EPA a question.

15	DR. LARSON:    Yes.

16	DR. HENDERSON:	I mean, relating the

17   roadway values in the memorandum that went around to

18   epi?

19	DR. ROSS:    Can you restate the question?

20   Because I'm struggling to think of where this fits.

21	DR. LARSON:    Yeah, there's -- at least

22   the memorandum that was sent around, and there was some

23   passing reference to it in the document for the EPA

24   monitors there were near-roads, there's a number of

25   them that are close to major roads and there's others

1  included in your summary or interpretation of that

2  data?

3	DR. THOMAS LONG:    Dr. Larson, this is

4  Tom Long from NCEA.

5	DR. LARSON:    It didn't seem to be in

6  there, so maybe it was a suggestion that   it might be

7  useful. I don't know.

8	DR. ROSS:    First, let me recognize that

9  there are very few epi studies that were done in the

10   United States, and we wouldn't have that kind of

11   information for studies done in Europe or even actually

12   Canada at our fingertips.  But Tom can address

13	DR. THOMAS LONG:    Yeah, I think the --

14   this is a very interesting question of how the monitors

15   used in the epi studies influenced the interpretation.

16   And it probably can only be addressed -- even with this

17   detailed analysis that was done in the memorandum -- it

18   probably could only be looked at in a qualitative

19   manner.

20	DR. LARSON:    Yeah.

21	DR. THOMAS LONG:    For the reason that

22   Mary just mentioned, and also, the epidemiology studies

23   are done over quite a long time period, and monitors

24   come in and out of the network and those sorts of

25   things.  But, I think it would be interesting for us to



EPA CASAC MEETING 05/01/08 CCR#15905-1	20

74	76

1  at least look at it in the qualitative manner.

2	DR. LARSON:    I think that's about all

3  you can do.  I agree that would be interesting.

4	So, anyway, I guess that's all I have to

5  say at the moment.  I thought it was an improved

6  document for sure.  And I really like the idea of a

7  fairly short document that's to the point and there's

8  been a lot of thought gone into it.  Thank you.

9	DR. HENDERSON:	Thank you, Tim.

10	DR. MENG:    This is -- may I say

11   something about this?

12	DR. HENDERSON:	Okay, you, and then

13   Frank.  Okay. Go ahead.

14	DR. MENG:    Okay. This is Qingyu Meng

15   from NCEA.  I just want to mention that you've seen a

16   ambient monitor as a surrogate for population exposure

17   in epi-studies.  Actually, qualitatively we have

18   addressed that issue in the annex with the exposure air

19   part.  So you've seen the ambient monitor as a

20   surrogate for population exposure related to the

21   spatial variation of NO2 and also the personal

22   activities.  So how does that affect the epi results --

23   we mentioned something in the annex.

24	DR. HENDERSON:	Thank you.  And Frank?

25	DR. SPEIZER:    Yes, I guess I was kind of

1  where they are in terms of where the measurements are

2  made.  Because, in fact, that's the criteria they use,

3  is distance from the road.

4	DR. ROSS:    Well, there actually are

5  different sets of studies.  And some of the

6  epidemiologic studies are associating effects with NO2

7  from the monitors and they don't necessarily say where

8  the monitors are.  And we could find out where they are

9  in the United States, but not at these others.

10	DR. HENDERSON:	Okay, Christian, you

11   have comments?

12	DR. SEIGNEUR:    Yes, we'll also find that

13   the draft seems to be an improvement of the previous

14   draft.  The comments that were made in the previous

15   meeting I think went very nicely otherwise.  So I only

16   have two comments.

17	The first one actually Ted touched upon

18   that a little bit, and that's the issue of the

19   uncertainty associated with the difference in the NO2

20   measurement technique.  In Chapter 1, there is

21   description of uncertainties and how they affect of

22   course the results of the health effects studies, and

23   then we move to section two where there is a very nice

24   complete description of all the uncertainties

25   associated with the NO2 measurement technique.



75	77

1  shocked by Mary saying that because it's in Europe and

2  Canada, you don't have the data.  I'm surprised.  I'm

3  not sure what you're saying.

4	DR. ROSS:    We don't know where the

5  monitors are in those studies.  We can identify them in

6  the United States and we can tell you which ones are

7  close to roads.

8	DR. SPEIZER:    Okay. But most of the

9  epi-studies really indicate the effects that are

10   related to proximity to roads.  So, at least in terms

11   of what the epi-data shows, it really indicates that

12   the closer you are to the roads, the greater the

13   effects that are seen.  So that tells you one thing.

14   One of the criticisms I was going to come back to later

15   with regard to the 18 page memo was it would really be

16   nice to know what is the population concentrations that

17   are within that 100 meters that you use versus, that

18   you're throwing out as terms of being too close to the

19   roads, are there populations there.  And if there is

20   nobody living there that's fine.  I mean, that's

21   perfect, and that's the right thing to do, but we

22   really ought to have some indication of what the

23   population load is in those spaces.

24	But in terms of the -- it seems to me

25   the studies that have indicated effects have indicated

1	As Ted mentioned, some of those

2  circumstances under which you have large uncertainties

3  may not be relevant -- actually, it was health effect

4  studies -- where typically you have high NO2

5  concentrations for your results and typically in

6  wintertime -- the appendix shows nicely that you tend

7  to have higher NO2 concentrations in winter compared to

8  summer.  I mean, it's an oversimplification but

9  perfect.

10	And of course as you move away from the

11   source, you have atmospheric dispersion and then you

12   have, you know, conversion of NO2 to nitric acid and

13   organic nitrates.  So the implication is that when you

14   have high interference, most likely, you don't have

15   high NO2 concentrations.  But that's not being said

16   explicitly at the end of section 2.

17	And then if the reader move to section

18   3, you clearly decided that you may have large

19   uncertainties due to the NO2 measurement technique,

20   which I don't think would be the case.  So my

21   recommendation would be that at the end of section 2,

22   you know, you bring closure to this issue of

23   interference and whether or not it does have an impact

24   on epi studies, which in my opinion, it probably does

25   not have a major impact.



EPA CASAC MEETING 05/01/08 CCR#15905-1	21

78

1	The second comment is actually related

2  to an appendix.  In the previous meeting, we discussed

3  air mode,  which was used for the exposure and risk

4  assessment.  And there was a discussion of  CMAC and I

5  brought up the point that if we're going to use air

6  mode we should describe air mode.  And now the appendix

7  has a nice description of air mode including, you know,

8  the pros and cons of the model.  But when I read that

9  section I was a little bit perturbed because, first of

10   all, there are some limitations of air mode which are

11   being brought up and some I don't think are very

12   pertinent to what we are doing here.  Air mode is a

13   steady-state model, which means that you assume the

14   wind direction to be constant for one hour, which I

15   don't think to be that major limitations because we are

16   looking at NO2 concentration in proximate sources in

17   most cases.

18	And then that paragraph in the appendix

19   ended up bringing up SCAMP F which is another

20   dispersion model implying that SCAMP F actually could

21   be a better model to be used for that type of study.

22   And I'm not sure I would agree with that because there

23   are problems with SCAMP F .  SCAMP F doesn't do the NO2

24   calculation correctly for instance.  So, either that

25   appendix will have a complete review of all dispersion

80

1  when you're trying to read it.

2	So, I'm going to -- I spend a little bit

3  more time on the dosimetry part of things since that's

4  my background, and the other sections in the chapter.

5  And first of all, there wasn't -- there really wasn't

6  much work to do there because there hasn't been much

7  progress since the last, last ozone, OS and NOx review.

8  So the appendix seems to include all the relevant

9  references.  There's considerable -- most of the new

10   work has to do more with what I would consider the

11   biochemical aspects of dosimetry rather than the

12   extrapolation modeling aspects that try to determine

13   internal dose from exposure, and try to extrapolate

14   that across different exposure conditions and different

15   species of animals.  So there hasn't been much progress

16   there.

17	Now, in this particular review that

18   posed no obvious problem because the actual health

19   effect that was chosen for risk assessment turns out to

20   be taken from clinical studies, or it'll say it's a

21   health effect that there's been a lot of clinical work

22   on in humans, so that it was possible just to use the

23   exposure conditions of the experiments directly rather

24   than extrapolating.  So, I'm not sure this is good or

25   bad.  I have this feeling in the back of my mind that



79	81

1  models; including air mode, SCAMP F and the other ones

2  like SCAQMD, which I think would be overkill or we

3  simply, I think my recommendation would be to eliminate

4  the comments based -- eliminate the comments on SCAMP

5  F.  Rewrite a little bit air mode in a more positive

6  light with respect to the application we have here, and

7  then I think then we would have a nice appendix

8  presented -- presenting the model that's being used in

9  the assessment.

10	DR. HENDERSON:	Thank you, Christian.

11	Jim?

12	DR. ULTMAN:    I agree with everybody else

13   in the sense that this chapter has vastly improved

14   since the first draft.  It was much more inclusive and

15   much more readable, I thought.

16	I have one very minor point, but it's a

17   nagging point.  It has to do with document development.

18   When you print out a color coded contour plot in black

19   and white, it just messes up all the data.  So, for

20   example, the background -- the plotted background

21   levels, I mean, you can't tell what's low and what's

22   high.  So there must be a way of overcoming this

23   problem when you print the documents out.  This was the

24   same problem that appeared in the ozone document, as

25   well.  It's a minor point, but it is -- it's disturbing

1  there's a kind of an implicit bias when choosing health

2  effect as a basis for risk assessment in choosing

3  something where human data exists so that you can --

4  you don't have to worry about extrapolation issues.

5  And as I said that's the good thing because it makes

6  the risk assessment very clean.  It's a bad thing

7  because you may miss some important issues because --

8  that have been uncovered in animal studies, because you

9  haven't gone through the trouble of trying to do the

10   proper extrapolation modeling.

11	So, the only -- the only real suggestion

12   that I have here is to possibly extend some of the

13   calculations that were done already in this document,

14   in terms of perhaps the cumulative inhaled dose that

15   was present in different studies, both animal and

16   human, and maybe display them in some type of a tabular

17   form.  In other words, there's various end points that

18   were measured in both human and animal studies.  And,

19   you know, the hours of exposure, the concentration

20   levels, the patterns of exposures, are roughly known.

21	So that you could get at least an

22   overall inhaled dose in those studies. And you could,

23   dose could be listed for animal and human studies as

24   well.  As I say, I know you've done some of that, I

25   think, for the human studies already, but not for the



EPA CASAC MEETING 05/01/08 CCR#15905-1	22

82	84

1  animal studies.

2	It would be nice to go one more step,

3  and that would be to use existing literature on

4  dosimetry to see if some of the animal studies could be

5  at least roughly extrapolated to what that would mean

6  in terms of human concentration.  And I realize in some

7  cases it's kind of irrelevant because some of the

8  effects in the animals were seen at very high exposure

9  concentrations and even if extrapolated, it wouldn't

10   make any difference.

11	But some of the animal studies, the ones

12   having to do with infectivity and things of that

13   nature, immune system compromise, some of those were

14   seen at very low exposure levels.  So it might pay to

15   do a, if it can be done on a basis, as I say, of

16   published studies, to try to do a rough extrapolation

17   to what it might mean in a human.  So as I'm saying, my

18   suggestion to the table that it at minimum would try to

19   estimate for the different types of endpoints and the

20   different species, you know, rats and humans, for

21   example, what the cumulative doses were for comparison

22   in the various studies. And I'm not suggesting you have

23   to do this in every study but rather maybe divide it

24   according to the particular endpoints we're talking

25   about.  And then it maybe even push the envelope a

1	So why am I bringing this up under this

2  charge question?  Well, because I think the issue is so

3  important that it shouldn't be held off until the risk

4  assessment

5  document. I think more has to be said, and you've

6  already mentioned this, more has to be said about how

7  on road data looks, the on road concentration datas

8  look relative to off road data and then even going

9  further.  Just as there are equations in this chapter

10   which, detailed equations on how exposure is

11   calculated, I think it wouldn't hurt to move forward

12   some of the quantitative equations on the on road

13   extrapolations.  To move them forward from the risk

14   assessment to this document and explain them in more

15   detail so that when we do get to calculations such as

16   are in this supplemental memo, that we can really

17   understand what the meaning of those particular

18   calculations are.

19	So those are the two major, major

20   comments I had.

21	DR. HENDERSON:    Thank you, Jim.  Now I

22   would open it up for anybody who has further comments

23   on Charge Question 2.  Lianne, is your hand up?

24	Okay.

25	DR. SHEPPARD:    I wanted to also talk



83	85

1  little further and see if the animal overall doses

2  could be extrapolated down to or up to what it would be

3  in human.

4	So that was one of my comments.  The

5  other comment has to do with the, I think, I'm

6  following on a comment that was already made that had

7  to do with the on road concentrations of NOx and NO2 in

8  particular.  This turns out to be, as we all saw, a

9  very important component of the risk assessment

10   document and I have to admit, I was really, I thought

11   the explanation in the risk assessment document of the

12   extrapolation method, by which you go from, you know,

13   ambient monitored data to roadside was not clearly

14   written.   I mean, I couldn't really follow it very

15   clearly.  But apparently there is an assumption in

16   there, for example, about not having the influence of

17   the roadside concentrations affect the calculations.  I

18   couldn't even, I couldn't follow that, and maybe it's

19   just my own thickheadedness, but I consider myself at

20   least an average reader.   So I thought the explanation

21   in the risk assessment document of the extrapolation

22   method was weak.  Now maybe if I had gone to the annex

23   and read it in a lot more detail, I would have caught

24   on more, but I did go to the annex and it really didn't

25   help me.

1  about this whole problem of monitor siting and the

2  data, interpretation of the data.  It, as you

3  mentioned, Jim, it's really important in the exposure

4  risk assessment document and as Tim mentioned, it

5  potentially has great implications for the

6  interpretation of epi studies.

7	Unlike particulate matter, oxides of

8  nitrogrn vary very highly as a function of distance

9  from road, and so some of these weak correlations that

10   we're seeing are presumably driven completely by that,

11   particularly when you're looking at hourly data,

12   because there can be huge differences due to wind

13   direction and the road and so on and so forth, that can

14   be driving that.  And that could be affecting

15   interpretations and abilities to actually make imprints

16   about health effects.  Very, very dramatically, I

17   think.  And it could also be very dramatically

18   affecting the estimates of exceedances in the exposure

19   and risk assessment document.  So it's difficult, but

20   we've got to do a better job with clarifying the

21   importance of monitoring siting and incorporating that

22   into the entire assessment.

23	So that's my main comment.  My other

24   comment is related in the sense that I thought, first

25   of all, I wanted to say overall, there were many good



EPA CASAC MEETING 05/01/08 CCR#15905-1	23

86	88

1  improvements, the entire document and so any critical

2  comments need to be taken in light of the fact that

3  you've done a great job integrating a lot of things.

4	The discussion of the correlations I

5  thought was a lot better, but I felt like I still

6  wasn't satisfied.  There's a couple of things.  One is,

7  I'd love to see the formulas, I still didn't find the

8  formulas, and the words were better so I felt like I

9  had a fairly good idea of what was being done, but I'd

10   still like to see the formulas, at least in the annex.

11	And I've, maybe I've missed them in

12   there, but, and cross-referencing so that we can find

13   them easily, too.  I mean, correlations are

14   standardized quantities, so I have a difficulty with a

15   comparing them because they're not only about the

16   co-variance of two things, but they're also about the

17   relative variability, and so when you, for instance,

18   look, compare correlations from annual data compared to

19   season restricted data, you can get much smaller

20   correlations in the season restricted just because

21   there's less variability within one season, there is

22   across a year.  But I don't think that's what we care

23   about  when we're trying to compare these correlations.

24   I don't think that's what the purpose of  our inference

25   is.

1	DR. THURSTON:    Yes, I wanted to follow

2  up on what Lianne was talking about, 'cause I too felt

3  this whole issue of personal versus population

4  exposures needs to be clarified. I think, you know, the

5  public comments we had at the beginning clearly showed

6  that what's written in the document is not being fully

7  understood.  I mean, they do address this issue that

8  what's relevant to the population studies is not the

9  correlation with individual NO2 levels, but with the

10   average of the individual levels. And, you know, if

11   you're going to do a personal level individual cohort

12   study, well, the, yes, the personal exposure's

13   important to consider. But if you're doing a population

14   study, what you're interested in is the average of that

15   population and what is found, what has been found is

16   that when you average over a cohort, where it sort of

17   eliminates the noise of each individual person's

18   personal variation in their like, indoor exposures and

19   things, that that average correlates with the central

20   site monitor much better and justifies the use of

21   central site monitors for epidemiology. So I think that

22   needs to be, you know, based on the conversations that

23   I've heard, you know, and the comments, that that needs

24   to be even said more clearly, I think.

25	One of the things is, that Lianne, I



87	89

1	And so these raw comparisons of a whole

2  host of correlations, I just feel like we're just not

3  getting at what we care about, and it's a difficult

4  problem, so, and I think, as I said, I think you've

5  made progress and you're doing a better job than the

6  first draft, but I still think there's a long way to

7  go.

8	DR. HENDERSON:    Thank you, Lianne, and

9  Ron is in my blind spot, so I'll call...

10	DR. WYZGA:    Thank you, Rogene. I think

11   one of the things we're going to be faced with later is

12   basically interpreting and then integrating health

13   information from many different studies and types of

14   studies.  And I think it would be useful to have an

15   explicit section in here that talks about some of the

16   difference in the characteristics of air quality from

17   the outdoor studies and the indoor studies, and you may

18   even want to add some of the clinical studies, in terms

19   of what's the relative composition of the different NOx

20   compounds, and also what are the co-pollutants?

21	There are differences between them, but

22   I think having some explicit articulation of that would

23   help us in interpreting the health studies later on.

24	DR. HENDERSON:    Thank you, Ron, was

25   there somebody else? George?

1  think what Lianne was getting at was, you know, there

2  was sort of mixing up the cases of where there's

3  correlations between individuals and the central site,

4  and the average of the individuals.  It is noted in the

5  table, but I almost think they should be just separated

6  completely and saying, these are the correlations that

7  might be relevant to an individual level study, and

8  therefore, you would need to individually monitor.

9  Which is what epidemiologists do.  So they recognize

10   that if you're going to look at an individual's

11   symptoms, let's say, you need to have that individual's

12   exposures, but if you're going to look at the

13   population of an entire city, then you don't need every

14   person's exposure, what you need is the population

15   weighted, you know, the population average.

16	And so I think that that disparity and

17   distinction in the correlations data, because the way

18   it is right now, it's easy to be confused by it and

19   say, well, look, some of these are quite low and some

20   of them are better, but so we need to make that, I

21   think even more, it is in the document, but clearly,

22   maybe by having a separate table for the two and

23   discussing the implications for the epidemiology might

24   make that clearer to people reading the document.

25	The other, the other thing is the



EPA CASAC MEETING 05/01/08 CCR#15905-1	24

90	92

1  question of the central site versus the, even when

2  you're looking at that average exposure of the

3  population and then looking at the central site or even

4  just the correlation of the central sites, you're going

5  to find that if the monitor is at a real hot spot, that

6  it's not going to correlate well with population-based

7  monitors.  And so if you look at that correlation then

8  you're gonna, and you do, and then you say well, the

9  correlation could be quite low, but epidemiologists are

10   aware of this, and when we do the epidemiology,

11   generally it's practice, if you're going to do a

12   population weighted exposure, you know, if you're

13   looking for a population exposure, not to use the hot

14   spot.  You use the community-based monitors.

15	And that should be in the papers.  I

16   mean, I'll give you an example.  If you're going to

17   look at hospital admissions, there are scheduled and

18   unscheduled admissions.  And someone could say, my god,

19   why are they doing hospital admissions, you could get

20   really disparate results if you include scheduled

21   admissions. 'cause they're going to happen irrespective

22   of air pollution.  Well, we recognize that when we do

23   the studies and we don't include these scheduled, the

24   people who are scheduled to come in on Tuesday for an

25   operation a month ahead.  Those aren't included in the

1	So I think that some of these

2  correlations could be presented in a way, in a context

3  more thinking ahead to the epidemiology and saying what

4  is general practice in the epidemiology and which

5  correlations relate to which types of epidemiology?

6  And I think that's sort of saying what has been -- you

7  know, what Lianne was getting at, I think, right?  And

8  certainly Tim was getting at that with the

9  epidemiology, that the central site hot spot monitor

10   may not be represented but it's not what is used, so it

11   looks like a bigger problem in the document than it is,

12   I guess.

13	DR. HENDERSON:    Okay, Lianne can respond

14   and then Ed has, is waiting in line here.

15	DR. SHEPPARD:    Just to follow up, it

16   matters quite a bit what the study design is, George.

17	DR. THURSTON:    Yeah, right, good point.

18	DR. SHEPPARD:    And so it can't, I mean

19   the time series studies you want may be a population

20   based monitor by and large, but a lot of the cohort

21   studies of course the focus is on getting that

22   variability that has to do with near roads,

23   capturing that.

24	DR. THURSTON:	Right, so I think that

25   that distinction of types of epidemiology for types of



91	93

1  count.  That would be a problem, but we don't do it.

2	Similarly I think that using a central site

3  hot spot monitor to represent the entire population

4  would be a problem, and generally, you know, as  to my

5  knowledge, nobody does that. So it's not the problem it

6  appears to be from the Table of Correlations.  In other

7  words, you might try what are the correlations of, if

8  you eliminate the traffic hot spot monitors, what's the

9  correlation of the community based monitors?   Which is

10   what someone's going to do with, if they're doing an

11   epidemiol -- now if they're doing a roadway study, sort

12   of what Dr. Speizer's talking about, yes, they would

13   include those and then they would look at the

14   population living very close, that would be a different

15   kind of study.

16	But for the population city-wide, I, my

17   understanding, and it's general practice to not use the

18   hot spot monitors, certainly not just that one. You,

19   what people do is average over, so that's another

20   distinction. If you're going to look at the correlation

21   of one monitor with the personal levels, that's not

22   what people do in the epidemiology, they average, you

23   know, all, especially the community-based monitors,

24   they average them on a day and they use that average to

25   get an idea.

1  data, although it really, you know, I think needs to be

2  made up here in the exposure, because people, you know,

3  in a way to relate it, I mean it's discussed, it is

4  discussed and I think maybe that could be clarified,

5  and make this distinction.

6	DR. HENDERSON:    It was discussed in

7  the --

8	DR. THURSTON:    It is.

9	DR. HENDERSON:	You want just to be,

10   have it clarified --

11	DR. THURSTON:	I think so, yeah.

12	DR. HENDERSON:    It was discussed and you

13   want to just have it clarified and you've been waiting

14   patiently.

15	DR. AVOL:   I just had a few brief

16   comments, some a little bit structural, and then to

17   amplify some things that previous speakers have said.

18   First I agree that the document is much improved and

19   that the terse sort of focused nature of it is welcomed

20   and appreciated, but I think in doing that you also

21   need to sort of present the key important issues and

22   then to provide some measured judgment about what those

23   issues are and I think sometimes that's gotten a little

24   bit away and two areas where I think it's worth

25   nothing.



EPA CASAC MEETING 05/01/08 CCR#15905-1	25

94	96

1	One is in the multi-pollutant

2  considerations.  We've talked a lot in previous

3  meetings and there have been comments and you'll

4  continue to hear comments about the issue of

5  multi-pollutants and how that factors in here and so I

6  think you sort of...you need to sort of talk about that

7  because as George pointed out, there's biological

8  plausibility, both in the chemical transformations and

9  in the physical world as well as in the biological, the

10   receptor effects and so I think it's important to bring

11   that out somewhere in the document, as appropriate in

12   the respective sections.

13	The other thing that I think is an issue

14   of integration, the ISA should take the data and make

15   some measured judgment about things, it seems to be,

16   and particularly in Chapter Two, one of the problems I

17   had was that you talk a lot about the measurements as

18   Ted and Christian both said, you describe the

19   measurement, you describe the potential interferences,

20   et cetera, but in that chapter at the end of it, you

21   don't come to some conclusion about where you are with

22   it.  There's a little bit of that later in Chapter Five

23   at the summary of the whole document but it would be

24   very useful I think in terms of once you get rolling

25   and going with the document, chapter by chapter, and

1  and what we do see in that figure is that at some point

2  there appears to be not really, I hesitate to call it a

3  threshold, but a level at which people do start having

4  an inflammatory response that is statistically

5  significant.

6	Unfortunately the dose response is

7  somewhat unclear above that point as to whether the

8  magnitude of the actual inflammatory response would

9  increase or not.  However, if we moved over to figure

10   3.12 where we looked at increases in airways

11   hyper-responsiveness.  Before I go back or before I

12   move on to that, for anyone that's on the phone that

13   cannot receive a handout, the figure that I just

14   referred to on inflammatory responses, that's figure

15   3.11.  That may be found in the first draft of the

16   document as Annex figure 51.  But moving on to the

17   airways hyper-responsiveness, what we've done in that

18   figure is we've tried to again look at the dose that

19   people received and whether there was a statistically

20   significant effect indicated by the plus sign or not

21   statistically significant effect indicated by the

22   negative sign and it's not really clear that we have a

23   nice dose response relationship at least for airways

24   hyper-responsiveness.

25	I mean there are some reasons for that.



95	97

1  building up the case, as it were, that when you

2  describe this, when you present this, you make some

3  judgment about what do you think about this and then go

4  on.

5	DR. HENDERSON:    Okay, are there other

6  comments?  I want to ask EPA, have our...are they

7  clear, do you need any clarification of these comments

8  for any reason?

9	DR. JAMES BROWN:    I would like to just

10   comment...this is James Brown and I would like to just

11   comment on the dosimetry issues that Dr. Ultman brought

12   up.  I have a few comments there, the first one is that

13   I would like you to know that we are actually working

14   on looking at all the end points in both the animal

15   studies and the clinical studies and trying to see if

16   we can identify a clear dose response relationship.

17	The second thing would be, we have a

18   figure where that works a little bit and Dr. Nugent has

19   that now and that's looking at inflammatory responses

20   in healthy individuals.  Unfortunately, right before

21   our draft went out, the wrong figure became substituted

22   in and some of you noted the...and were a little bit

23   confused by that figure, and in that figure we looked

24   at responsiveness and increased inflammatory responses

25   in healthy individuals, as a function of inhaled dose

1  One thing Dr. Samet brought up in his comments in

2  relation to another chapter where I said that the

3  responsiveness is less in exercising individuals.

4  Well, that clearly goes against an increase in dose

5  with an increase in response.  I think that there are

6  reasons that we might have that perhaps.  It's a change

7  in responsiveness in asthmatics just due to the

8  exercise itself, making the increase in responsiveness

9  to the challenge agent less evident.

10	Another complication in establishing

11   that dose response relationship would be the delivery

12   of the challenge agent, the method of delivery, the

13   dose that's delivered and the underlying sensitivity of

14   the individuals, also limits our ability to establish

15   the dose response relationship.  So we are working on

16   that and we have provided two figures that weren't

17   necessarily included with dosimetry but in the

18   discussion of the clinical studies to address that

19   issue.

20	DR. HENDERSON:    Thank you very much. and

21   I think it's time for a break...

22	DR. SAMET:    Could I make just one quick

23   follow up.  I was really not so much thinking about the

24   dose response issue that you, as you said that you've

25   outlined part of it in the report now as I was the





98	100

1  animal extrapolation issue, and you know, there's

2  nothing on the comparison between the animal dose in

3  experiments where a health effect was seen in the

4  animal versus what you might see in the human and the

5  reason I think that's important is because it's

6  possible that in the risk assessment, we focused on a

7  particular endpoint where there might be another end

8  point that a person's even more sensitive to and the

9  only way to find that out, to compare sensitivity from

10   the animal experiments, would be to do some kind of a

11   dose extrapolation so I'm not even thinking  anything

12   sophisticated as a dose response but just the question

13   that if you compare two different...if you compare

14   different animals...

15	I'm sorry, I'm saying that wrong, if you

16   look at an endpoint that shows some sensitivity or some

17   effect due to NOx and you know the level of exposure of

18   the animal and you can estimate the lung size or the

19   weight of the animal and you can estimate the

20   ventilation rate of the animal, if you do a simple

21   extrapolation to, you know, an anthropometric

22   extrapolation basically to a human, would that be in

23   the same range as some other end point so for example,

24   if you look at the work that was done on the...that's

25   been done on I guess on rats and effect of...I'm sorry,

1  (WHEREUPON,   a brief recess was taken.)

2	DR. HENDERSON:    Okay, it's time to get

3  back together and discuss Charge question three.  For

4  anybody who wants them, there are copies of the slides

5  presented by NCEA this morning, out on the table so if

6  you want copies of those slides they are available.

7  Our helpers have been passing out revised copies of Ann

8  Smith's remarks, so if you find those in your chair,

9  they're to replace what's in your folder and with that

10   we will start off...

11	DR. COTE:    Rogene, could I ask a

12   question?

13	DR. HENDERSON:    Yes, go ahead, Ila.

14	DR. COTE:    I was sitting in the back and

15   continuing to think so I wanted to go back and ask Jon

16   a comment, about the application of the Hill criteria,

17   I mean, what he said is if you look at all those

18   criteria, it's hard to make them all kind of work very

19   well, but by the same token I think there are

20   situations, we're trying to design this so it works for

21   a broad array of chemicals, not just here.

22	So analogy is a perfect example, there

23   might be chemicals where you would rely on sort of

24   structured activities, so Jon, I just wanted to follow

25   up, would you suggest that we cut down the number or I



99	101

1  challenges on rats after exposure what that does to the

2  immune system or what NO2 does to the immune system.

3	Those studies, some of those studies are

4  very low concentrations, so estimating a equivalent

5  human dose would that effect with the number there,

6  that effect come in at the same levels as the effects

7  that you've seen, other effects that you've seen

8  clinically in humans, so I was thinking more in terms

9  of animal extrapolation rather than dose response from

10   the clinical experiments.  Does that make sense?

11	DR. JAMES BROWN:    I would just say yes,

12   that makes sense.  One of the difficulties that we'll

13   have in doing any of that is that we don't have the

14   degree of information available to us that we had for

15   ozone, for instance, with regard to any inter-species

16   extrapolations.  I did add a little bit of material to

17   the annex, not much, dealing with that, as well as the

18   effective age in humans.  I added two studies to the

19   annex to address some of those issues, but I'm not sure

20   how far we can go with them.

21	DR. HENDERSON:    Maybe we could continue

22   these discussions during the break.  The Chair needs a

23   break, I don't know about the rest of you.  But could

24   we come back at just a few minutes after eleven, take a

25   fifteen minute break?

1  mean our preference was to keep it broad in case one of

2  those were useful, you'd kind of have it incorporated

3  into the scheme but I understand what you're saying, a

4  lot of those are not useful frequently.

5	DR. SAMET:    So your point about if

6  you're using this elsewhere is fair, for example, for

7  analogy in thinking about let's say mechanism,

8  structure, activity, relationships or something.  You

9  know, what would probably be useful is to take some

10   test cases and see how useful they prove and that and

11   then make the decision, I mean I think try it out.

12	DR. COTE:	You know, where we've used

13   this scheme sort of on the toxic side of the house the,

14   more. I mean each one of these in certain situations

15   can prove to be useful but you're right that there's

16   probably about five of them that are most useful so...

17	DR. SAMET:    But again, but I think since

18   you're sort of setting out the process I mean if the

19   next version of the annex had, well, we tried this out

20   with two or three examples and here's what we learned I

21   think that would be a useful complement.

22	DR. COTE:    Okay, thanks.

23	DR. CRAPO:    I would add to that, Ila,

24   just say that I think you're wise to keep the broad

25   range, that to me the great value to the Hill criteria



104



1  is it makes one think about the types of logic and

2  materials that one wants to consider in making an

3  overall decision and by  having that full range of

4  concepts that he laid out there, it, even though you

5  may reject two or three of them as not being relevant

6  to any particular topic, it does force one to go

7  through and be sure that you're looking at it from many

8  different perspectives, many of which are important and

9  often are, by people who don't do this, are ignored, so

10   I think you're wise to keep it this way, given the

11   broad range that you want to use it.

12	DR. COTE:    Okay, I just wanted some

13   clarification on that.  Thank you.

14	DR. HENDERSON:    Good, okay, well, let's

15   move on to Charge question three which discusses the

16   integration of the health effects studies from animal

17   tox and clinical studies in epidemiology and that

18   discussion will be headed up by James Crapo.

19	DR. CRAPO:    Let me begin by saying that

20   I think that this draft is markedly improved and is

21   very at an appropriate point to launch this discussion

22   and I completely agree with the global overall

23   decisions that are recommended by the EPA staff in

24   this, that is the global one being that there is

25   clearly a substantial body of evidence that suggests

1  we start looking at these...at the risk assessment and

2  the decision to look at specific levels, literally two

3  hundred part per billion, two hundred fifty and three

4  hundred part per billion as ranges that one might

5  consider for numbers of people exposed and therefore is

6  that...are there..is there evidence of health effects

7  at those really, really low levels.   I think it

8  becomes your next question.

9	So you have to decide what range you

10   want to test to see what levels of population are

11   exposed and on that, I think I have a few comments that

12   I think really need to be looked at.  The...when you

13   really look to see what the data is to say you can take

14   this down to let's say .26 parts per million, and look

15   at it, the data is... really comes down as far as I can

16   tell, to two, well, actually one set of studies,

17   probably best illustrated in figure 3.1-2 on page 318,

18   where you have the airway responsiveness of allergen

19   challenge.

20	The document clearly demonstrates that

21   for normals the evidence of there's something going on

22   is probably ten times higher than it is asthmatics so

23   we're...we've picked a sensitive population and we're

24   looking to see if that sensitive population is at risk

25   for any kind of a meaningful adverse effect that might



103

105



1  that there's a health effect, that it's associated with

2  short term exposures, and raise the need to look at our

3  current standard and consider it in another form

4  besides an annual standard, suggesting that the high

5  levels that can occur in short term excursions can

6  cause disease in a variety of forms and on that I think

7  that you're right on the money and it's been well said

8  and the data completely supports it and I think if you

9  follow through all the Hill criteria, it is very strong

10   in meeting the ideas of consistency, strength of

11   associations, specificity, temporal relationships,

12   animal experiments, a gradient is shown, there's

13   plausibility to it.   I mean it actually meets all of

14   them, you can put almost every one of the Hill criteria

15   to support this basic concept that an annual standard

16   wouldn't address the issues of the short term

17   excursions that can cause these effects, so I think

18   that's very well done.

19	I think the challenge here comes when

20   you then go the next step and start to say okay, where

21   should I set that standard, what is it, what are the

22   low level effects and where should I begin to consider

23   it.  And there it gets a little more difficult.  When I

24   really push, if I integrate this with the discussion

25   we'll have tomorrow...this afternoon or tomorrow when

1  be at this very low level and when you look at

2  figure...I think figure 3.1-2 probably could be better

3  done is hard, you've got the plus and minus there and

4  in reality the only data that bears relevance to the

5  really low levels we're talking about are the ones

6  right against the left axis, and if you just globally

7  look at this one, there's three points above this line

8  and four below.

9	When you look at who those are and what

10   the studies are, there's really two studies that become

11   the relevant one and they are the ones by Strand and

12   the one by Brock, both done at the Carolinska Institute

13   and using the same technique and the same equipment,

14   same strategy and if you flip over to the other

15   analysis of this same question, is in summarized in

16   table 5.3-2 on page 510 which are the key human health

17   effects of exposure to nitrogen dioxide, looking at

18   clinical studies, and I'm focusing on these because the

19   epidemiology studies clearly demonstrate that there's

20   something going on and it's...the way it's done there's

21   mixed pollutants, it's difficult to interpret it

22   completely and it comes up with an association that

23   relates to a change that might be seen with let's say

24   twenty parts per billion alteration.

25	But it can't really tell you what the



108



1  low level is.  The animal experiments don't show

2  effects that go down to really, really low levels.  So

3  you're really, when you're talking about those low

4  levels, finding that threshold where we have a concern

5  and we want to start to test it, it really comes from

6  the human clinical studies dealing with asthmatics, and

7  then you find that it comes down to these two.  Table

8  5.3-2 again summarizes the first one as a summarizing

9  the Brock and the Strand articles and then the Meta

10   Analysis by Follinsbee and then it goes on to several

11   other studies, but as you go through these you find

12   consistent reports that...of inconsistent results and

13   negative results as well.

14	And I guess what I'm trying to say is

15   that the, when I change the question to where should we

16   be looking at levels in the range of two hundred to

17   three hundred parts per billion, I find some evidence

18   that is positive but I don't think this really meets

19   the Hill criteria, and I don't think the, well, I think

20   the discussion of this is a little biased on the

21   positive side.  I don't see a good discussion of

22   the...I don't see a really good discussion of the

23   conflicting negative studies

24	If you read pages 316, 317, 318 where

25   this is primarily discussed, there's a lot of

1  validations of that in terms of really meeting the

2  strength of the Hill criteria.

3	The...and I note that if you go back and

4  look at the figure that was just passed out, the

5  revised figure 3.1-1, appropriately revises the one

6  that was in there and clearly shows that there's, in

7  terms of inflammatory responses, this would indicate

8  that there's not any significant inflammatory response

9  measurable by any study until you start getting up over

10   two hundred ppm minutes.  Now to translate those, if

11   you set the standard at .2 parts per million for one

12   hour, that would be... let's see...it would be like six

13   ppm minutes, right.

14	When you translate to a total dose, so

15   I'm trying to express this number in terms of the same

16   thing on this curve.  What?  Well, .2 would give

17   you...you have to go for six minutes, is it 1.2...you

18   have to go for six minutes at that level to get one ppm

19   minute and then you'd have ten of those, let's say, so

20   roughly you're at five...so one point, yeah, you're

21   correct, it's 1.2 ppm minutes would be the exposure

22   that somebody gets from getting .2 parts per million

23   for one hour and figure 3.1-1 doesn't show any effect

24   in terms of the attempt to measure an overt

25   inflammatory response until you're at least up over two



107

109



1  discussion of the positive findings and yet I had to

2  really read between it and go back to some of the

3  original studies to find the negatives and why I have

4  points below the line that will balance it, so in terms

5  of consistency, it's not consistent yet at that level.

6  We have some studies that are positive and we have a

7  bunch that are negative and we have different end

8  points.  The strength of association is getting weak

9  and that's what you'd expect as you start going to the

10   really, really low level.

11	But what is trying...the question I'm

12   trying to ask is whether or not the range we should be

13   testing is much broader than two hundred parts per

14   billion to three hundred parts per billion, should it

15   be two hundred parts per billion to eight hundred or a

16   thousand parts per billion to consider that range and

17   what is the data within that.

18	As you move up to the higher range, the

19   data becomes more consistent and strength of

20   association increases dramatically.  I'm a little

21   concerned that the two studies that drive this entire

22   conclusion, that will lead most of our discussion

23   tomorrow, are two studies done at the same institution,

24   using the same technology and therefore it's not really

25   validating it...they're not strong independent

1  hundred ppm minutes so that's not a sensitive parameter

2  to detect what we're talking about.

3	The only parameter that's detecting

4  adverse effect at these really low levels is allergen

5  challenge to an asthmatic sensitized subject, and it's

6  only been done at one institution so far so I guess I

7  just want to put on the table and I think that the

8  discussion, as I read the discussion, I'm not sure is

9  balanced enough to pick up this issue and I think this

10   is the key thing at the end the day is where are we

11   going to go to set our recommendations and what....and

12   where do we want the EPA to go in terms of doing the

13   evaluation for subjects at risk, what's the range they

14   want to consider and so I think that we...I'd like a

15   little more work on this low level data in both its

16   presentation and its analysis of the conflicting data

17   as well to help us come to that conclusion.

18   Fundamentally I think we're on the right track.

19	We do have to set a, recommend a short

20   term standard and we need to analyze and determine what

21   the risks are and what the population at risks are so

22   the global conclusions that I have, I completely agree

23   with, and that's just addressing the lower end of this

24   that we need more work.

25	DR. HENDERSON:    Thank you, James.  And



112



1  maybe we can go on and get some more comments and then

2  discuss that.  The next person is Ed.

3	DR. AVOL:    I don't really have very much

4  in the way of comments, I actually generally thought

5  the chapter was much improved, and I thought it

6  presented a lot of information in a useful way and

7  particularly appreciated the summary comments at the

8  end of it, sort of trying to tie together each section.

9  I focused mostly on the longer term assessments

10   and...because that's the work that we've been doing and

11   found most of that to be reasonably presented and

12   communicated.  So I'll defer to later.

13	DR. HENDERSON:    Okay, now, I know John

14   Balmes had some comments on wanting to use the epi data

15   more, John, are you on the phone?

16	DR. BALMES:    I am.  Can you hear me

17   okay?

18	DR. HENDERSON:    Fine.

19	DR. BALMES:	Okay, yeah, well, I just

20   missed...I just came back on the phone so I just missed

21   most of Ed's comments....

22	DR. HENDERSON:    Well, it was mainly

23   James, I, we can summarize, James Crapo had some very

24   pertinent comments that at the low levels and he

25   pointed out we're using, basing things on the human

1  figure in the document, figure 3.1-1.

2	DR.  CRAPO:    They passed a new one out

3  to replace that  about twenty minutes ago.

4	DR. BALMES:    Pardon?

5	DR. CRAPO:    Yeah, they gave us a

6  replacement twenty minutes ago.

7	DR. BALMES:    There was a replacement

8  figure?

9	DR. CRAPO:    Yes, that's wrong, they took

10   that one out and gave us a new one.

11	DR. BALMES:    It really threw me for a

12   loop.  Okay, but nevertheless, the data with regard to

13   allergen responsiveness which is in figure 3.1-2 really

14   going back to James Crapo's point, you know, there's a

15   relatively limited amount of data, I think it's

16   accurately described in the document but I guess we'll

17   have a discussion later about how useful that is for

18   helping to determine an air quality standard that would

19   be appropriate and I guess my sort of major concern is

20   with regard to the discussion of respiratory effects

21   associated with long term exposure to NO2.

22	I don't necessarily disagree with the

23   final conclusion about these data being suggestive of a

24   causal effect but not sort of clearly enough to use for

25   an air quality standard, but I think that...I guess I



111

1  clinical studies and we only have, see responses at the

2  low level in allergen challenged asthmatics and at one

3  institution and at only one institution so there's a

4  concern, you know, does that meet the Hill criteria?

5  Now you can go ahead, John.

6	DR. BALMES:    Well, I would agree with

7  that comment of James or your summary of it.  So I

8  guess I would just follow up on the discussion that Jon

9  Samet led earlier in that I think that Chapter Three

10   would be strengthened if there were efforts to apply

11   the framework for causal determinations along the way

12   as opposed to just at the end...as opposed to Chapter

13   Five.

14	There are summaries about short term

15   health effects and long term health effects but the

16   summaries don't really analyze the quality of the

17   evidence as carefully as I'd like with regard to

18   fitting into the ultimate decision about the level of

19   causal determination so I just think the framework is

20   good but we need to apply it throughout the document,

21   and I guess overall I think the discussion of health

22   effects of short term exposure to NO2 in Chapter Three

23   is appropriate and fairly well integrated but there are

24   some problems, and one problem and maybe somebody else

25   brought this up, but I actually think there's a wrong

1  don't feel like the discussion is adequate enough.  I

2  think that these studies, one of which Ed has been the

3  co-author of many of the publications, you know, it's a

4  prospective, the children's health study and the Mexico

5  City study both which showed effects of NO2 on the

6  growth of lung function are important studies, they're

7  high quality studies, the document indicates that it

8  can't really be used with regard to NO2 air quality

9  standard setting because they are likely confounded by

10   effects of other pollutants, and while I don't disagree

11   that it's hard to disentangle the effect of NO2 from

12   other pollutants.

13	I have two points to make about this.

14   One, I think they are strong studies and they can't

15   just be so easily dismissed.  And there was a

16   tremendous investment with regard to the children's

17   health study that I think it was done, that investment

18   was made to try to provide good data that could be used

19   for air quality standard setting so just to dismiss it

20   because other pollutants aside from NO2 were also

21   associated with the same outcome, to me is a bit

22   problematic and too easy.

23	I think that, so that's number one...the

24   second point I would make is that to me this brings up

25   a discussion that we had with regard to ozone, is it

113



116



1  really appropriate to be focusing on single pollutants

2  with regard to health effects and the control of,

3  prevention of these health effects when actually it's

4  the combined pollution mix, and I realize that we're

5  stuck with a process and under the Clean Air Act that

6  is focused on single pollutants at this point but I

7  think it's really a problem.  We had a long discussion

8  about this with regard to ozone and I don't want to

9  forget it for subsequent pollutant discussions such as

10   for NO2.

11	SPEAKER:    Not long enough.

12	DR. BALMES:    Pardon?

13	DR. HENDERSON:    That wasn't us.

14	DR. BALMES:    I think that as somebody

15   else mentioned this morning that integration of the

16   animal tox studies with the epi studies needs to be

17   better, especially with regard to the fact that there

18   is a difference with regard to the dose response and

19   certainly high dose exposures, relatively high dose

20   exposures both with humans and animals have produced

21   health effects that are coherent with the epi but at

22   doses that are much higher, I think that George

23   Thurston's comment that like for ozone that with real

24   world NO2 exposures, lower levels are associated with

25   health effects because in fact there are co-pollutant

1  respiratory symptom epi studies, I mean it's almost

2  like a disconnect here.

3	I think that these paragraphs are

4  technically correct summary of what we know about

5  immune responses and increased risk of infection

6  secondary to NO2 from toxicologic studies but I

7  don't...they don't connect well, they don't integrate

8  well at this point in the chapter, in fact I find them

9  distracting and I would delete them.  There's

10   really...in fact there's nothing even stated about

11   respiratory symptoms in this...on these pages and

12   suddenly there's this long thing about immune responses

13   and infection.  That's a specific example of where I

14   think that the tox and epi data need to be better

15   integrated.

16	DR. HENDERSON:    Okay, thank you, John.

17   We'll go on to Terry Gordon.

18	DR. GORDON:    Most of my points have

19   already been brought up so I guess I can be a bit

20   brief.  I want to emphasize that what was said earlier

21   that at the end of each section it seems like there

22   could be...because this is supposed to be the

23   integration, there should be better conclusions on the

24   end.  Some of these sections just ended with a

25   description of a study and I like to be spoon fed and I



115

117



1  effects, you know, again we can't just ignore the fact

2  that pollutants and the real world come as a mixture

3  even though we seem to be stuck with a regulatory

4  process that is focused on single pollutants so there

5  has to be more discussion of that in the document in my

6  view.  Maybe I'll stop there, probably gone on too

7  long.

8	DR. HENDERSON:    Is that all, John,

9  you're not going on...I mean, we're interested, and we

10   just, but if you're finished....

11	DR. BALMES:    Well, I guess one other

12   thing...one other specific issue is that going to 390,

13   the section on...is it 390?  No, I'm sorry, section

14   3.4.5, there is...this is an effort to...I'm sorry,

15   just one second...there's a section that talks about

16   respiratory illnesses...oh, yeah, it's page 398 and

17   399, so this is in the summary integrating evidence on

18   long term NO2 exposure and respiratory effects and I

19   actually think that the discussion about growth of lung

20   function is fine and then there's a short paragraph on

21   398 about asthma and then there are several paragraphs

22   about the effects of NO2 on immune defense and the

23   increased risk of infection and I don't find that

24   this...that these two paragraphs actually provide much

25   support for asthma, the asthma's prevalence and

1  think the public should be spoon fed a little bit about

2  why that whole section was written.

3	Again, minor, but it still bugs me that

4  the figures in the legends, they still weren't

5  complete.  I still think anybody should come in and

6  just be able to read through and look at the figures

7  and understand what's going on and I found that that

8  didn't happen too often.

9	I don't know if this is a major point or

10   a minor point but when we got to airway height

11   responsiveness being the benchmark that's going to be

12   used in the REA, which I don't necessarily agree with

13   using that solely, there is a Oreck paper from 1976

14   that found effects at .1 ppm, which would be another

15   lab, it's significantly lower than what was been found

16   in other labs and it's not because George and I have a

17   really good old library to go to, it's just I remember

18   reading this back in '76 or '77 and being surprised

19   someone found something so low, so I would use that.

20	I semi-disagree in the figures and some

21   comments earlier, I don't know why people like to use C

22   times T, there's some papers that have been reviewed in

23   this document and they say that C times T doesn't work,

24   so I don't know why we always have to go to ppm

25   minutes.  Then finally in the balance of the



120



1  description integration of the controlled human

2  exposures in animals, I thought it was pretty nice and

3  I was actually very pleased with this until I got to

4  the REA and then I was sort of shocked that the epi was

5  not used as a benchmark and if that was the case, it

6  seems like the balance of this is off because so much

7  time was spent on the epi data and just so linking the

8  ISA to the REA I was surprised.  That's all.

9	DR. HENDERSON:    Thank you, Terry, and

10   again, we're bringing up questions I think we may

11   discuss at the end with, you know, with the NCEA, our

12   NCEA friends.  Kent Pinkerton.

13	DR. PINKERTON:    Okay, thank you.  I also

14   agree with many of the points that have been made and

15   will just simply reiterate that this really is an

16   improved document over the original one.  I find that

17   the new literature that's been included is extremely

18   helpful.  I really understand that the epidemiology and

19   the human clinical studies really far exceed any of the

20   animal toxicology studies and it's very clear that that

21   has a lot to do with the fact that there is a

22   disconnect between animal toxicology and human clinical

23   as well as epidemiology studies in terms of the

24   response based on concentration.

25	However, I think that the addition of

1  critical and then you've gone on in the document to

2  state that there have been some studies that have been

3  allowed...have allowed you to still show robust NO2

4  effects when adjusting for ozone or CO or PM and again

5  I think that just needs to be further confirmed that

6  that is the case and that's all the comments I have

7  right now.

8	DR. HENDERSON:    Thank you very much,

9  Kent.  Jon Samet.

10	DR. SAMET:    I think my written comments

11   really go back to what James said at the start, and I

12   think it's a little hard to separate sort of discussion

13   of Charge question three and our last one in fact, and

14   it seems to me the issue here is in the framework set

15   out there's a question asked about causation and then

16   this two stage process, the next is, is there...what is

17   the quantitative effect I guess to the extent that it

18   can be determined in the exposure range, a

19   concentration range of interest, and that I think is

20   distinct from asking that you...because you might ask

21   two different questions, one is, are causal effects

22   observed and then second is, are causal effects

23   observed in the range of interest for setting the

24   standard and it's a little hard not to blur this

25   thinking I think, and I think James is getting there a



119

121



1  these new studies that you have for the animal

2  toxicology would show effects at lower levels is good

3  and also the fact that we can establish some biological

4  plausibility at higher concentrations I think is

5  important to give some credence to what we're seeing

6  with the human epidemiology and clinical studies.

7	I think it's important, and I think

8  you've tried to emphasize this but maybe even more so,

9  that susceptible populations are really critical.  It

10   was very striking to see the difference in response

11   between a healthy adult and someone who has an

12   asthmatic condition as well as those effects that are

13   seen in children who have a physician diagnosed asthma

14   and again I think that that part of it needs to be

15   emphasized.

16	I think that it has been stated that

17   there is some concern that for some of these studies

18   that it may be very difficult to separate NO2 effects

19   from ultra fine particle effects and I think that

20   continues to be an area of concern, but one of the new

21   studies, well, relatively new studies, the Australian

22   study, the Piloto study I thought was really quite

23   fascinating, the intervention study showing significant

24   effects at one hour peak levels that are in the forty

25   to eighty parts per billion range that I think are very

1  little bit.

2	And then that hinges on whether you know

3  enough about the mechanisms in which the data are

4  available, whether that's observational or experimental

5  or hopefully both and sometimes you can tie this all up

6  nicely, tobacco smoking, radon, for example, I mean

7  these are easy ones, now we're dealing with a tough one

8  and so I think this is a very difficult...is a

9  difficult problem, and I think my comments then relate

10   back to how much can the epi and the toxicological

11   information understanding the mechanisms be used to

12   integrate and have certainty about the existence of

13   effects in the range of interest and then how can they

14   inform us on the quantification of those effects.

15	And that said, I mean I think the

16   document sort of skirts around these issues...it's

17   tough to formulate specifically and I don't think

18   things are brought together sharply enough around these

19   important questions to address them and maybe if there

20   was a better definition of what the questions are that

21   need to be answered so that the policy basis for the

22   NAAQS is laid out, it would help and I think charge

23   question three and charge question five, there's this

24   sort of missing piece in between which is, how certain

25   are we as to the existence of effects and the range of



124



1  interest, that's sort of the causation question, and

2  then related to that is do we have any understanding of

3  the quantitative aspects of those responses.

4	And charge question three is the very

5  general one and charge question five is truly general

6  and I think there's this missing piece which I think

7  we're sort of skirting around here.  I think James

8  probably addressed it most specifically in his

9  comments.

10	DR. COTE:    Jon, I'm sorry, the missing

11   piece being not a clear articulation of...

12	DR. HENDERSON:    Ila, I can hardly hear

13   you and I'm sure they can't hear you on the phone.

14	DR. COTE:    This is Ila Cote, I wanted

15   just to ask a clarifying question for Jon. When he said

16   the missing piece, do you mean a clearer articulation

17   of or certainty that this chemical causes the effect or

18   what....I was a little unclear about the missing piece.

19	DR. SAMET:    Well, so what is really

20   the....so what do you have to glean from all the

21   scientific evidence in the ISA to inform the selection,

22   the decision as to whether the NAAQS needs to be

23   revised and if so what form it should have, what level

24   and so on, and so what are the missing pieces?

25	Well, some idea of the certainty with

1	DR. SCHLESINGER:    Can you hear me?

2	DR. HENDERSON:    Oh, fine.

3	DR. SCHLESINGER:    Okay, good.  Well,

4  again, some of the comments that I had after re-reading

5  this again after I sent my initial comments on have

6  been addressed.  I do think it's an improved draft, I

7  do like the summary comments after each section and I

8  did like table 5.31 on page 523, I thought it was a

9  good summarization of the points being made, and it was

10   interesting to compare the previous document with this

11   document.

12	I think John Balmes alluded to...unless

13   I misinterpreted him he was talking about what sounded

14   like equal weighting related to causality of all the

15   studies that are being presented without any discussion

16   of whether some of the studies may be less valuable in

17   that regard and if that's what he meant then I agree

18   with it and if it's not what he meant then maybe I

19   don't agree with it, so, John, is that what you meant?

20	DR. BALMES:	In part, yes.

21	DR. SCHLESINGER:    Okay, good, because

22   there seems to be an equal weighting of the studies, in

23   the epi, I guess John can address it more, but in the

24   toxicology and the clinical human, they seem to be

25   equally weighted and some of these studies may have



123

125



1  which effects exist at exposures experienced in the U.

2  S. at present and second, what are the quantitative

3  relationships that may exist between exposure and risk,

4  and this...the charge questions are very general,

5  chapter three kind of ends generally and then, you

6  know, chapter five is not pulled together in that way

7  so...I think it's very hard to draw out those

8  quantitative relationships.

9	I think that's what you're trying to do

10   with those figures, there's scant evidence so there has

11   to be uncertainty but some of the uncertainty in this

12   specification of how uncertain we are, has to come from

13   putting the epidemiological observations together and

14   there you're trying to make the story of, you know,

15   coherence and so on and consistency among the studies

16   with what we can infer from about mechanisms observed

17   at these higher effects and whether we think those

18   mechanisms are important in underlying the effects that

19   are observed observationally.

20	DR. HENDERSON:    This is a tough one as

21   people have said, it's not real -- and I thought,

22   James, that you outlined it very well.  We'll add a

23   couple...well, is Rich Schlesinger....

24	DR. SCHLESINGER:    I'm here.

25	DR. HENDERSON:    Okay, Rich.

1  some flaws.

2	Getting back to the issue that was

3  brought up by our panel and also by specifically one,

4  at least one of the public commenters, was the

5  integration of the toxicology, human clinical and epi,

6  and there have been a number of reasons why there is an

7  apparent sometimes lack of coherence and this has come

8  up in other issues as well, it clearly came up, as Jon

9  Samet will likely remember, in the discussions for the

10   PM  health effects during our long period of preparing

11   for the NRC report and there's a number of reasons and

12   so, but I don't think it's addressed in the document

13   and there are a number of reasons why there can be this

14   apparent lack of consistency.

15	George Thurston brought up one, which

16   had to do with the interaction of gases, whether it's

17   gas/gas interaction or gas/particle interaction and the

18   Borne paper, I remember that one, too, so there's an

19   issue of that, there's an issue of dose differences

20   between the animals and the humans and dose differences

21   between human clinical and epidemiology exposures but

22   one thing that wasn't brought up is there can be major

23   differences in the sensitivity of biological endpoints

24   that are being evaluated and you can do even one human

25   clinical study and have significant differences at the



128



1  same exposure level between different end points, and

2  the epi and the tox and the clinicals don't always look

3  at the same end points even though sometimes they

4  appear to so that's another issue that could be

5  addressed and not relate to the differences in the

6  models that are used, whether it's an asthmatic or a

7  normal human or a mouse or a rat so.

8	So dosimetry differences in the models

9  used, animals to human and human to human and

10   differences in the end points, are all issues that are

11   hard to tease out but could explain some of the

12   differences so just because there are differences in

13   the dose response does not necessarily, should not

14   necessarily result in the conclusion that there's no

15   consistency or no coherency, but I think again that

16   needs to be addressed in the document.  I did pick up

17   on what John said about page 399, that the discussion

18   of defense mechanisms did not seem to fit in with the

19   rationale for asthmatic effects, but I thought it was

20   my imagination but I'm glad John brought it up because

21   it didn't seem to fit with me.

22	And I guess Terry brought up the issue

23   of C times T and it may not be valuable but one problem

24   is that most of the time when the C times T is done, is

25   C to the one power times T to the one power and there

1  create a dose response from the controlled exposure, I

2  mean that's, when you look at this plot, you know, that

3  was handed out, the tendency is oh look and to fit a

4  sigmoid response curve to this and fit it in and to me

5  that would be so wrong in so many ways because you

6  know, the fact is, this is exposure not dose and a lot

7  of this NO2 probably is not reaching the lung at these

8  low levels so I wish somebody would do like a

9  calculation and say, knowing what we know about NO2

10   alone how much would be scrubbed out and you know, is

11   it plausible that, let's say it's 200 ppm minutes could

12   be scrubbed out.

13	Well, then suddenly, you know, those two

14   non-response are zero exposures in actuality, zero

15   doses in reality because there were no particles to

16   carry them deeper into the lung so they never got to

17   the lung to do the damage.

18	So I think that they are proof of

19   principal, but can't really be used for dose response

20   and I think that it's the way the document is written

21   that you fall into this trap and especially since the

22   REA then goes on to use these as benchmarks and ignores

23   the epidemiology for risk assessment that you tend to

24   focus on these in trying to look for dose response

25   where we shouldn't be looking, you know, I don't want



127

129



1  was a paper, I think it was someone named Ken Berge a

2  while ago where depending on what you're looking at,

3  there may have to be an exponent for the C and an

4  exponent for the T and sometimes if you use those then

5  you come up with a better model that relates C times T

6  to the ultimate biological effect so it may not be that

7  you shouldn't use C times T, it may be that you may

8  have to use an end....an exponent other than one for

9  the C and/or for the T, depending on the mechanism of

10   action of the material.  And other than that I think

11   that's it for now.

12	DR. HENDERSON:    Thanks, Rich, I

13   appreciate your coming in on the phone and George

14   Thurston is the last one and then we'll have a general

15   discussion.

16	DR. THURSTON:    Okay, thank you.  Well, I

17   just wanted to follow up briefly on the discussion.  I

18   think we've had a good discussion, but I really think

19   that Kent Pinkerton was right in saying that the

20   exposure studies, the controlled exposure studies are,

21   they give credence to the epidemiology and that is the

22   way they should be used, not to set dose response and

23   you know.

24	I think that was where, you know, I'd

25   say I sort of disagree with James Crapo's efforts to

1  to knock the toxicology, we need that and we need the

2  controlled exposures for proof of principal and to the

3  extent we can make them like the real world we can

4  learn more and more.

5	I think that we need a set of

6  experiments where we expose, look at these end points

7  with both NO2 and particles and do those, you know, the

8  paper I submitted by Borne. But do it using today's

9  methods, today's end points and more realistic levels

10   of exposures to try and get down and bring these two

11   together, but I think the way the document is written

12   now there's too much of an emphasis on the toxicology

13   for use that it's not really intended for. To me it's

14   intended for looking at an important part of the Hill

15   aspects as he calls them but not for dose response.

16	And then so, and I thought that Rich

17   Schlesinger, I agree with what he said, he brought a

18   much more rich, you know, I'm just talking here about

19   the particle interaction, but as Rich Schlesinger

20   pointed out, there are other reasons and I thought he

21   did a very good job of doing that and that and his

22   discussion should be incorporated into the document of

23   reasons why you might not see an exact correspondence

24   between the controlled exposure studies and the

25   epidemiology, so that really is important to add a



132



1  discussion of, you know, and I point this out in my

2  written comments, there's a gap between the two and

3  there's to me it's not inconsistent at all, that's what

4  I would expect, I would expect these bottom ones to go

5  to have no response because there's no way the NO2

6  could get deep in the lung to have effect so it seems

7  very consistent to me.

8	And then the only other thing was that I

9  talked about in the written comments was that these

10   questions at, and I think everybody else is sort of

11   saying similar things, but these questions on page 5-1

12   that are laid out at the beginning of the integrative

13   summary, it might be useful at the end to say, well,

14   how can we answer these questions now that

15   we've...because it never really...they are never

16   directly addressed at the end.  In other words, they're

17   laid out as something to think about and then it goes

18   through the document, but at the end it would be useful

19   to say okay, what do we...what are our conclusions vis

20   a vis each of these questions and what are the gaps?

21	And the one that I thought was most

22   important and I've heard other people say it, you know,

23   and I think this is what Dr. Crapo was trying to get at

24   is at what levels of nitrogen oxides exposure do health

25   effects of concern occur and that's come up in the

1  minutes or one hundred and fifty ppm minutes but for

2  those things that are responsive...I think from the

3  standpoint of sort of looking at it, it might be useful

4  just to sort of get a gestalt feel for whether or not

5  you're seeing an effect that you could at the end of

6  the day say yes, it meets the criteria reasonably well.

7  It gives us some indication that there's an effect that

8  we ought to consider as important and then from there

9  separate out to see how it applies to the more

10   quantitative basis because I think really what the

11   toxicology and the human studies has to tell us is

12   whether there is a concept of a mechanism and I think

13   that might be useful and it can be done in a whole

14   series of these little sort of panels of running

15   through the criteria.  It allowed me anyway to sort of

16   look at it in one picture rather than over ten pages.

17	DR. HENDERSON:    Can you describe that

18   gestalt, I mean what is...

19	DR. SPEIZER:    Well, I looked at the,

20   let's say for host defense and immunity, I got more

21   pluses and I had mostly pluses.  Now, you know,  I

22   don't know what the quantitative levels are but it

23   looks like there's an effect in my mind, if I look at

24   the short term airways responsiveness, I got a lot of

25   plus minuses, you know, but not as consistent across,



131

133



1  public comments and I think throughout here that

2  everybody wants, that's the sixty four thousand dollar

3  question as we used to say, and that's what we have to

4  get at and I, personally I think that the epidemiology

5  is important in answering that question.

6	DR. HENDERSON:    Thank you, George and

7  Ellis thanks you for...he was also, I mean, Ellis

8  Cowling had that same request to answer the question so

9  now I would open it up to others and I think Frank

10   Speizer, and then I'll have Ron.

11	DR. SPEIZER:    Yeah, I was struck by the

12   sort of independent assessment of each of the

13   parameters and so what I tried to do is I constructed

14   sort of mini-tables of the Hill aspects, across the

15   various response categories and then just qualitatively

16   tried to indicate pluses and minuses for some of the

17   studies and I think the...

18	I think we need to separate whether or

19   not there's a consist....whether or not there's a

20   gestalt really of an effect that we can agree...that

21   can be agreed upon and separate that out from the

22   quantitative aspect of it.  Because I think if you run

23   down the line and find null effects, let's say, for

24   lung function change then you don't have to worry about

25   whether or not it's occurring at two hundred ppm

1  down the line, that's just another way of looking at

2  it, I mean they summarized it here, the data is all

3  here, it's just...it was hard for me to sort of read

4  through this and come to a conclusion so I tried to lay

5  it out in a picture.

6	DR. HENDERSON:    I like the idea that,

7  you know, you call it the gestalt, I mean whether

8  there's causality, you know, there's consistency that

9  suggests a causality and then you go to what level, you

10   know, then you have to ask what level and that is the

11   fifty thousand dollar question.  Ron?

12	DR. WYZGA:    I had two comments.  Number

13   one, I'd like to see a little bit fuller discussion of

14   the whole issue of co-pollutants.  At least in some of

15   our work the co-pollutants that have turned out to be

16   the most important have been EC and CO and the emphasis

17   of the document is given to PM and I know that EC is

18   not regulated per se right now, and that may make it

19   difficult to look at but I'd like to sort of, sort of

20   as you look at each study, indicate whether or not they

21   attempted to consider co-pollutants, whether or not and

22   if so which co-pollutants were considered.  I think it,

23   the difficulty is in interpreting the epidemiology

24   studies is really trying to sort out NO2 from what else

25   is there and I think anything you can give us further



136



1  to help along those lines is going to be helpful.

2	And then the second thing several times

3  in the document you've mentioned studies were not

4  considered because, quote, they did not inform. I think

5  it would be very useful, they are in the appendices and

6  I think it would be useful to sort of explicitly say,

7  you know, why they were excluded, and not inform. And

8  in some cases they may inform qualitatively.

9	For example, there may be a study that

10   says a negative study but it didn't give you very

11   specific data so that you could really quantify it but

12   I think it's important to know that it was supportive

13   or not supportive, of a specific argument, so to the

14   extent that you could put in anything qualitative about

15   these studies I think would be helpful as well as when

16   you did exclude them be more explicit as to why they

17   were not included.

18	DR. HENDERSON:    Thank you, Ron.  Dale.

19	DR. HATTIS:    Yeah, I think it's quite

20   important to make an attempt to bring together the

21   toxicology and epidemiological observations and in

22   doing that you need to note some particular aspects of

23   the toxicological experiments.  Often experiments are

24   done in relatively modest population sizes and there's

25   a tendency to do when you're measuring a continuous

1	DR. HENDERSON:    Yes.

2	DR. CRAPO:    I would suggest it would be

3  really helpful for our discussion now to put the rubber

4  right straight down on the road and talk about the real

5  question which is what level...I think there's a

6  consensus that we need to talk about a short term

7  standard and consider the possibilities of that and the

8  question that we should really ask and answer is what

9  are the levels that raise our concern?

10	We can say a level at which we're not

11   concerned, a level we're concerned at, for which

12   there's a consensus, and to kind of initiate that

13   discussion, I'm going to propose that the correct level

14   is not two hundred to four hundred...two hundred to

15   three hundred parts per million but rather, no, point

16   two to point three, that it's really one to two for the

17   point of discussion so just let me make a counter

18   argument to the one that's being used for our

19   discussion tomorrow.

20	That is, that they're five to ten times

21   too low and the data doesn't support that and I'm

22   curious to see how you'll defend this, George, 'cause

23   you want to argue that the epidemiology is going to

24   answer it, no, I would...what I would point out is that

25   if I'm arguing in favor of a higher level for the area



135

137



1  variable like lung function parameters, there's a

2  tendency to do group averages and a comparison of group

3  means, and this tends to obscure the variability among

4  people, of people within particular groups so there's a

5  tendency if you're studying the asthmatics to say,

6  okay, these are the asthmatics but asthmatics

7  themselves may well have an appreciable variability as

8  we saw I think in the analysis of the SO2 observations.

9	So we had some people...some of the

10   asthmatics were very much more responsive than others.

11   Some of the asthmatics were very much more responsive

12   in that case than others and so you could conceivably

13   produce an effect that's observable epidemiologically

14   on the instance of a rare outcome by having, you know,

15   a small minority of the asthmatics say respond at a

16   particular level and not really be inconsistent with

17   the human exposure studies.

18	And so the human exposure, you should

19   note when you're reviewing the human exposure studies

20   whether the effect was examined on an individual basis

21   or on a group average basis and if you...where it's

22   possible to do the, a threshold distribution type of

23   analysis for the human observational studies then I

24   think it's important to do that.

25	DR. CRAPO:	Rogene?

1  of concern, I would point out that the only data at the

2  low level are a couple of studies on asthmatics out of

3  one institution and if your approach is to take the

4  lowest possible study of any kind that shows an effect,

5  that's one approach and then you could use that, then

6  you clearly would go to the level that was proposed in

7  the risk analysis for tomorrow.

8	If your proposal is to use a Hill

9  criteria to look at it, you're going to look for a

10   range at which there's lots of evidence from different

11   sources where there's consistency, where's there's

12   strength in association, where there's plausibility,

13   where there's analogy and now if you're looking at the,

14   for example, at the inflammatory data and the figure

15   that was just passed out, you don't start to see any

16   effects of inflammation until you get up to the one

17   study that was done at .6 parts per million for three

18   hours so you know, a one hour standard at one part per

19   million would give you sixty ppm minutes so you're...

20	I'm trying to argue that this would say

21   the earliest inflammatory effect that's measured in

22   these studies came at the range of say one to two

23   parts per million.  If you looked at the adult, I mean

24   the non-asthmatic humans, the earliest nonspecific

25   effects were identified in the one to two parts per



140



1  million range so I'm starting to get some consistency

2  across different results.

3	I'm going to argue that that higher

4  range would more adequately meet the Hill criteria as

5  having a plausible place where you could defend that

6  there's going to be a significant biological effect and

7  to drive it lower you're going to need something more

8  than just the two asthmatic studies and I'm not

9  necessarily telling you what I believe the right answer

10   is.   I'm trying to put it on the table so you can

11   either prove me, you can prove me wrong if you're going

12   to take a lower level but you need to show me your

13   logic for why we would do that.

14	DR. HENDERSON:   I propose this, I think

15   you're very good, James, at putting, laying down the

16   gauntlet and getting our discussion going but maybe we

17   pause for lunch and then at 1:00 o'clock we come back

18   and see, we've got some time here. Really charge

19   question four is just about susceptible populations,

20   and I think we can do that but we're really talking

21   about charge question five which is, you know, kind of

22   where the rubber meets the road and so I think we can

23   afford to eat lunch now and come back at 1:00 and take

24   up James's challenge and why isn't it in the range of

25   one to two ppm.  Okay.

1	We haven't actually pulled that out, so

2  I think we're taking a message that we need to pull

3  that out into a nice one place where it's more concise,

4  but your questions about the studies, for the clinical

5  studies,  is whether there are just two studies at

6  which low levels are seen?

7	DR. HENDERSON:    Yes, I'm -- that was

8  what was proposed, that there were two clinical studies

9  on asthmatics at the Karolinska that suggested

10   responses at levels lower than the current standard.

11	DR. BROWN:    Do you want me to...

12	DR. HENDERSON:    Go.

13	DR. BROWN:    ...go ahead?  This is James

14   Brown and first I would like to say -- I'll start over

15   again.  First I would like to say that I think there

16   are three that we had up there marked as...

17	SPEAKER:    Could you speak up, please?

18	DR. BROWN:    ...as positive, but...

19	SPEAKER:    We can't hear you.

20	DR. BROWN:    You can't hear me?

21	SPEAKER:    No.

22	DR. BROWN:    Okay, let's try again.

23   Let's move the microphone in the direction that I'm...

24	DR. HENDERSON:    And I'm laying this out

25   as just, you know, are we right?  I just want to see



139

141



1  (WHEREUPON,   a lunch break was taken.)

2	DR. HENDERSON:	Okay, back to work.  We

3  ended with the start of a good discussion on answering

4  the question of -- really the basic question of do we

5  need to lower the -- do we need to alter the current

6  standard and the -- I believe, James, I'm right in

7  you're saying that, as far as you could see, this was

8  the -- the suggestion that we lower the current

9  standard was based on two clinical studies in

10   Karolinska.  Is that -- I mean am I paraphrasing that

11   right?

12	DR. CRAPO:    Yes, that's the way I read

13   the document.

14	DR. HENDERSON:    So what I would like to

15   hear is does NCEA see it that way?   I mean, can you --

16   is that a correct statement?

17	DR. ROSS:    Well, can I remind us that

18   we're talking -- when we're talking about the ISA,

19   we're not going to answer the question of whether or

20   not we should lower the standard.

21	DR. HENDERSON:    Okay.

22	DR. ROSS:    But we will try -- we will --

23   we will try to characterize the levels and in Chapter 5

24   in the bullets we've tried to list, you know, what are

25   the levels at which these effects are seen.

1  where we are, so tell me if I'm wrong.

2	DR. BROWN:    I'd like to have everybody

3  look at or think about that figure, the 3.12, and we

4  see studies that are on the positive side and we see

5  studies that are on the negative side and what I would

6  like to note there with the positive and negative is

7  that that refers to statistical significance in -- as

8  in a change in airways hyperresponsiveness and that

9  only shows studies to which people were challenged with

10   an allergen following NO2 exposure.

11	DR. HENDERSON:    Okay.

12	DR. BROWN:    So the first thing with

13   regard to hyperresponsiveness, Dale Hattis noted that

14   that response is log normally distributed, at least

15   with regard to a non-specific challenge like SO2.

16	So we might also expect these specific

17   airway resistant -- or these -- these specific

18   challenges to also be a log normally distributed

19   response.

20	So we've got small studies with limited

21   statistical power.  So we see some that are

22   statistically significant, we see some that are not.  I

23   don't think that means that there's not an effect.

24	If we went back and we looked at what

25   was in the '93 document, we see that these studies and



144



1  that figure don't stand on their own.  In the '93

2  document, we went through 29 or 30 studies that looked

3  at changes and specific challenges or challenge to a

4  specific agent such as histamine, SO2, et cetera.

5	There were very few studies that had

6  looked at allergen challenges.

7	Now, looking at those non-specific

8  challenges, those responses also, one study was

9  statistically significant, one was not.  Dr. Gordon

10   mentioned the study by Oreck  in 1976.  That was one of

11   the studies that was mentioned in our 1993 document.

12	I don't remember whether that one was

13   statistically significant or not.  There were 13 out of

14   20 people that had an increase in responsiveness, but

15   when we looked across all of those studies, the 29 or

16   30 studies, and did a meta-analysis of those, indeed we

17   found that there was a statistically significant effect

18   of NO2 on airways responsiveness to non-specific

19   challenges.

20	So I think it's dangerous to only look

21   at those few studies looking at the allergen challenges

22   in Figure 3.12 because there was a lot of data

23   available in the '93 document that I don't think we

24   should lose sight of and in Chapter 5, indeed we had

25   two bullets.

1  significant, but what you're not showing is the point

2  estimates and any indication of the precision of the

3  findings.

4	So what you really have is bits and

5  pieces of information, some of which are statistically

6  significant and some not, but the figure really at

7  least should show sort of what is the delta of airways

8  responsiveness and a confidence interval around it and

9  then after that you're left with a question of whether

10   you should do some sort of quantitative summary of the

11   evidence and then do a meta-regression on the exposure

12   to see if there's an exposure/response relationship.

13	So I think what you've done is taken the

14   information to sort of its possibly least informative

15   dimension in this plus/minus display when you could be

16   doing far better.

17	DR. BROWN:    Can I go ahead and follow up

18   on that?

19	SPEAKER:    Can't hear.

20	DR. HENDERSON:    Okay.  You follow up and

21   Leanne will come after that, okay?

22	DR. BROWN:    And I was just going to say

23   that I actually tried to do that for these studies, but

24   in these studies, the protocols are very inconsistent

25   as to whether they were looking to produce a 20 percent



143

145



1	One mentioned that there was some new

2  data available looking at allergen challenges, but the

3  second bullet -- and that was on Page 513 -- tried to

4  reflect back on that former data from '93 to say that,

5  well, if we look at half-hour exposures, there was a

6  statistically significant effect for exposures between

7  .2 and

8  .3 ppm  from a meta-analysis.

9	Not in that bullet, the Oreck study and

10   at least one other one -- I think it was Ahmed or

11   something like that -- did go down to .1 ppm and had

12   more subjects showing a change in responsive -- an

13   increase in responsiveness following NO2 than did not

14   and I'll stop with that.

15	DR. HENDERSON:    Okay, John?

16	DR. SAMET:    I had a specific comment

17   about this table, which had -- figure, which had

18   troubled me because of your plus/minus interpretation

19   and this conversation makes me want to say that the

20   figure really has to be redone and there are several

21   things you should be doing, but what you should not be

22   doing is either showing the figure as you did or

23   talking about it as you just did because, as you said,

24   these are small studies.

25	They may or may not be statistically

1  change in FEV1 across the board by changing the dose of

2  the allergen challenge or whether they were trying to

3  give a constant dose of allergen which was not

4  consistent between studies.

5	So I was not able to do a magnitude of

6  change comparison between the studies and nor was the

7  data available in the majority of the studies to

8  actually be able to say exactly how many subjects had

9  an increase in responsiveness or a decrease such as was

10   done in the meta-analysis in '92.

11	DR. SAMET:    Right, but your figure can

12   -- it could still be better even if you can't do the

13   quantitative summary.  Then, you know, the question of

14   whether this is important enough that you need to go

15   back and get the data is a different matter because,

16   you know, you can write to the investigators and say,

17   you know, we're doing this quantitative synthesis of

18   the evidence and, you know, are your data available to

19   us?

20	DR. CRAPO:    I would also like to add

21   that I have difficulty giving credibility to justifying

22   this with a meta-analysis published in 1992 that I

23   don't have any access to right now.  I mean I -- that's

24   -- if it's important, if it's shaping the most

25   important decision we're doing, it should be in the



148



1  center of this document.

2	DR. BROWN:    It was actually in Chapter

3  15 of the '93 document.

4	DR. CRAPO:    Well, I understand that, but

5  that wasn't given to me and I didn't intend to read it

6  to start with.

7	This is -- I'm reading this document and

8  if -- if -- if that is the -- if that is the data that

9  convinces you where we ought to go with this, why is it

10   not front and center in bold type on this document?

11	DR. BROWN:    We could pull forward tables

12   out of that old document and tables out of the

13   Folinsbee  1992 publication to illustrate some of...

14	DR. CRAPO:    I would just like to say we

15   shouldn't consider anything that's not in this

16   document.  So, if you think it's important, put it in

17   this document.

18	DR. HENDERSON:    Okay.  Leanne, you had

19   something you wanted to say?

20	DR. SHEPPARD:    It's just a minor follow

21   up on John's comment, which he stated so well and, you

22   know, by presenting a figure this way we're making

23   statistical significance synonymous with scientific

24   importance and we shouldn't be doing that.   We just

25   can't be making that binary categorization and drawing

1  speaking and I don't think that this is enough to hang

2  our hat on and I did understand that the staff was

3  using the '93 criteria document and the Folinsbee

4  meta-analysis in their -- with regard to their thinking

5  in the risk and exposure assessment, but I also have to

6  say that I agree with James, that you have to bring

7  that more front and center. I mean, if James couldn't

8  figure it -- figure that out, it means that other

9  readers won't either.

10	DR. HENDERSON:    Good point.  Thanks,

11   John.  Are there others?

12	We can go on and discuss Charge Question

13   4, which is just about susceptible populations and then

14   get to Charge Question 5, which is really the bottom

15   line of this ISA and includes some of the same things

16   we're talking about.

17	So  why don't we do that?  John, you're

18   the lead discussant on Charge Question 4, which is

19   related to susceptible populations.

20	Why don't you go ahead and give us your

21   comments in response to that charge question and we'll

22   continue our discussions as we go along?

23	DR. BALMES:    Yes, I can see in my

24   written comments that I mislabeled the charge question.

25   It should have been four.  I said two there in my



147

149



1  conclusions from it.

2	DR. BALMES:    This is John Balmes.

3	DR. HENDERSON:    Okay, John.

4	DR. BALMES:    As the senior author of

5  Study 11 on this figure, I just wanted to support what

6  John said and what Leanne extended.

7	You know, these are very small studies.

8  I think P values are totally problematic with regard

9  to, you know, whether an effect was really seen or not.

10	In our study, even though it's listed as

11   a negative study, which was, in fact, the way it was

12   reported, we also reported that there were some

13   individuals that appeared to be responsive to NO2 with

14   regard to increased lung function changes after

15   allergen exposure, after allergen challenge.

16	So even in, you know, a so-called

17   negative study here, there were some responders and I,

18   for one, would be happy to give EPA individual subject

19   data if they do try to pursue a more rigorous

20   meta-analysis, but -- yes.

21	I guess I -- I would just sort of echo

22   James.  I find the -- as a contributor to this data

23   area, part of the reason we did our study is that we

24   were impressed with the paucity of data and I'm still

25   impressed with the paucity of data here relatively

1  written comments, but in any event, I thought Chapter 4

2  did an adequate job of discussing the public health

3  significance of exposure to ambient NO2, especially

4  within the context of susceptible and vulnerable

5  subpopulations.

6	The document indicates that individuals

7  with pre-existing respiratory disease, especially

8  asthma, children, and older adults may be the most

9  susceptible to the effects of NO2 exposure and I think

10   that's an appropriate assessment of the evidence.

11	The only thing I really disagreed with,

12   with regard to how the chapter came out is that it

13   seemed to me they were trying to stretch -- the author

14   was trying to stretch towards a gender age-based

15   difference in susceptibility based on the fact that

16   incidence of asthma differs between boys and girls with

17   age with, you know, younger boys being more susceptible

18   and then in adolescence and especially late adolescence

19   girls become -- have a higher incidence with asthma.

20	You know, the few studies that looked at

21   gender differences, you know, the results are

22   inconsistent, so I think we shouldn't go there at all.

23	So I think that part of the chapter --

24   it's only a few sentences -- should be deleted, but

25   otherwise I thought Chapter 4 was fine.



152



1	I did see Jon Samet's comments that,

2  while he agreed with the assessment of the chapter with

3  regard to susceptible subpopulations, that he wanted to

4  make sure the agency thought with each pollutant about

5  whether there are differences and didn't just sort of

6  throw up the usual people with pre-existing respiratory

7  disease, children, and older adults.  So I agree with

8  that point of Jon's.

9	DR. HENDERSON:    Okay.  Thank you, John.

10   Steve, do you have comments?

11	DR. KLEEBERGER:    I do.  I have a few.

12   So I actually thought the document, like John, was --

13   this particular chapter was well written and I don't

14   have much to argue with what was written, but rather

15   what was perhaps not written.

16	I took this charge question at face

17   value and so the question asked, "What are the views of

18   the panel on the characterization of groups likely to

19   be susceptible or vulnerable?"

20	The word "likely" is what I took most to

21   heart and so the populations that we have identified in

22   this chapter, of course, are the asthmatics and

23   children and in particular the children with an

24   increased risk of infections, but there are a number of

25   other potentially likely susceptible groups that I

1  differential responsivity.  In fact, they might be more

2  responsive to ozone exposures in terms of inflammation

3  and airways reactivity.  It's, to my knowledge, not

4  been investigated with NO2 at any level.

5	A pre-existing condition might be very

6  low birth weight infants.  They are at risk to severity

7  of infection in terms of RSV.  There are papers that

8  are going to be coming out shortly that show that they

9  are also more at risk in terms of responsivity to viral

10   infections aside from RSV like influenza and air

11   pollutants certainly could also be put into

12   consideration as a environmental factor that could

13   enhance risk for hospitalization and there are a couple

14   of studies that actually have shown that.

15	And I think in light of the fact that

16   prematurity in our population of premature kids is

17   increasing at a fairly high rate according to a

18   National Academy of Sciences study, this is potentially

19   a very important subgroup that should be -- should be

20   tracked and I would argue that we need to be -- we need

21   to be studying.

22	Age, independent of sort of this

23   premature cohort or very low birth weight kids, infants

24   I think are at risk.  There was a paper that was

25   published in 2006 in pediatrics that showed a



151

153



1  think ought to be at least discussed and especially in

2  the face of some of them being at increased risk in

3  studies with other criteria of pollutants and I think

4  this is where the -- at least from my perspective, the

5  problem with the NO2 studies that are already out there

6  or are not out there with respect to other criteria of

7  pollutants where they've actually investigated some of

8  these susceptible groups.

9	And I guess what I would suggest, and in

10   a second I'll go through a couple of these

11   subpopulations, is that we create a table of likely

12   susceptible groups.

13	Whether this then comes as a suggestion

14   of where research needs to go, holes in the literature,

15   I don't know.  I mean this is something that we could

16   sort of work out, but I think limiting our discussion

17   to only the -- the -- the asthmatics, perhaps the

18   elderly and the children is not sufficient in terms of

19   understanding susceptibility.

20	So I'll bring up a couple of other

21   possible conditions, which may enhance responsivity to

22   a number of criteria of pollutants and perhaps NO2.

23	Obesity is one of these pre-existing

24   diseases that I think the literature is coming out now,

25   at least with respect to ozone, that obesity imparts

1  relationship between -- this is an epidemiologic study

2  showing the relationship between NO2 and SIDS.  This

3  wasn't in the report, but probably should be included,

4  so this is another potentially important subgroup that

5  is susceptible.

6	And finally, although genetics was

7  characterized and discussed in the Chapter 4, the

8  problem with most of the studies that have been done

9  with NO2, and the same argument could be made for ozone

10   and particulates, is that most of these studies are

11   sort of under-the-lamp-post kind of studies where these

12   are genes that are potentially important based on some

13   biologically plausible rationale, but a thorough

14   investigation of the genetic background and importance

15   of genes has really not been done very carefully,

16   especially for NO2, and again these are likely

17   populations and these are likely susceptibility

18   factors.

19	The literature doesn't bear out in terms

20   of NO2, but for other pollutants, I think most of these

21   likely sort of susceptible populations could be

22   included and I think that's the end of my comments for

23   now.

24	DR. HENDERSON:    Thanks, Steve.  Ed?

25	DR. POSTLETHWAIT:    Yes, I would just



156



1  like to reiterate that, as most of the reviewers have

2  said, that the...

3	DR. HENDERSON:    You need to get close to

4  the mike just for the telephone people.

5	DR. POSTLETHWAIT:    I'll yell.  I would

6  like -- I would just like to reiterate that I think

7  that this revision has shown commendable improvement.

8	Adding to Steve's list of potential

9  susceptible populations, again I like your idea of a

10   table.  Any pro-inflammatory state -- diabetes could be

11   included -- and one of the things that was brought up

12   earlier in discussions about exposure and dosimetry, et

13   cetera, that the chapter did not address was both

14   proximity to high concentrations, roadways, et cetera,

15   and it also extends into this issue of how high is --

16   James has brought up how high can spikes be and people

17   live in areas that may experience spikes of NO2.

18	It still, because of a dilutional factor

19   from the long sampling time, meets the current NAQS.

20   Those perhaps should also be considered to be

21   susceptible populations.  In fact, I think we discussed

22   that a bit at the last meeting.

23	The other point I would like to bring up

24   is -- you know, again I raised this at the last meeting

25   and I don't think it's appeared in this document -- is

1  at the scientific literature, ultimately the idea of

2  form and foundation for setting a standard, I think

3  that does need to be considered.

4	The only other thing I think I would

5  like to add is I agree with George's comment earlier

6  about the business of considering particulates, but

7  that again becomes an extraordinarily complex issue

8  unless you're dealing with chemically inert particles

9  that simply act as a vector to deliver NO2 to the deep

10   lung.

11	If you have reactive particles, now

12   you've got both potentially NO2 chemistry going on on

13   the surface of the lung and also NO2 reacting with

14   goodies in the particles that could produce a whole new

15   class of compounds that is not even considered and so

16   again all that does is add increased challenge to what

17   we all recognize as a very challenging situation and I

18   think with that I'll end.

19	DR. HENDERSON:    Thank you, Ed.  For my

20   own education, Ed, have people actually measured

21   endogenously formed NO going to NO2?

22	It makes sense what you said, but I just

23   didn't know if anyone had actually measured.

24	DR. POSTLETHWAIT:    That's almost an

25   impossible measurement to make.



155

157



1  the whole issue of the endogenous generation of

2  reactive nitrogen species and I'm the first to admit

3  I'm not sure how that factors in, but conditions like

4  asthma, for example, anytime you have inflammation --

5  in fact, exhaled NO is now used as a common marker for

6  inflammation -- through its oxidation and nitrate and

7  the activity peroxidases, you're going to make NO2.

8	And so how much endogenous NO2 is

9  actually produced relative to the exposures from

10   environmental sources is I think an extraordinarily

11   complex thing to try to dissect out, but I think it

12   warrants mentioning because in reality what you have,

13   for example in an asthmatic, is somebody who is already

14   endogenously generating NO2, driving protein nitration,

15   and now you're adding a little more NO2 on top of that

16   and the reactivity may come from the increase in

17   oxidant stress, which certainly will set off twitchy

18   airways or it could come from location of inhaled NO2

19   reacting with goodies relative to where NO2 is

20   generated on the surface of the lung from either

21   neutrophils or eosinophils and I think it's an issue

22   that needs to go into the document.

23	How you handle it in terms of dosimetry,

24   biological plausibility, et cetera, does become a bit

25   of a sticky wicket, but I -- again, for trying to look

1	DR. HENDERSON:    Yes, that's what I

2  thought.

3	DR. POSTLETHWAIT:    But if you

4  up-regulate nitric oxide synthase, which you find in

5  virtually every inflammatory situation...

6	DR. HENDERSON:    Yeah.

7	DR. POSTLETHWAIT:    ...the NO, through a

8  variety of mechanisms, reacts to form nitrate.  It

9  actually can directly form NO2 within hydrophobic

10   regions, but by and large it ends up as nitrate.

11	Peroxidase enzymes and the presence of

12   hydrogen peroxide as the other substrate make NO2.

13   They -- it directly will nitrate proteins and NO2 is

14   the nitrating agent.

15	And so if you do a bronchoalveolar

16   lavage or even induce sputum from an asthmatic, you

17   find nitrated proteins.  I mean that chemistry is going

18   on.  There's no question about that, but in terms of

19   how much NO2 is being made from an endogenous source

20   versus how much you're inhaling, please, somebody do

21   those calculations because I would love to see them.

22	DR. HENDERSON:    No, I was trying to

23   mentally do something like that.  Okay, thanks a lot.

24	Now, do other people on the panel have

25   comments on -- it's really Chapter 4 and Charge



160



1  Question 4 related to -- yes, John?

2	DR. GORDON:    I guess my comment actually

3  in just thinking and looking at the charge question

4  again is whether we've talked enough about the last

5  little bit, which is the potential public health

6  significance of NO2 effects and we've focused on the

7  susceptible populations and I think there's two pieces

8  of that that need to be discussed.

9	One is the interpretation of the effects

10   observed and if emphasis is going to be given to the

11   allergen challenge studies, what is the public health

12   relevance of those -- of those effects, which I think

13   probably merit some discussion and consideration.  This

14   is one of the comments I brought up around the risk

15   assessment.

16	The ATS guidelines aren't going to give

17   any sharp guidance, I don't think, on how to interpret

18   this effect in terms of sort of adversity.  So I think

19   that is going to need some expansion.

20	The other is the potential public health

21   significance.  So I guess I would take that as to the

22   -- some indication of the burden of impact at the

23   population level and there's bits and pieces of stuff

24   there that are sort of like lots of people are exposed

25   to NO2 and other things, but I don't think this is

1  become much larger.

2	DR. HENDERSON:    Okay.  Are there others?

3  Well, we'll go to Leanne and then George.

4	DR. SHEPPARD:	Yes.  I wanted to affirm

5  or embrace the idea that Steve brought up of listing

6  all the possible susceptible populations, likely

7  susceptible, that we can think of and then identifying

8  which ones there's information on and which ones more

9  research is needed and I actually had that similar

10   comment in Chapter 5 when we look at the effects.

11	I think the same thing should really be

12   done there.  There's a number of different areas where

13   no research has been done, so we just don't know.

14	We should try to move -- in this

15   document I think we should try to move away from

16   looking under the lamp post and at least identify what

17   we think could be important and then, if there's no

18   information, there's no information, but at least it

19   also lays out potential research areas for the future

20   by being very clear about that.

21	DR. HENDERSON:    Thank you, Leanne.

22   George?

23	DR. THURSTON:    Yes.  I just wanted to

24   follow up on what Ed said.  I agree with him that

25   exposure can be a source of vulnerability and that



159

161



1  really brought together and the only place this is

2  treated is in the summary, so I think that last little

3  bit of the charge needs probably some development.

4	DR. HENDERSON:    Okay, good.  Yes, Ed had

5  something to add.

6	DR. AVOL:    Yes, this is following up on

7  both what Steve Kleeberger said and John and this is

8  the issue of public health significance and

9  susceptibility and vulnerability and this comes out

10   more and is developed more in the risk assessment

11   document, but here I think it seems like it's a little

12   confusing about what the issues are in the public

13   health context if we're not going to get into the

14   discussion of who's vulnerable and who's susceptible

15   and I think it's a little gray about how those terms

16   are used here.

17	We certainly should think about

18   vulnerability in terms of exposure as John alludes to

19   because, if you think about that and think about the

20   size of the population that is exposed in addition to

21   the number of different populations that Steve alluded

22   to in terms of listing out other things than just

23   children, or elderly, or infirm, I think then the

24   public health context becomes more important because

25   the size of the populations that we're talking about

1  people living next to highways in particular are what I

2  would see as more vulnerable.

3	If you follow, you know, the logic that

4  I've been putting forth that the co-presence of

5  ultra-fine particles with their large surface area

6  would be optimal vectors for carrying gases into the

7  lung, then people living within 100 or 200 meters of a

8  highway are exposed to very high levels of ultra-fine

9  particles before they have time to coalesce into larger

10   particles.

11	So that I would see, from an exposure

12   point of view, as not susceptible because those people

13   are pretty much the same as everybody else, but if I'm

14   understanding these terms right, that people living

15   near highways would be a vulnerable population because

16   of the co-presence of ultra-fine particles to be a

17   vector for the NO2.

18	DR. HENDERSON:    Okay.   Terry?

19	DR. GORDON:    Just on that last -- that

20   last issue, would people living near the roads also be

21   susceptible because they might be a different

22   socioeconomic level?

23	DR. THURSTON:    Yes, that would be

24   another issue, if they are different, if there is a

25   different socioeconomic status that we've seen in other



164



1  studies to be important in terms of health effects,

2  which I don't know if that's been shown for NO2.

3	I know we've done it for ozone in New

4  York City where we showed what appeared to be a racial

5  difference in susceptibility to ozone, you know, and

6  responsiveness to ozone.

7	We got high relative risks in the --

8  well, no significant relative risk in the white

9  non-Hispanic, but we saw it in everybody else, but then

10   when we looked at the white non-Hispanics who were on

11   -- who were poor, we saw the same effects as in the

12   minority populations.

13	So it was a socio   what looked like a

14   racial, maybe biological difference actually turned out

15   to be a socioeconomic difference, that you saw the same

16   effect in the non-Hispanic whites when you looked at

17   the poor subgroup and the working -- basically the poor

18   and the working poor.

19	DR. HENDERSON:    Okay.  Are there any

20   other comments now for -- it sounds like you're getting

21   a longer list of susceptible population.

22	If -- are there any requirements from

23   NCEA for clarification of what was said about Chapter

24   4?

25	DR. ROSS:    No, I think that's been very

1  Well, go right ahead.  I didn't mean to cut you off.

2	DR. SCHLESINGER:    No, I think you cut

3  everybody off.

4	DR. HENDERSON:    Oh, my goodness.

5	DR. SCHLESINGER:    I don't think you

6  could...

7	DR. HENDERSON:    I'm a powerful woman.

8	DR. SCHLESINGER:    ...hear us.  Anyway...

9	SPEAKER:    ...you can.

10	SPEAKER:    Rich, we just don't want to

11   hear you.

12	DR. SCHLESINGER:    I'll tell you

13   something.  This is the first time that I couldn't

14   hardly hear you.  Regarding -- one of the issues I had

15   written down here was in relation to the risk and

16   exposure document, but it just -- it came up and that

17   is I was confused over the way susceptible and

18   vulnerable were used and in the other document the

19   writers -- the EPA seems to use -- both terms are used

20   for specific populations.

21	I've showed you on the elderly

22   vulnerable and susceptible, but when they define the

23   two terms, it seems that the age related differences

24   were defined as susceptible rather than vulnerable.

25	So I really think, based on my confusion



163

165



1  clear.  Ellen is our -- Ellen Koran was the lead author

2  on this.

3	Just to verify, I think we were using

4  susceptibility in the sense of the biological --

5  biological conditions like pre-existing disease and

6  vulnerable meaning something like SES or distance to

7  roads or something that's where they are, not

8  necessarily a biological thing and we'll just try to

9  clarify.  Maybe

10   -- I can see sometimes we weren't always clear about

11   that.

12	DR. HENDERSON:    Okay.  Then we'll go on

13   to this, what I consider a major chapter, Chapter 5,

14   where

15   -- and it's -- the Charge Question 5 is related to that

16   and I -- I think we might have a longer discussion on

17   that.   Do we have Douglas Brown on the phone or not?

18	SPEAKER:    Regine?

19	DR. HENDERSON:    Yes.

20	SPEAKER:    Oh, you can hear us now?

21	SPEAKER:    We were cut off for a while.

22   Somebody has been trying to make a comment here.  Was

23   that you, Rich?

24	DR. SCHLESINGER:    Yes.

25	DR. HENDERSON:    Oh, I'm sorry, Rich.

1  and apparently other people's confusion, there needs to

2  be a better definition.  You can be both vulnerable and

3  susceptible.

4	DR. BALMES:    Yes, I think, Rich, the way

5  I read the chapter, and the EPA staff can correct me if

6  I'm wrong, vulnerable had to do with exposure.

7	DR. SCHLESINGER:    Right.  They...

8	DR. BALMES:    Children can be both

9  vulnerable and susceptible.

10	DR. SCHLESINGER:    It's not clear.

11	DR. BALMES:    But -- yes, but I would

12   think older adults, who probably spend -- you know, may

13   spend more times indoors if they're...

14	DR. SCHLESINGER:    Would be vulnerable.

15	DR. BALMES:    ...debilitated would be

16   susceptible, but not necessarily...

17	DR. SCHLESINGER:    Vulnerable.

18	DR. BALMES:    ...vulnerable.

19	DR. SCHLESINGER:    That's correct.  The

20   vulnerable applies more to the exposure scenario.

21	DR. BALMES:    But if it's not clear to

22   you, that means it's not going to be clear to other

23   people...

24	DR. SCHLESINGER:    That's right.

25	DR. BALMES:    ...as well.



168



1	DR. SCHLESINGER:    It wasn't clear to me

2  even after they defined it in the risk and exposure

3  assessment, so they may want to think about it.  That's

4  it.

5	SPEAKER:    That came up in the written

6  document, too, the written comments.

7	DR. HENDERSON:    Are there any more

8  comments from people on the phone?  We didn't meant to

9  cut you off.  Okay.  Do you know if Douglas Brown is on

10   the phone?

11	SPEAKER:    We're having a hard time

12   hearing you.

13	DR. HENDERSON:    Is Douglas

14   Crawford-Brown on the phone, please?  Yes.

15	DR. CRAWFORD-BROWN:    Oh, yes.

16	DR. HENDERSON:    Yes, you're just on.

17	DR. CRAWFORD-BROWN:    Yes, I've been

18   here.

19	DR. HENDERSON:    I thought I heard you.

20	Well, Doug, could you lead us off in our

21   discussion of the Charge Question 5?  And it's actually

22   Chapter 5, too.

23	DR. CRAWFORD-BROWN:    Yes, and that was

24   very convenient.  Thanks very much.  Well, as I say in

25   my summary, I'm really torn about Chapter 5 and the

1  certain questions such as what would be a

2  representative exposure in the United States, how would

3  that representative exposure be different in various

4  vulnerable subpopulations, I mean quantitatively

5  different in there, what should be the averaging time

6  for the exposure to compare it against the relevant

7  epidemiological and clinical studies, and a variety of

8  other issues that I raise in my review.

9	So, it's not possible in either Chapter

10   5 or in any technical supporting documents to know how

11   you would conduct an actual exposure assessment and so

12   that brings me then to my major issue of what the ISA

13   is intended to do.

14	As I say, it does do a good job of

15   letting people know what kinds of information are out

16   there, of letting people know what the strengths and

17   limitations are of the current exposure information,

18   and of letting them know what the relevant health

19   effects are and what the strengths and limitations

20   conclusions about the different health effects are.

21	It's just that there isn't anywhere near

22   the level of information that's needed to let people

23   understand how to take the available exposure

24   information and convert it into exposure estimates for

25   actual populations or subpopulations in the United



167

169



1  reason I'm torn is that I'm not yet clear, and I don't

2  think any of us are, as to how much detail the

3  scientific part, the ISA, should go into on exposure

4  and risk-related issues.

5	So what I say in my summary and what I

6  would say now is that Chapter 5 does a good job of

7  picking out the conclusions that I think are relevant

8  in the earlier chapters for anyone who is eventually

9  going to go in and do an exposure or a risk assessment.

10	I think it does a good job both of

11   saying what was said in the previous chapters and

12   summarizing it, but also picking those points out of

13   the previous chapters that really are relevant to the

14   question of exposure and risk.

15	So on the one hand I think it does a

16   good job of pointing the exposure assessment team,

17   pointing the risk assessment team, towards the relevant

18   conclusions in the earlier chapters.

19	Having said that, and then this may be a

20   problem of the way in which the question is worded, but

21   having said that, you can't go into Chapter 5 and

22   actually figure out how you're going to do an exposure

23   assessment based on the information that's provided

24   either there in Chapter 5 or in earlier chapters.

25	And so it's not possible to answer

1  States.

2	Now, if the EPA were to say, well, that

3  latter is not the purpose of the ISA, then I feel very

4  comfortable with the way in which the ISA is currently

5  written.

6	If the intention is to produce a chapter

7  in which an exposure assessor could go in and know what

8  the next step is or how to conduct an actual exposure

9  assessment, then it doesn't get you anywhere near that

10   level.

11	But I'll sort of end it with that very

12   general consideration.  There are a number of specific

13   comments that I make throughout my review, but that

14   seems to me to be by far the most important issue,

15   still a sort of lack of clarity over exactly what it

16   means to support an exposure assessment to support a

17   risk assessment.  So I'll end with that.

18	DR. ROSS:    Dr. Henderson?

19	DR. HENDERSON:    Yes.

20	DR. ROSS:    Can I speak to that?  I would

21   like to actually inform Dr. Crawford-Brown that the

22   second is actually the intent of the Integrated Science

23   Assessment, in that  as we're laying out the evidence

24   that would support risk and exposure assessments and

25   that's the plan, but we have a lot of faith in the



1  competence and skills of our colleagues and OAQPS to

2  actually conduct the risk and exposure assessments, to

3  gather the data, to do the actual assessment itself, so

4  we don't go so far as to lay out the data or to tell

5  them exactly what to do or what to use, but that's the

6  gray areas.

7	We're trying to lay out the evidence

8  and, you know, what effects seem to be most strongly

9  associated with NO2 that can inform their assessment,

10   but not go so far as to...

11	DR. CRAWFORD-BROWN:    I would like to say

12   I can sort of live with that kind of characterization

13   of Chapter 5.  It's just that, as someone who has done

14   exposure assessments before, when I read Chapter 5 and

15   get to the end of it, I know something about the kinds

16   of issues I need to be concerned about, you know, what

17   sort of weaknesses there are in the existing

18   information.

19	It's just that there are certain key

20   questions that I'm going to have to ask myself as an

21   exposure assessor that I don't find addressed in here

22   and I already mentioned one of them just as an example,

23   which is I just don't know what averaging time is being

24   pushed in this chapter and so I don't know do I go and

25   take a running average of daily results for exposure,

1  hope you're -- you know, I know it's later where you

2  are than we are, so I'm glad you're staying with us.

3  Now we go to Jon Samet.

4	DR. SAMET:	Well, I thought that Doug

5  had made some very thoughtful comments both in writing

6  and now orally.

7	I think I'll add to it.  There's  this

8  great line in Raiders of the Lost Ark, number one --

9  number four is out -- where they stop and they go, "Oh,

10   you mean we're just making this up as we go along?" and

11   since we seem to be making this up as we go through it

12   the first time, I think we better be pretty careful

13   about what we make up and, if this is the model, it

14   ought to be explicit and I think, going back and

15   looking at what's here in Chapter 5, I'm not sure the

16   purpose as we just heard it articulated or what it

17   should be is actually said here.

18	I mean if you read it, it's got the

19   title, "Integrative Summary and Conclusions."  It's not

20   too integrative and it does pull out, you know, bits

21   and pieces of what's been learned that are thought to

22   be most relevant from the prior chapters.

23	The framing -- these framing questions

24   on the first page I'm not sure relate explicitly to

25   this purpose of the chapters as we've talked about

172



171

173



1  do I do hourly results, and so on and I think that's

2  important for the exposure assessor and there is

3  information in the document in earlier chapters that

4  would give some insight into what the answer are --

5  answers are to some of the points that I raise here.

6	They just don't quite make it all the

7  way in Chapter 5, but I certainly am willing to hear

8  the idea that, if you give the strengths and the

9  weaknesses and raise the important questions as you

10   have, then the actual exposure assessors in the agency

11   will be able to take that and do what's needed in the

12   exposure assessment.

13	DR. HENDERSON:    Thank you, Doug.  You're

14   going to be with us for a while, right?

15	DR. CRAWFORD-BROWN:    I'm sorry?

16	DR. HENDERSON:    You're going to be on

17   the call for quite a while more, right?

18	DR. CRAWFORD-BROWN:    I've been on here

19   the whole time.  It's just that...

20	DR. HENDERSON:    Are you just...

21	DR. CRAWFORD-BROWN:    ...periodically we

22   -- I think those of us on the phone actually get cut

23   off from -- not cut off from the conversation, we can

24   hear the conversation, but we're not able to respond.

25	DR. HENDERSON:    Oh, okay.  I hope -- I

1  them, particularly as it sets up the foundation for the

2  risk and exposure assessment and presumably the ISA

3  together with the risk and exposure assessment become

4  the basis for making policy -- making policy decisions.

5	In fact, I'm not even sure I know what

6  all the charge questions really -- really mean and I

7  indicated that in my comments.

8	For example, what are the air quality

9  relationships between short and long-term exposures to

10   nitrogen oxide?

11	I just really don't know what that means

12   and don't bother to tell me, Mary.  It's okay.

13	I -- and then -- and then when you turn

14   to what I think is the integrative part, which is this

15   last bit called "conclusions," at least in my view it

16   short of fell short about getting to some of the key

17   questions and I think this goes back to what we

18   discussed around Charge Question 3 and whether the

19   question of sort of plausibility of effects and the

20   existence of effects at the range of concern for the

21   risk assessment has really been addressed in the

22   document and I sort of highlighted this one sentence on

23   Page 522, and if you look at -- this is Line 21 -- I

24   just -- you know, as sort of an integrative sentence,

25   the only thing that's really integrative about it is it



176



1  starts with integrating and then -- and then it really

2  doesn't -- you know, the evidence has just not been

3  brought together to support it and I think this is

4  probably I think one of the most critical

5  determinations with regard to the risk assessment and

6  its interpretation.

7	So maybe the bottom line here, a

8  clearer, up-front statement about what the purpose of

9  the chapter is because I don't think that's clear and

10   again this goes back to -- since we are making this up

11   for the first time through, we should be very clear

12   about what this is about.

13	And then I think -- and I think we're

14   going to have a longer discussion as we sort of go back

15   to Charge Question 3 and 5 in any case about sort of

16   this broader concern about has the document established

17   clearly what is we know and what it is we don't know

18   and the limits of the evidence, which I think is what

19   James has gotten at before.

20	DR. HENDERSON:    Thank you, John.  I had

21   also highlighted that one sentence, which kind of goes

22   against what you said earlier, Mary, where you say

23   we're not going to say, in the ISA, whether, you know,

24   there is evidence that there are health effects at less

25   than the current NAAQS.  I mean I believe you said

1	We are not, though, making implications

2  about what any potential level of a standard might be.

3  We're just saying here's what we see.  That's what I --

4  that's the distinction I meant.

5	DR. HENDERSON:    Okay, that -- I see the

6  distinction.  That's a fine distinction, but it's -- I

7  think that has to be justified up the kazoo to make a

8  statement like that and I'm not sure that it's -- at

9  least I came away with questions, but let's go on to

10   other people's comments.  Frank Speizer?

11	DR. SPEIZER:    Well, I don't have -- I

12   don't have too much more to add except that I was sort

13   of disappointed and maybe -- maybe it's the titling of

14   this.

15	SPEAKER:    Frank...

16	DR. SPEIZER:    Maybe it shouldn't be

17   called...

18	SPEAKER:    ...the microphone.  We can't

19   hear.

20	DR. SPEIZER:    ...summary and conclusions

21   because it doesn't really read like a summary and

22   conclusion.  It reads like more of a catalog of...

23	DR. HENDERSON:    Some people can't hear

24   you, Frank, so I think you're going to have to shout.

25	DR. SPEIZER:    Okay.  I just -- I was



175

177



1  that's for later on, but -- and yet that's a very

2  strong conclusion right here.

3	DR. ROSS:    Well, the distinction I was

4  making there is that we weren't making decisions about

5  the NAAQS, but I think from the epidemiologic

6  evidence...

7	SPEAKER:    Mary, we can't hear you.

8  Could you speak up?

9	DR. ROSS:    Sorry.  What we did was look

10   at the epidemiologic studies and there quite a few of

11   the studies were conducted in areas where the entire

12   air quality distribution is below the current level of

13   annual standard for the NO2 NAAQS.

14	So we felt confident stating there that

15   there are associations in these studies in areas where

16   all the data are below the current standard levels,

17   even though it's an annual standard, and that, to me,

18   seems like a scientific conclusion.	Here's

19   what the studies are.  It's the same as saying what the

20   levels are within the clinical studies.  Here's where

21   the effects are being seen.  Here's something we can

22   say about the range of it and in the bullets about the

23   epidemiologic studies we characterize the sort of

24   distribution of data from the studies in which the

25   effects are seen.

1  commenting I think it may be just in the title of the

2  chapter calling it summary and conclusions that may be

3  inappropriate because I read this much more as sort of

4  a catalog of status, so to speak, and it is providing

5  sort of a listing of things that will be I think useful

6  later on.

7	One of my concerns was in Table 5.34.

8  It is that there's nothing about exposure in there.  I

9  mean there is -- you're basically describing how the

10   studies were laid out and how they were used, but you

11   don't give us any levels of exposure really and I would

12   have thought that should have been added to these

13   tables.

14	Now, it makes the tables even bigger

15   than they are.  You would probably have to put them in

16   landscape format if you did that, but I think probably

17   some range of exposures in each of the studies would

18   have been useful to have in this cataloging of the

19   studies.

20	DR. ROSS:    Can I speak to that?

21	DR. SPEIZER:    Yes.

22	DR. ROSS:    That table is actually a

23   legend.  This is not pretty, but that's the legend for

24   the figure with the...

25	DR. SPEIZER:    Oh, I see.



180



1	DR. ROSS:    ...with the red, the blue,

2  and green.

3	DR. SPEIZER:    I'm sorry.

4	DR. ROSS:    If you go down to Appendix

5  Table 5A, those are...

6	DR. SPEIZER:    Yes, that's...

7	DR. ROSS:    There's a little more detail

8  about the levels...

9	DR. SPEIZER:    ...a little more detail.

10	DR. ROSS:    ...and the studies.

11	DR. SPEIZER:    Right.  That's true.  Yes,

12   you're right.  Okay.  I guess my major concern was that

13   I didn't think this was an integrative summary and

14   conclusions.  I think it's an important piece.  It just

15   may be labeled wrong.

16	DR. HENDERSON:    Thank you, Frank, and

17   Ron?

18	DR. WYZGA:    Okay.  Let me say that

19   overall I thought the document was vastly improved, but

20   I felt that this chapter was improved the least of all

21   of them and I guess there's several questions.

22	First of all, I'll echo what some of my

23   colleagues have said.  I had hoped that somehow the

24   transition from this document to the exposure risk

25   assessment would be more seamless and there is a lot of

1  be very helpful and I think a lot of the information is

2  elsewhere in this document, but it somehow is not tied

3  together and organized in a way to answer this.

4	So my two things, I would ask that you

5  address that explicitly and then secondly, you sort of

6  think about what's the transitional information that's

7  needed for the following document and should it be here

8  and, if so, can you somehow summarize it?

9	DR. HENDERSON:    Thank you, Ron.  Now I

10   would open up for everybody to discuss this Chapter 5.

11   Yes, Ed?

12	DR. POSTLETHWAIT:    Yes, this may be a...

13	SPEAKER:    Ed, please speak into the

14   mike.

15	DR. POSTLETHWAIT:    Yes, I am trying.

16	This may be actually a semantic issue or

17   sentence structure, but on Page 519, there's part of a

18   sentence here that reads, "There's little evidence of

19   an effect threshold," and I really thought that that

20   needed to be clarified and that same concept appears in

21   other places in the document.

22	My assumption is that that means that

23   there isn't scientific evidence to point a finger at a

24   number for a threshold as opposed to NO2 effects are

25   linear down to zero and to me that's a very important



179

181



1  information here, but somehow it's not summarized I

2  think in an effective way in Chapter 5.

3	For example, I would hope that this

4  chapter would provide more guidance for the risk and

5  exposure assessment to draw on in terms of which

6  studies to use.

7	Do you use the clinical studies?  Do you

8  use the epidemiology studies?  Basically what averaging

9  time?  Are you worried about short term?  Is it an

10   hour, 24 hours?  What's the summary of the evidence for

11   both?

12	So I had hoped there would be a lot more

13   information to address these and maybe it's not what

14   this document is supposed to do, but somehow I had sort

15   of -- something is missing in that transition and I'm

16   not sure if it should be here or elsewhere, but I was

17   hoping it would be here.

18	And then also I felt that it needed to

19   really take on more head on -- you know, the basic

20   issue and I think the difficulty with the

21   epidemiological studies is is NO2 per se the agent

22   which is responsible for a lot of the effects that are

23   being seen or is it an index?

24	And I thought if it had, you know, some

25   very specific paragraphs that addressed that, it would

1  distinction that needs to be clarified in here and it

2  actually goes back to multiple of the charge questions.

3	Whoever wrote that, I would love to hear

4  whether -- unlike James who says he's a pretty decent

5  reader, maybe I'm a really bad reader, but...

6	DR. ROSS:    What we meant was there was

7  little evidence of a threshold and especially referring

8  to the epidemiologic studies, which are the ones

9  conducted at the lowest levels, but then that this is a

10   shorthand way of characterizing why it's difficult to

11   distinguish thresholds in epidemiologic studies.

12	I think we discuss it a little bit in

13   more detail in Chapter 4, some of the difficulties in

14   interpreting epidemiologic data with regard to

15   thresholds, but we're not saying that one doesn't

16   exist.  It's just hard to find.

17	DR. COTE:    And I think this is a laundry

18   list of reasons why it's difficult to see a threshold,

19   whether or not it exists or not.  Even if a threshold

20   exists, all these would make it difficult to see.

21	DR. SAMET:    Most of the studies did not

22   seek to find a threshold in their analysis, so I think

23   you have to be really careful.

24	DR. COTE:    So would you -- is it that --

25   I think what we were trying to say is we didn't observe



184



1  why...

2	SPEAKER:    Are you guys having trouble

3  hearing?

4	DR. COTE:    Sorry.

5	SPEAKER:    Yes.

6	SPEAKER:	Isn't the point actually that

7  there's little evidence relevant to assessing whether

8  thresholds exist period and that's...

9	DR. POSTLETHWAIT:    Yes, I mean does --

10   is there an exposure threshold or do we just not have

11   enough information to identify one?

12	DR. COTE:    I think we don't have enough

13   information to identify one.  We don't know whether one

14   exists or not.

15	DR. POSTLETHWAIT:	So my suggestion

16   would be to idiot proof...

17	DR. COTE:    Say that?

18	DR. HENDERSON:    Okay.  Other comments?

19	SPEAKER:    You guys need to speak louder

20   or closer to the mike.  We're really having problems

21   hearing.

22	DR. HENDERSON:    Yes.  Go ahead Ila.

23	DR. COTE:    This is Ila Cote.  I just

24   wanted to say something before there was a general

25   discussion following up on Ron's comment, which might

1  together.

2	And the threshold -- well, obviously the

3  threshold varies from outcome to outcome and I think

4  someone brought that up earlier that, you know -- so

5  that defining a threshold is going to be outcome

6  dependent and of course difficult to do.

7	And then lastly, the thing that I

8  brought up on my written comments, which you really

9  haven't discussed and people may not agree with me, but

10   these framing questions I thought were very good at the

11   start of Chapter 5.1, and I wonder if that isn't the --

12   shall I call it the framework that we should do the

13   integration under at the end and then go back at the

14   end and say, okay, how can we or can we not and to what

15   extent can we answer these framing questions?   Does

16   anybody -- how do people feel about that?

17	DR. HENDERSON:    I think that's a good --

18   I mean this is just me speaking personally.  I think

19   it's a -- I think it's a good idea and again it's the

20   idea that Ellis Cowling  suggested.

21	She said, "Why don't we just answer

22   those questions?  We put them up there and we need an

23   answer."

24	And I'm not sure that they'll be

25   answered all in the ISA.  It may be that some of these



183

185



1  help, is that the tables and figures in Chapter 5,

2  maybe we weren't as explicit as

3  -- we clearly weren't as explicit as we should have

4  been, but those are our attempt to identify what we

5  think are the best studies that one could utilize for

6  exposure and risk assessment.

7	You know, so it's kind of whittling down

8  of this is our best bet and I guess we weren't as clear

9  about that as perhaps we needed to be, but when you

10   follow through on the discussion, at least that was our

11   intent there.

12	DR. HENDERSON:    Okay.  Did -- George,

13   were you trying -- or -- yes, go ahead, George.

14	DR. THURSTON:    Well, just to respond

15   quickly to some...

16	SPEAKER:    You have to speak up.

17	DR. THURSTON:    ...to some of the

18   comments that were made, one thing with the NO2 PM

19   issue that Ron brought up, I think it's important to

20   remember that this is not an either/or kind of thing.

21   This might be an and/or because of the issues that I

22   brought up.

23	So it's not like, if it's PM, then it's

24   not NO2 and if it's NO2, it's not PM.  The reality is

25   -- or let's say I believe the reality is that they work

1  answers come in later documents and so, Mary, how do

2  you feel about that?

3	DR. ROSS:    I was just going to say

4  exactly what you said.  The review doesn't stop with

5  the ISA.

6	DR. HENDERSON:    Yes.

7	DR. ROSS:    So I think we will definitely

8  try to be more specific and pull together the

9  information that we may not have organized as well as

10   we could have to try to address those questions as much

11   as the science does, but the policy assessment is going

12   to build upon that further, so I think it's fair that

13   we'll go part of the way, but not try to answer them,

14   you know, this is the end of it all, the discussion.

15	DR. HENDERSON:    Yes, that sounds fair.

16	DR. COTE:    Just as background, we

17   actually wrote it both ways and found that, in

18   answering the questions very explicitly, there were

19   some assumptions that needed to be made or some policy

20   decisions, science policy decisions that needed to be

21   made, that we thought were more appropriately made in

22   the exposure and risk assessment and so, as a

23   consequence, dropped the explicit answers of the

24   question and went to a more -- presenting information

25   that could be used to answer those questions.



188



1	DR. HENDERSON:    Okay, Leanne?

2	DR. SHEPPARD:    I'm interested in

3  following up on that particularly since you already

4  wrote it.  I wonder if that's worth putting in the

5  annex and being very clear that, if you make these

6  policy decisions, that this follows without actually

7  making a recommendation.

8	DR. COTE:    I think we don't want to do

9  that because we just got uncomfortable with what our

10   role was in the process and answering those questions

11   and again we thought it was better put off to the

12   exposure and risk assessment in which those questions

13   will be addressed I think.

14	DR. HENDERSON:    There's some -- there's

15   a question in my mind.  This is about the whole thing

16   and I'll throw it out there and you can tell me how

17   wrong I am.

18	In looking over all this Integrated

19   Science Assessment, I see that the lowest exposure

20   information that we have is in the epi studies, but

21   they are -- they have, of course, multi-pollutants and

22   that's a big problem, so that information has its

23   problems.

24	We have the tox data, which is done at

25   much higher levels.  It provides mechanistic

1  studies are a very strong body of evidence.

2  Combined with the clinical study data and the

3  toxicological study data, we think, you know, it's a

4  pretty strong case for NO2 having respiratory effects.

5	DR. HENDERSON:    I'll let Ed go and then

6  George.

7	DR. AVOL:    Then isn't that -- the

8  Integrated Science Assessment, isn't that what you

9  should say in the document?

10	JOHN:    Speak up, Ed.

11	DR. ROSS:    I think we were trying to say

12   that.

13	DR. AVOL:    I said -- John, I said isn't

14   that -- the Integrated Science Assessment, isn't that

15   what should be said in the document?

16	DR. ROSS:    That's what we meant...

17	JOHN:    I think so.

18	DR. ROSS:    ...to say, so we'll --

19   we'll...

20	DR. HENDERSON:    Well, yes, I mean what

21   you just said is good.  It makes me understand it.

22	SPEAKER:    Write that down.

23	SPEAKER:    Yes, I was going to say...

24	DR. HENDERSON:    I am one of those

25   not-so-careful readers, but that was -- I was giving



187

189



1  information that suggests that some of the things

2  you're seeing in the epi studies -- you know, it's

3  plausible mechanism consistent with some of the animal

4  data, but then I saw that the emphasis is going to be

5  on human clinical studies because it's done in humans

6  and it's, you know, controlled exposures, but of course

7  you have the problem of healthy people being exposed

8  for short periods of time.

9	But is that, in general, what I was

10   supposed to draw from this that, if I were to go on and

11   do a risk assessment, the best data to use was human

12   clinical?

13	DR. ROSS:    No.  Actually, let me tell

14   you I think when the staff -- and we sat down and

15   discussed this, I think we believed the epidemiologic

16   data, especially for respiratory effects, and we

17   believe that there are effects occurring in those

18   levels and the levels are quite low.  What's difficult

19   is interpreting the quantitative magnitude of the

20   effect considering that there's part of an ambient

21   mixture.

22	So I'm not -- I don't think we're saying

23   that you should only use clinical study data.  I think

24   it's easier to determine causality from clinical

25   studies, but as part of the mixture, the epidemiologic

1  you the actual impression I got, so that's good.  Okay,

2  George?

3	DR. THURSTON:    Well, I was just going to

4  differ with you on the statement you made that in the

5  epidemiology we have the multi-pollutants and that's a

6  problem and I just want to point out that that's a

7  strength, that we're talking about real people, real

8  situations, real exposures, and real exposures include

9  multi-pollutants and that's where we get the most

10   useful information about the questions we have to

11   answer.  So there are challenges, but the

12   multi-pollutant aspect is a strength I think, not a

13   problem.

14	DR. HENDERSON:    It's a very strong

15   strength.  I couldn't agree with you more, George.

16   It's just that we don't know how to handle it well I

17   don't think.  I mean I think it's a real tough

18   question.

19	DR. BALMES:    Can I follow up on that,

20   Regine?

21	DR. HENDERSON:    Yes, please do.

22	DR. BALMES:    So going back to the point

23   that I made earlier today when we were talking about a

24   different charge question, I'm still bothered by the

25   assessment in Chapter 5 about the effects of long-term



192



1  exposure to NO2 and respiratory morbidity.

2	Again the overall judgment is that the

3  data are sufficient to infer -- are suggestive, but not

4  sufficient, to infer a causal relationship at this time

5  and I don't necessarily disagree with that global

6  judgment, but the bullet points --

7	I guess there are four of them that

8  follow that statement -- refer, one, to the lung

9  function data, two, to asthma prevalence and incidence

10   and then, three, to respiratory symptoms and then

11   there's a fourth on animal tox and I think that

12   throwing the -- what I think is pretty strong evidence

13   epidemiologically of chronic exposure to NO2 and

14   effects on growth of lung function in with the asthma

15   data, which I think are not very particularly

16   convincing with regard to asthma prevalence and

17   incidence and long term exposure and the respiratory

18   symptom data with long-term exposure, which also aren't

19   very convincing or are inconsistent, it really, as I

20   said in my written comments, dilutes what I think is

21   reasonably strong evidence about the effect of

22   long-term exposure on lung function.

23	And basically both earlier in Chapter 3

24   and here the statement is made, well, since there's

25   high correlation among traffic-related pollutants in

1	SPEAKER:    Jon.

2	DR. HENDERSON:    Jon?

3	DR. SAMET:    Go ahead.

4	DR. HENDERSON:    I'm so sorry.

5	DR. NUGENT:    I think -- I think I have

6  to agree with Terry that we can't go down this route.

7  We cannot say that we can't do this because it's too

8  hard.

9	I mean I think we're going to have to

10   make some judgments as to what part of what we see as

11   an effect we can attribute to NO2...

12	DR. HENDERSON:    NO2.

13	DR. NUGENT:    ...and we're going to have

14   to -- we're going to have to bite the bullet and admit

15   that there is some -- there is some uncertainty in

16   this, but we have to come up with something.  We can't

17   just say it's too complicated.

18	DR. HENDERSON:    Too hard.  Good point.

19   Jon?

20	DR. SAMET:	Just a somewhat contrary --

21   and I think part of the reason that Figure 16-2 on Page

22   113 is here is because of these potential relationships

23   that NO2 could have with effects and this gets to the

24   heart of interpreting these multi-pollutant models and

25   I think Ed should speak up and Leanne probably, too,



191

193



1  the children's health study and in the Mexico City

2  study, we just -- you know, we can't use it and that's

3  too easy of an answer and I don't think it addresses

4  George's point.

5	I just think that we need to have a

6  little bit more thought here rather than just

7  dismissing these data because they're so called

8  confounded by copollutants.

9	DR. HENDERSON:    Thank you and Terry has

10   something to add.

11	DR. GORDON:    Yes, I want to agree with

12   the last two comments.  It seems like -- I'm not an

13   epidemiologist, but it seems like a slippery slope and

14   it might be setting a precedent to say it's too hard

15   for us to separate out NO2s and the copollutants and

16   what happened with PM?

17	PM 10, PM 2.5, one is sure part of the

18   other and this committee or administrator separated

19   them out somehow to have two separate standards for

20   NAAQS, so I know it's not the same and it's a weird

21   correlation, but we did it for PM 2.5 and PM 10, so I

22   think it's worth it.

23	DR. HENDERSON:    Okay, Frank?

24	DR. SPEIZER:    I wasn't next really.

25	DR. HENDERSON:    Who was next?

1  but the children's health study investigators have been

2  very cautious about interpreting their findings and

3  pinning them on a specific pollutant and they have

4  avoided that because they understand that, in terms of

5  the pollutant index, that they are correlated to a high

6  degree and for several it doesn't exactly matter which

7  one you pick.  The association is about the same and I

8  think they've been very judicious in their

9  interpretation of those papers and I think Ed should

10   speak to that.

11	I think the other issue is -- and I've

12   had this discussion now in multiple venues the extent

13   to which the multi-variable regression of tools that we

14   use can sort out these relationships, particularly in

15   terms of the challenge that we have and the data.

16	Again, so this is where I fall back then

17   to the plausibility of the effects observed and they

18   have to be interpreted I think in a broader context.

19	So I would say that it would be bad for

20   us to say, gee, you know, five variables were in the

21   model and NO2 was significant, therefore there's an

22   independent effect.

23	I think we have to be very cautious

24   given these relationships among the variables and how

25   far we're going to delve into this as a committee, as



196



1  opposed to saying that that's a charge back to EPA I

2  think is another matter, another very difficult aspects

3  of this pollutant.

4	DR. HENDERSON:    I see us needing to

5  advise the EPA on what we think they should do, not

6  that we should do it.  I mean, that -- that...

7	DR. SAMET:    I think that was -- that was

8  my last -- the last point.  You know, there's a tough

9  problem here, but they're supposed to solve it.  Good

10   luck.

11	DR. SPEIZER:    But we really need to help

12   them.

13	DR. SAMET:    Agree, agree.

14	DR. SPEIZER:    And use the best judgment

15   that we can to get there.

16	DR. HENDERSON:    Yes, and that's why I'm

17   glad we're having this discussion because I see the

18   multi-pollutant problem is at the heart of -- you know,

19   of what we're talking about here and so I'm very

20   pleased that everybody is joining in.  George?

21	DR. BALMES:    This is John Balmes.

22	DR. HENDERSON:    Oh, John wants to talk.

23   Go ahead, John.

24	DR. BALMES:    I don't disagree with Jon

25   Samet's characterization of the children's health study

1  somehow grasp that and address that issue.

2	We have tried -- as Jon Samet pointed

3  out, we have tried time and time again.  We have been

4  very careful to look and try and tease out the

5  individual effects because we realize that's the way

6  the current NAAQS are listed, but the fact of the

7  matter is that there is very high correlation of these

8  constituents in the air because they come from a common

9  source, combustion exhaust, and they behave, as George

10   talked about and several other people talked about this

11   morning, in common ways through the air.

12	They may, in fact, interrelate in the

13   sense of -- in the context of providing a vehicle for

14   delivery, an approved source delivery, and that's then

15   sort of an exacerbation rather than the individual

16   components and so in the public health significance

17   sense, it is important and relevant that these things

18   do act together and somehow we need to sort of address

19   that here.

20	In our analyses, in our publications, in

21   our presentations we have tried to separate out, and

22   negate, and minimize, and diffuse the issue of them all

23   being a package and the fact of the matter is, after

24   all those analyses, all those adjustments, all those

25   models, NO2 continues to stay in there as an important



195

197



1  data, but I think that there has to be an

2  acknowledgment of the multi-pollutant issue and I think

3  we do -- it is our responsibility to advise the agency

4  that -- or at least I believe that there are effects

5  from the combustion pollution mix that's been studied

6  by the children's health study and to say that, oh,

7  well, we don't find an independent effect of NO2 or we

8  don't find an independent effect of PM 2.5, therefore

9  we can't use these data is problematic.

10	We have to figure out a way to use these

11   data because I think these data very strongly suggest

12   that exposure, chronic exposure, to combustion source

13   pollution is bad for your lung function.

14	DR. HENDERSON:    Ed?  Did you have

15   anything to say about the children's study?  I mean...

16	DR. AVOL:    This Ed, not that Ed.

17	DR. HENDERSON:    Oh, I'm so sorry Ed.

18	DR. POSTLETHWAIT:    We look a lot alike.

19	DR. HENDERSON:    No, I didn't know which

20   Ed John was referring to.  Ed Avol?

21	DR. AVOL:    So with regard to the

22   children's health study, I mean I think the issue for

23   the agency is -- and this may be simplistic, but the

24   reality is, in the public health context, you are

25   exposed to multiple pollutants and we just need to

1  player in the effects.

2	DR. HENDERSON:    Yes, go ahead, James.

3	DR. CRAPO:    I agree with the concepts

4  that are being talked about completely, but I'm having

5  real difficulty in understanding the document and what

6  its ultimate arguments are settling down to be.

7	I know that your next step after this is

8  to go into the risk assessment document and you had to

9  pick a range and you picked 200 to 300 parts per

10   billion as a range that you're going to test for

11   susceptible population being -- isn't that right?

12	SPEAKER:    They did.

13	DR. CRAPO:    What?  They did?  Okay, they

14   did.  No, but the basis for that needs to come out of

15   the risk assessment, this -- the ISA document and the

16   logic that would support that needs to be clearly

17   elucidated in this document and defended.

18	And as I said earlier, if I use the

19   human exposure data, the clinical human experimental

20   exposure data, I have real difficulty justifying that

21   range.  It would have to be a higher range.

22	We're arguing now and I think well that

23   the epidemiologic data argues for a lower level, that

24   there's effects that need to be explored, but it's not

25   really in the document to provide that support and I



200



1  don't see how I'm going to get to that range.

2	I would really like to know what the --

3  how the agency staff would present their argument that

4  there is a range of actual exposures that occur that we

5  need to test and what the arguments are that would

6  support your position on that.

7	DR. ROSS:    I think -- you're going to

8  tell me to speak up.  I think there's two things.  One

9  is, as you're aware, there often are two different ways

10   that the agency evaluates the information.

11	It's been classically known --

12   classically.  Most recently there's a new tradition of

13   evidence based and risk based analyses of data and

14   there's a way of looking at the evidence in a

15   qualitative sense and I think you can look at the

16   epidemiologic evidence in a qualitative sense.  What

17   OAQPS is doing is quantifying some portion of the data.

18	So in the risk and exposure assessment

19   they're looking at what they can quantify and doing

20   some quantitative analysis and I'm -- I'm speaking out

21   of my realm of expertise here and they're going to

22   speak to this very shortly after the break.

23	But -- so you can look at the

24   epidemiologic evidence in a qualitative sense.  I would

25   say semi-quantitatively, let me say, of saying, yes, we

1  .053.

2	DR. CRAPO:    So does that mean that we

3  should be -- I mean the OAQPS should be looking at

4  possible risk assessments that go all the way down to

5  the current standard instead of stopping -- going just

6  to 200 PPB?

7	DR. ROSS:    See, we're not saying that.

8  We're just saying here's what the evidence is and it's

9  not actually required that you quantify risk estimates

10   for everything and so OAQPS is looking at what they can

11   quantify.

12	Again here I'm speaking outside my turf,

13   but there's a difference between what you can actually

14   quantify, what you can get data for and what you want

15   to quantify in risk.

16	There is -- it's certainly possible to

17   look at the evidence qualitatively as the EPA has done

18   in the past and say here's what the epidemiologic study

19   states, here are the ranges of levels shown in those

20   studies and use that to draw inferences about the --

21   whether the standard is requisite for protecting public

22   health and...

23	DR. CRAPO:    Okay, so...

24	DR. ROSS:    ...what potential levels

25   might be considered.



199

201



1  believe there are associations and here are the range

2  of levels in those studies in which effects are found

3  knowing that there's no dividing point where there's --

4  there's never a point in an epidemiologic study where

5  effects begin and end.

6	It's a -- it's risk, sort of slope of

7  effects per part per billion, but we can describe the

8  range of those and that's the intent of that Appendix

9  Table 5A to say here are the levels of the

10   epidemiologic studies and then we had some specific

11   ranges listed in the bullet points in Chapter 5, so

12   that's intended to inform the policy decision.

13	DR. CRAPO:    Didn't I hear you say that

14   you found that some of the epidemiological data

15   supported effects where no -- none of the population

16   had exposures exceeding the current annual standard?

17	DR. ROSS:    Right.

18	DR. CRAPO:    Does that mean there were no

19   hourly excursions above the annual standard?

20	DR. ROSS:    Well, the annual standard is

21   -- is...

22	DR. CRAPO:    .053.

23	DR. ROSS:    ...53, so then what we did is

24   look at mean levels in the studies and you can -- but

25   you can even say 24-hour levels sometimes were below

1	DR. CRAPO:    So let me focus primarily on

2  the ISA on my last set of questions.  As I read the

3  ISA, its most potent argument that I -- that came out

4  to me really for there being some risk to the

5  population of adverse health effects from nitric oxides

6  dealt with the human clinical experimental data, not

7  the epidemiologic data.

8	It seemed like the epidemiologic data

9  was dismissed and that the focus was on the asthma

10   patients with -- and is that the way -- I'm questioning

11   whether the ISA should be written with that focus given

12   the kind of discussion we're having.

13	DR. ROSS:    That wasn't the focus that we

14   believed we -- I mean a lot of the emphasis here was on

15   the epidemiologic data and the conclusions draw heavily

16   from the epidemiologic data.

17	We present all of the evidence and then

18   -- but then we try to integrate and some have commented

19   we didn't do that as well as possible, so we'll try to

20   improve on that.

21	We'll try to integrate from

22   epidemiologic, clinical, and toxicological findings to

23   say what do we know about respiratory effects of NO2 or

24   the other effects?

25	DR. HENDERSON:    I think you do need to



204



1  improve that because I got the impression that James

2  did.  I mean not that you threw out the epi data, but

3  that you couldn't use it in the quantitative fashion

4  and you're saying you're using it in a qualitative

5  fashion.  Is that what you're saying?

6	DR. ROSS:    We actually are trying not to

7  say which way you could use it or not, but we're just

8  trying to draw conclusions about what we think the

9  evidence says and I think we say quite -- I mean we

10   certainly meant to say quite strongly that we believe

11   the epidemiologic evidence is supportive of a likely

12   causal association between respiratory effects and NO2

13   exposure, especially short term...

14	DR. HENDERSON:    Okay.

15	DR. ROSS:    ...NO2 exposure.

16	DR. HENDERSON:    That's the short term,

17   okay.

18	DR. ROSS:    Yes.

19	DR. HENDERSON:   I'll accept that.  Dale?

20	DR. HATTIS:    Yes, just following right

21   along on that, this discussion, if you believe the

22   epidemiological evidence indicates the effects of NO2

23   more or less in the range of ambient level of

24   distributions, okay?

25	You still have to ask -- answer, I

1  you know, that it's either the high level excursions

2  that matter or it's the long term average that matters

3  or something in between.

4	And unfortunately, the epidemiological

5  evidence hasn't been analyzed to -- as far as I can

6  tell to shed light on that question because you could

7  have analyzed the epidemiological evidence in such a

8  way that you could have made your dosemetric the

9  incidence of particular high exposures rather than the

10   long term average.

11	DR. HENDERSON:    Okay, are there -- who

12   else is...

13	DR. HATTIS:    He may want to respond

14   because I just made a nasty comment about the

15   epidemiological analysis.

16	DR. AVOL:    If you had health outcomes on

17   the same time line as you had as the excursions.

18	DR. HATTIS:    Well, you could have long

19   term health outcomes that are attributable to

20   occasional excursions or you could have long term

21   health outcomes that are attributable to a long term

22   average concentration.  There's no necessary

23   relationship between the time scale for the measurement

24   of the outcomes.

25	DR. AVOL:    Parenthetically, if, in fact,



203

205



1  think, Dr. Crapo's question.

2	Are there effects at ambient levels

3  because there are these occasional excursions to a very

4  high level or is it because there is, in fact, a long

5  term -- a long acting cumulative impairment of defenses

6  against respiratory infections that causes these

7  impairments to the growth of lung function in children

8  or things of that sort?

9	And if you think that you can't answer

10   that, then maybe the advice you have to give is -- to

11   the OAQPS people and to the public general is that

12   maybe it's safer to do risk assessments on the basis of

13   -- on both kind of bases, either long term average

14   levels or on the incidence of particular excursions to

15   very high levels because we don't know what the causal

16   factor really that's important is, okay?

17	See they -- the OAQPS people, in their

18   preliminary document that we're going to review in a

19   little while, have made the inference that, if the

20   effects -- epidemiological findings are right, then

21   they're due to high excursions.

22	Well, if you don't think that you are

23   confident of that, then it seems to me you should be

24   advising them that they should do it more than the way

25   they're doing it because it could be right that it's --

1  you're following along the line of the excursions as

2  being the important trigger in terms of either short or

3  long term health effects, that identifies a potential

4  public health population of exposure, which gets back

5  to the people alongside roadways and proximity and

6  identifies another group in the area of identifying

7  significant populations exposed.

8	SPEAKER:    Right, the susceptible

9  populations.

10	SPEAKER:    Yes.

11	SPEAKER:    Yes, exactly.

12	DR. CRAPO:    I would like to add one

13   thing.

14	DR. HENDERSON:    Sure, James.

15	DR. CRAPO:    It's not only just a couple

16   of us like Regine and I that interpreted the ISA as

17   saying that it focused on human clinical studies and

18   plus, if you look at the -- at the risk assessment

19   document on Page 12, middle paragraph, it says exactly

20   what I've been saying, that it cites the ISA as their

21   source, so I mean OAQPS interpreted it the same way I

22   did.

23	DR. HENDERSON:    There you go.

24	DR. CRAPO:    No, I think -- I think that

25   the -- I think that the ISA does, in this tenor, go



208



1  this direction and I'm really challenging it to change

2  it to be more critical of the -- of the human clinical

3  studies and the asthmatic responses and show the

4  weaknesses of that as well as the strengths much more

5  clearly and then more appropriately emphasize the

6  epidemiology because otherwise we don't have a

7  scientific basis for going onto the risk assessment.

8	DR. ROSS:    Well, clearly we can work to

9  clarify that, but all I can say is the bulk of Chapter

10   5 is really discussing epidemiologic evidence and

11   epidemiologic findings with support from clinical and

12   toxicological studies if you walk through it line by

13   line.

14	DR. CRAPO:    But part of the problem is,

15   when you did discuss the other clinical studies, you

16   didn't discuss the weakness or the weaknesses are

17   there, but they're hidden in small phrases behind whole

18   paragraphs on the positive and without any discussion

19   of the weaknesses of the positive, so when you read

20   that, the casual reader just comes out and says, wow,

21   that's got it.

22	DR. ROSS:    Okay.

23	DR. CRAPO:    And that's what I did,

24   that's what others did, that's what OAQPS did.

25	DR. HENDERSON:    Well, okay.  Thank you,

1  exposure values in the main ISA was in the figure

2  caption.

3	So I think possibly, if you can somehow

4  highlight exposure levels on the epi studies in the

5  same way that you do with the clinical studies and then

6  kind of bring that together rather than just looking at

7  relative risk factors, it might help things.

8	DR. ROSS:    So that's a good point

9  because in the clinical studies you actually have a

10   dose.  You have an actual number that you can

11   attribute.

12	So what we have tended to do is, for

13   example, on respiratory visits and hospital admissions,

14   on Page 5-15 at the top of the page, we report the mean

15   -- the range of mean concentrations from the studies in

16   which these are lower concentrations -- locations with

17   lower concentrations where studies were showing

18   associations range from 15 to 20 parts per billion with

19   maximum concentrations of 28 to 82 parts per billion.

20   That's a 24-hour average.

21	So that's the kind of data we were

22   pulling from epidemiologic studies, which is about the

23   best we can do because again you don't have a point to

24   say the whole population was exposed to a certain

25   level, but this gives you sort of an approximate range



207

209



1  James, for finding that.  Maybe that's where I got --

2  you know, I just read that document, too.  I'm going to

3  ask, if the EPA needs more clarification, perhaps you

4  wish we would shut up at this point, but if you...

5	DR. ROSS:    No, no.

6	DR. HENDERSON:    But if you want...

7	DR. ROSS:    If I looked frustrated it's

8  because we were actually trying to say exactly what we

9  meant to say and I don't think we intend to trash the

10   clinical studies.  I mean we believe they are as solid

11   as they are, but we'll just try to make sure we're as

12   clear as possible when we integrate when we come to

13   conclusions about them.

14	DR. HENDERSON:    You might just read it

15   through and see why we got -- if you can figure out how

16   some of us got that impression.  Jim, did you want to

17   say...

18	DR. ULTMAN:    Yes.  One reader's

19   impression.  I thought that the tenor of the chapter,

20   in fact, the whole ISA was that the epidemiological

21   studies were providing the basis of causality, but not

22   the basis of particular exposure levels.

23	In the clinical trials and the

24   toxicological data on animals, the exposure values were

25   much more up front.  I mean the only place you saw

1  of the types of concentrations, points in the

2  distribution of the concentrations for the

3  epidemiologic studies and I'm realizing that clearly we

4  did not bring this out well enough so that it was

5  obvious to the readers.

6	So one of the things we will definitely

7  do is try to bring that kind of information out, but

8  let me ask if this is the kind of information that

9  seems helpful.

10	DR. HATTIS:    Let me try to say that that

11   is helpful by itself, but it's not fully answering the

12   issue that I think that Dr. Crapo rightly brought up,

13   which is, in fact, that if it's the high excursions

14   that matter, what you want is not only the range of the

15   means, the long term averages, but some indication of,

16   you know, if there were cities that were more variable

17   versus less variable with the same mean, you know, it

18   would be nice to have that information.  I don't

19   believe that the epi studies usually give you that.

20	DR. ROSS:    We don't have that

21   information and I don't actually think we concluded

22   that it was necessarily...

23	DR. HATTIS:    Yes.

24	DR. ROSS:    ...the high concentrations

25   that mattered here.



212



1	DR. HATTIS:    Correct.

2	DR. ROSS:    That wasn't one of the

3  conclusions.

4	DR. HATTIS:    If you -- if that -- yes,

5  that wasn't one of your conclusions, so then you need

6  to be -- to communicate the fact that there might be

7  multiple ways of doing an exposure assessment that

8  would be relevant to the prediction of the health

9  outcomes.

10	DR. HENDERSON:    Okay.  Terry, did you

11   want to say something?

12	DR. GORDON:    Just a real quick one.  So

13   a process question.  So you guys did your ISA

14   completely separate from the REA and you don't have to

15   answer this, but were you surprised that they went with

16   the benchmarks of the clinical studies only?

17	DR. ROSS:    No, and actually we do

18   work...

19	DR. GORDON:    Based on James' question

20   about the...

21	DR. ROSS:    We share information.

22	DR. GORDON:    ...tenor.

23	DR. ROSS:    No, we share information like

24   Ila said.  We try to share early drafts of information

25   as soon as we can, as soon as they're close enough to

1  PM, yes.

2	DR. SPEIZER:    And wasn't there an effect

3  that you could quantify?

4	MR. AVOL:    There was an effect, but the

5  effect was...was significant for PM.  It was...the

6  trend was certainly there for NO2, but it was stronger

7  for PM.

8	DR. HENDERSON:    Lianne?

9	DR. SHEPPARD:    Yeah, I wanted to follow

10   up on...on John's comments and Ed's follow-up and...and

11   suggest that if we can't separate the NO2 effects in

12   reality because of the way things occur in the

13   environment and in models, because it's just...because

14   they occur together in the environment, it's really

15   difficult to do, but if we can make the plausible

16   argument that decreasing NO2 will, in fact, decrease

17   the impact on the population health effects, then I

18   would say the public health perspective says that

19   that's enough and that we can argue that needing to

20   separate this is really an artificial thing and

21   that...that it doesn't matter, and we can proceed from

22   a public health perspective to say that...that

23   regulation on an NO2 will have a positive impact on the

24   public health.

25	DR. BALMES:    Lianne, right on.



211

213



1  be able to inform them.

2	So we provide information to OAQPS, what

3  information we have, but it's certainly their call in

4  terms of what they feel like they can quantify because

5  what -- there are different ways to inform the policy

6  decision and the risk and exposure assessment is one

7  way, but there are also qualitative means that I'm

8  sure, when you see the policy assessment, they're going

9  to look at various ways of looking at the data.

10	So, no, I'm not going to rat out my

11   OAQPS colleagues that they did it all wrong.

12	DR. HENDERSON:    Okay, Frank?

13	DR. SPEIZER:    Can I ask Ed did your

14   in-and-out migration studies focus at all on NO2?

15	DR. AVOL:    We looked at NO2 as well as

16   BM, yes.

17	DR. SPEIZER:    And wasn't there an effect

18   that you could quantify?

19	DR. AVOL:    There was an effect, but the

20   effect was significant for PM.  It was -- the trend was

21   certainly there for NO2, but it was stronger for PM.

22	DR. HENDERSON:    Okay.  Frank?

23	DR. SPEIZER:    Can I ask, Ed, did your in

24   and out migration studies focus at all on NO2?

25	MR. AVOL:    We looked at NO2 as well as

1	DR. HENDERSON:    Someone said right on.

2  Was that...

3	DR. BALMES:    John Balmes.

4	DR. HENDERSON:    ...John Balmes?

5	DR. BALMES:    I'm complimenting her for

6  her articulate presentation of the point that I was

7  trying to make.

8	DR. HENDERSON:    Thank you, John, and

9  Terry has got his hand up.

10	DR. GORDON:    And...and...and you just

11   said what I was going to say.  In the integration of

12   the whole thing and what James is getting at, I...I

13   agree with you completely, and I was just...a warped

14   sense of thinking about things, but the epi for PM says

15   there's pulmonary effects, and the NO2 says there's

16   pulmonary effects, and what does the clinical and

17   animal tell us?  Well, the clinical and animal and NO2

18   says there's only pulmonary effects, and the epi for

19   NO2 says there's no cardiovascular.

20	So, if we try to separate PM and NO2, I

21   mean, I could make the crazy supposition that PM's not

22   causing any pulmonary effects, because it's actually

23   NO2.  PM's only causing cardiovascular effects which,

24   indeed, the animal and the human tox data says there's

25   very, very little pulmonary effects of PM, but there,



216



1  of course, is cardiovascular effects from the panel

2  studies and the animal studies.

3	So, it's...it's so tightly entwined, I

4  like what you...what you put forth.

5	DR. HENDERSON:    So, we protect public

6  health when we are reducing the NO2.  Good...good

7  point.

8	Did you have your hand up, Ted, or did

9  you just...

10	DR. RUSSELL:    I just...I don't think we

11   really want to go down that route, because there are

12   controls that will act on NO2 that won't act on other

13   combustion byproducts and vice versa.  So, I would hate

14   for us to start dreaming up standards or something that

15   could lead to controlling the wrong pollutant.

16	DR. SHEPPARD:    Well, that's why...that's

17   why my comment said, you know, if we can...if reducing

18   NO2 will, in fact, reduce the effects, because yeah,

19   it's more complicated.  I realize that.

20	DR. HENDERSON:    I think eventually, long

21   after I am dead, the Clean Air Act will be revised to

22   just multi-pollutants, and then we can say, you know,

23   regulate combustion products, for instance, rather than

24   single pollutants, and maybe that would help us with

25   our problems.

1  in.  I, obviously, won't be speaking to each person

2  individually, but they could all look at it and decide

3  if I've captured what they said properly.

4	DR. HENDERSON:    Yeah.  You can...the

5  idea is that those who are developing these paragraphs,

6  you know, you work any way you want to develop it, but

7  I can envision you drafting something and emailing it

8  to others and saying is this...have I got it, you know,

9  about right?  So...

10	DR. CRAWFORD-BROWN:    Yeah, okay.

11	DR. HENDERSON:    I've got a schedule.

12   Did...did Doug get the schedule?

13	DR. NUGENT:    Yes, and...

14	DR. CRAWFORD-BROWN:    I did, a schedule

15   of when various things should be?

16	DR. HENDERSON:    Yeah, yeah.  So, what...

17	DR. CRAWFORD-BROWN:    I got that.

18	DR. HENDERSON:    What we're hoping is in

19   about a week, that Angela and I will get these draft

20   discussions, and we can put together a letter which

21   will be sent out to all panelists on May 16th, and then

22   you'll make comments back by May 23rd, we hope, and

23   then we're going through this...we'll go through a

24   process where, you know, if you have comments on the

25   letter or there are problems, of course, we'll respond



215

1	Right now, I want to sum up where we're

2  going from here.  I hope everyone who was a lead

3  discussant realized that they're responsible for

4  getting together with others to draft a...a paragraph

5  in our letter to the Administrator.  There will be a

6  letter to the Administrator expressing our consensus

7  views on the answers to these charge questions, and so,

8  those responses will be drafted by the...the lead

9  discussant.

10	Doug, do you feel comfortable doing that

11   from overseas?  Are you there, Doug?  He's prob...is

12   anybody on the phone?

13	DR. BALMES:    Yeah, I'm here.

14	DR. HENDERSON:    Okay.  Well, we'll have

15   to get in touch with Doug, but the...the schedule...

16	DR. CRAWFORD-BROWN:    I hit the wrong

17   star-6-pound-6 combination there.

18	DR. HENDERSON:    I...I figured.  I could

19   just see you there punching the buttons.

20	DR. CRAWFORD-BROWN:    What do you mean,

21   you figured?  You figured I had the wrong combination?

22	I...I have listened to what everybody

23   said.  I've been taking notes on this and have looked

24   back through everybody's comments, and so, I feel

25   comfortable writing up a paragraph and...and sending it

1  to that.

2	So, everybody will have a chance for

3  input on the letter.  And then, the revised draft

4  letter will be posted on the web by Angela, and we have

5  a telephone...a teleconference scheduled for June 11th

6  where we will meet a legal requirement for the charter

7  CASAC to approve this letter in a public fashion so

8  that the public can express themselves if they so

9  desire.

10	And, again, as Vanessa said, we

11   didn't...when I first came on as chair, we weren't

12   doing this, but one of...there's a lawyer at EPA who

13   said we must do this, so that's why we're doing it.

14   I'm being really open about why we're doing that.

15   That's a little change.

16	But that's what we hope will happen.  We

17   hope we can stick to the schedule, and work with your

18   colleagues to get those draft paragraphs in.

19	Are there any questions

20   about that?

21	DR. SPEIZER:    Just to confirm, that's

22   just the charter members that will meet.  Right?

23	DR. HENDERSON:    That need to be on the

24   phone call, but I don't want anyone else to feel like

25   they're not in on it, because you will have your chance

217



1  to modify it and...and...and, really, we're meeting a

2  legal requirement when we have the charter phone call.

3	And, Angela, you want to tell us about

4  dinner tonight just so you can...

5	DR. NUGENT:    I set up a reservation at a

6  nearby restaurant for dinner tonight for the

7  chartered...for the members of the committee, and I'll

8  circulate a list for anyone who wants to participate to

9  join in, and there's information on that about when

10   we're...when the van will take us on over to the

11   restaurant.

12	DR. HENDERSON:    So, you...you've got it

13   written down?

14	DR. NUGENT:    Yeah, I'm going to

15   circulate that.

16	DR. HENDERSON:    I think your reservation

17   was for 6:30 tonight?

18	DR. NUGENT:    That's right, with

19   the...the van leaving at 6:15.

20	DR. HENDERSON:    Yeah.  So, just show of

21   hands, how many would like to participate in this

22   dinner?

23	SPEAKER:    Where is it?  Where is this?

24	DR. HENDERSON:    Where is it?

25	DR. NUGENT:    (inaudible).

220

1  presentation on our next document.

2  (WHEREUPON  , a brief recess was taken.)

3	DR. HENDERSON:    So, we have a new cast

4  of characters here.  Okay, we're going to start on our

5  new document now, and we have some people from the Air

6  Office present.  Maybe we could get...may we get you to

7  introduce yourselves so we all know who you are?

8	MR. RICHMOND:    Harvey Richmond, the AMPI

9  Centers Group, OAQPS.

10	DR. JENKINS:    Scott Jenkins, same.

11	DR. GRAHAM:    Stephen Gra...oop.  Hey,

12   I'm speaking into the mike.  Stephen Graham, OAQPS.

13	DR. ROSENBAUM:    Arlene Rosenbaum from

14   ICF International.

15	DR. HENDERSON:    Thank you very much.

16   Okay, and who is going to be the lead person to start

17   out?  We've got copies, I think, of the overview.

18   Everyone should have them.  So, just...

19	DR. JENKINS:    I'm going to talk for the

20   first couple of slides, and then Stephen's going

21   to...going to chime in.

22	DR. HENDERSON:    Okay, proceed.

23	DR. JENKINS:    All right.  Can I have the

24   next slide, please?

25	Okay, I'm going to talk for just as few



219

221



1	DR. HENDERSON:    So, you've got a

2  good...it's seafood grill or something?  What is it?

3	DR. NUGENT:    Rockfish Seafood Grill,

4  yeah.

5	DR. HENDERSON:    We've been there before.

6  Okay, so you've got a good group.  Okay.  Well, we're

7  going to take a 15-minute break.

8	DR. ROSS:    Rogene?

9	DR. HENDERSON:    Oops.

10	DR. ROSS:    Before you leave, we were

11   talking about next steps, and I...

12	DR. HENDERSON:    Oh, yes.

13	DR. ROSS:    ...I just want to remind you

14   that our next step is a court-ordered deadline of July

15   11th for the completion of this document, putting it on

16   the web, and we're going to move forward as soon as

17   possible after this meeting and try and make revisions.

18   So, it will be good to get your letter as soon as we

19   can.

20	DR. HENDERSON:    Okay.  Well, you have

21   our schedule, and I think Angela is very good at this.

22   She's pushing us, but we'll try to meet that.  Well, we

23   have that teleconference is definitely set.  So, okay.

24   So, now we know what all the deadlines are, and we can

25   take a 15-minute break and come back at 3:00 for our

1  slides and give a little bit of background, the time

2  line, purpose and scope of documents, and then give a

3  very brief overview of our approach, and then Steve,

4  I'm going to turn it over to Stephen, and he's going to

5  talk about our risk characterization based on the air

6  quality analysis and then on the...based on the

7  exposure analysis.  Next slide.

8	So, the only thing I'm going to point

9  out on this time line slide is the...the bolded...the

10   bolded line there.  This is where we are right now.

11   We're going to be coming back to CASAC with a second

12   draft of this risk and exposure assessment document in

13   August for a meeting that, I believe, has been

14   scheduled for early September or tentatively scheduled

15   for early September, and then we're...we're planning to

16   finalize the document in November.

17	Then we move into the rulemaking process

18   with our Advanced Notice of Proposed Rulemaking in

19   December of '08.  Okay, next slide, please.

20	So, obviously, the purpose of these

21   documents is to convey the approach that we've taken to

22   assessing risks and exposures and to convey the results

23   that we've...we've obtained.  Ultimately, when we get

24   to the point of finalizing the document, it's

25   going...along with the IS...the final ISA, this is



224



1  going to inform the rulemaking process.

2	The first draft, which is what we're

3  here to...to discuss today, considers recent levels of

4  NO2 as well as...as NO2 levels that are associated with

5  just meeting the...the current standard, and we can

6  talk a little bit more about why we...why we do that in

7  a minute during the discussions if there...if there are

8  questions.

9	The exposure assessment, we conducted

10   that in a single location, Philadelphia County, and in

11   subsequent drafts of the document, we're also going to

12   address levels that...levels of NO2 that are associated

13   with just meeting alternative or potential alternative

14   standards and also potentially expand the exposure

15   assessment to include additional areas, and that will

16   depend on some of the feedback that we get from the

17   committee here today.

18	Okay, just a very, very brief overview

19   of the approach that we've taken.  The exposure

20   characterization, we...we essentially proceeded in two

21   steps.  First was the air quality analysis.

22	DR. GRAHAM:    I think this is where I

23   chime in.

24	DR. JENKINS:    Oh, actually...

25	DR. GRAHAM:    I'm sorry.

1  the epidemiological literature, at least qualitatively,

2  as part of what we call our evidence-based approach to

3  assessing potential alternative standards.

4	So, for example, when we identify

5  potential alternative standards, one thing we...we can

6  do and one thing we anticipate doing is actually

7  comparing the levels of those potential alternative

8  standards to the levels that have been obs...that are

9  observed in different epidemiological studies and

10   making a qualitative comparison to inform the judgment

11   on...on an eventual decision on a standard.  So, that

12   is...at this point, that's how we envision using the

13   clinical versus the epi literature.

14	Okay, now, with the next slide, I'm

15   going to turn it over and let Stephen go ahead.

16	DR. GRAHAM:    Well, actually, I just want

17   to put both of these components in a better frame, and

18   we could have left the earlier slide up.  The exposure

19   characterization was done in these two components, and

20   they were mainly complementary, not necessarily

21   designed to have one feed into the other.

22	The air quality analysis was really a

23   broad analysis that included all the NO2 monitors

24   across the United States, and the purpose was to serve

25   as a surrogate for personal exposure, and those ambient



223

225



1	DR. JENKINS:    No, no, no.  Actually, let

2  me go through this one slide, and then you go on the

3  next slide.

4	DR. GRAHAM:    Okay.

5	DR. JENKINS:    Because it's...this is so

6  general that it's not even worth your breath.

7	The...we proceeded in basically two

8  steps.  The first is the air quality assessment, and we

9  used the ambient levels derived from a combination of

10   monitors and modeling, and then the second was the

11   exposure analysis where we actually used APEX and

12   considered time spent in different micro environments.

13	With the risk characterization, we just

14   spent some time at the end of the last conversation

15   talking about how we characterize risks, and that, you

16   know, we chose potential health benchmarks from the

17   human clinical literature, and you can see them right

18   there, and, basically, we...we compare these benchmarks

19   to our estimates of exposure that were identified, that

20   we derived.

21	Now, I do want to make one comment, one

22   brief comment, on the epidemiological literature, since

23   I can see that's going to be an issue for discussion

24   today or tomorrow, at the...at the least, and that is

25   that one way that we would choose to use...we would use

1  concentrations of NO2 were evaluated at each of these

2  monitors, and we considered the spatial and temporal

3  variability in those concentrations.  And in

4  recognizing that there is this vari...variable

5  relationship between the personal exposure and the

6  ambient monitors, we decided to take a simple approach

7  to estimate on-road concentrations using those ambient

8  concentrations as input.

9	Then, the second component was the

10   exposure analysis that Scott mentioned where we did a

11   combined dispersion and exposure modeling exercise, and

12   the purpose of this was two-fold, to improve the

13   spatial characterization of the air concentrations,

14   because in most of these loctations, there are a

15   limited number of ambient monitors, and, in addition,

16   the second purpose was to simulate the contact of

17   persons with the air concentrations.

18	So, Philadelphia was...was selected as

19   that case study location.  Again, it's a more complex

20   analysis, so it wasn't like we could actually do that

21   across the United States.  And it serves about 1.5

22   million...I'm sorry...the population is about 1.5

23   million, and we did it for three years, 2001 through

24   2003, and modeled 17,000 receptors at various

25   locations, including census block centroids and



228



1  linked...roadway links.  Sorry.

2	The AIRMOD portion, we considered

3  stationary sources, fugitive, and airport emissions, as

4  well as these mobile source roadway link-based emission

5  estimates, and 1-hour concentrations were estimated for

6  those locations, the census block centroids, available

7  ambient monitors that were within Philadelphia County,

8  and these major roadway links.  And APEX was used

9  taking those air concentrations as input, APEX was the

10   exposure model used which, of course, considers time

11   spent in various micro environments and the variable

12   concentrations that occur within those micro

13   environments, and we estimated longitudinal profiles,

14   exposure profiles for each of those three years for

15   each simulated person.

16	And the reason why I...I wanted to go

17   into this description of everything first is because

18   well, as I said, there isn't necessarily a direct feed

19   between one analysis and the other, but there is a

20   little bit of overlap in that for the air quality

21   analysis, we did use that information to select the

22   locations for more refined analysis in the exposure

23   assessment as well as the air quality monitors

24   themselves were used to estimate a local concentration

25   for the air concentrations that were modeled by AIRMOD.

1  mean number of exceedances that were estimated to occur

2  within a year in that particular location.

3	And I did mention that all the monitors

4  were represented here.  We had specific locations, and

5  anything that wasn't defined as a specific location,

6  they were grouped in this...these other two categories,

7  non-MSA and other CMSA/MSA.

8	DR. HENDERSON:    Would you mind, just

9  because we've just had this discussion of these

10   benchmark values, saying exactly where the benchmark

11   values came from, the 200, 250, and 300?

12	DR. GRAHAM:    Well, in...in the ISA, in

13   Chapter 5 of the ISA, there were...there were two main

14   conclusions regarding the human clinical literature.

15   One was regarding the literature on airways or on

16   the...on airways hyperresponsiveness.  So, the...the

17   first was with regard to hyperresponsiveness

18   following...following inhalation of...of an allergen,

19   and that...in that...for that, the ISA concluded that

20   effects were seen as low as 0.26 ppm for a 30-minute

21   exposure.

22	The other major conclusion regarding

23   hyperresponsiveness was with respect to responsiveness

24   following non-specific...inhalation of a non-specific

25   stimulus, and for that, the ISA concluded that effects



227

229



1	So, if we can go to the next slide, this

2  is, I guess, a little bit more elaboration on the air

3  quality analysis, including the...the description of

4  the data that were used, and it was '95 to 2006.  We

5  separated it into two separate groups.  One we termed

6  history, the other more recent data, and then we

7  selected particular locations to...to do a more focused

8  analysis, geographic analysis, based on these criteria

9  of whether they had a high annual average concentration

10   as well as a current of p concentrations above 200 ppb

11   for...for one hour.

12	And then we looked at a variety of

13   scenarios that you see here, quality as is, just

14   meeting the current standard, and that simplified

15   version of how we estimated the on-road concentrations,

16   both for concentrations as is as well as just meeting

17   the...the current standard.

18	Next slide, please.

19	So, here's a...an example or a summary

20   of some of the data that were generated for each of

21   these scenarios, and it just provides a comparison

22   across the various simulations that were performed,

23   and, again, this is the results for the air quality

24   simulation for each of the cities, and we've got some

25   means and 98th percentiles represented.  And this is a

1  were seen at levels as low as 0.2 to 0.3 ppm for 30

2  minutes.

3	So, the...the range that we identified

4  here for this...for our assessment, 0.2 to 0.3 ppm, was

5  based on essentially those two conclusions, and the ISA

6  doesn't distinguish between 0.2 and 0.3 ppm for

7  that...for that non-specific hyperresponsiveness, so

8  we...we evaluated the range.

9	DR. HENDERSON:    That's good.  That's

10   what I thought it came from, but I just wanted to be

11   sure.

12	DR. GRAHAM:    I just want to make two

13   additional notes on this table.  These actually are

14   absolute counts of occurrences, so if there were p

15   concentration or a concentration above whatever

16   benchmark we were looking at, this is 200, a summary of

17   the 200 ppb.  If there were sequential occurrences,

18   this would...this would count that, and that's actually

19   distinguished...I'm sorry...distinct from the exposure

20   modeling, and I'll...I'll get to that in a moment.

21	The other thing to note is I did submit

22   a technical memorandum, and I noticed that this

23   contains the original analysis.  The numbers for the

24   technical memorandum are slightly different, and, for

25   the most part, you can see most locations ended up with



232



1  lower means.  It was particularly evident at the higher

2  percentiles for the on-road concentrations.

3	Next slide.

4	So, as far as the exposure analysis, I

5  probably described that in...in greater detail than

6  what was here, so we estimated the hourly

7  concentrations for simulated individuals across three

8  years and then, again, performed summary analyses to

9  determine the frequency of occurrence of these

10   exposures above those given potential benchmark levels,

11   but the difference between the absolute counts that

12   were estimated from the air quality characterization,

13   these were actually the maximum, if there was an

14   occurrence of the...I'm sorry...there was an occurrence

15   of one of these exceedances in a given day, then it was

16   counted as one.  There wasn't any counting of

17   sequential hourly exceedances for...for the exposures.

18	Next slide.

19	This is just an example of the exposure

20   results for...for the different years, and we also...I

21   guess I didn't mention that earlier.  We did do indoor

22   sources, gas stoves, and this is a summary of the

23   results, the maximum exposures that individuals, the

24   model simulated individuals, experienced, and these are

25   distributions for that, the...the CBFs, and we've got

1  different years.

2	Next, please.

3	Then, when we did a model run without

4  the indoor sources, you can clearly see the numbers of

5  individuals that are exposed to each of the potential

6  health benchmarks at least once within the year is...is

7  much lower.

8	Next slide.

9	And here's an example of results where

10   we see the fraction of the population, the asthmatic

11   population, mind you, above each of the potential

12   health benchmarks with just meeting the current

13   standard.  And that's with indoor sources, and then we

14   have without indoor sources.

15	And the next...next slide.

16	And, of course, the numbers are...are

17   lower.

18	If you hit the next slide.

19	Because we're estimating the

20   longitudinal profile, the exposure profile for each of

21   these simulated individuals, we do have ability to

22   capture whether these individuals receive multiple

23   exposures, and that's what this graph is showing, is at

24   least one exposure is in the back, and as you progress

25   towards the...the front of the graph, that's where you



231

233



1  three years here, 2001 through 2003, with indoor

2  sources and one model run where we had excluded the

3  indoor sources, and it's distinct that those

4  concentrations, the exposure concentrations people are

5  experiencing, the max, was...was much less than with

6  the indoor sources.

7	Next slide.

8	And here is another example output from

9  the model where we looked at those three potential

10   health benchmark levels across the different years, and

11   this includes the indoor sources, and what we have on

12   the...the y axis is the percent.  And the focus here in

13   this analysis was on these susceptible/vulnerable

14   populations.

15	And this is asthmatics, asthmatic

16   individuals, total.  We looked at children as welll,

17   children asthmatics, and the...the results for those

18   were similar, and all of the results have been

19   presented in the document.  This is just

20   illustrative...an illustrative example.

21	And you can see that, of course, more

22   individuals are...or the fraction of the population

23   that are exposed to the lowest potential health

24   benchmark is greater than the...the higher benchmark

25   levels.  And there's some variability across the

1  would see multiple exposures, up...upwards to six for

2  each of the different benchmark levels.

3	And this one has with indoor sources,

4  and the next slide, no indoor sources for each of the

5  three years.

6	Next slide.

7	And then, where we simulated just

8  meeting the current standard, of course, the numbers of

9  individuals or the percent of individuals exposed

10   is...is greater and with indoor sources, and the next

11   slide is without indoor sources.

12	That concludes that.

13	DR. HENDERSON:    Thank you.  Does that

14   conclude the whole presentation?

15	MR. RICHMOND:    That's right.

16	DR. HENDERSON:    Okay.  So, now it's time

17   for public comments.  Do we have public commenters?

18	DR. NUGENT:    Yes, ma'am.  We have one

19   public comment from Ann Smith from CRA International,

20   and she is presenting some comments on behalf of the

21   Utility Air Regulatory Group.

22	I'd like to note for our people on the

23   line...for our members on the line that her

24   presentation is available on the website if you go to

25     HYPERLINK http://www.epa.gov/casac  www.epa.gov/casac and look u
nder calendar, you'll find



236



1  the meeting date for today.  It's one of the meeting

2  materials that's appended there for the...this meeting

3  record.

4	DR. SMITH:    Thank you very much.  I've

5  sent you all written comments that are going to cover

6  the topics that I...that cover the topics that I've

7  listed here.

8	If you could go to the next slide...and

9  I'm going to cover mainly two of the issues that...what

10   I'm trying to do in the comments here is not dwell on

11   the numbers that are coming out of the analysis at this

12   point but, rather, to look at the analysis as it stands

13   in its structure and the kind of data that are being

14   used to understand if we have a useful construct that

15   can be interpreted and used in policy making.

16	So, I...you can read later more or less

17   the second, third, and fourth of these comments, but

18   I'm just going to focus for the moment on the first

19   two.  One is we're talking about the benchmark issue,

20   level issue, and then I'd like to comment on the site

21   selection.

22	Next slide.

23	In order for a risk assessment to be

24   useful, it has to have some grounding back to some

25   interpretation of what the benchmarks mean in the way

1  from the lowest dose up to the highest dose in those

2  studies which was information on the previous slide,

3  and now we can break them...break these studies out

4  that are in the table into the three areas where the

5  benchmarks are set.  And the thing you'll notice is

6  that for the exposures that are around 200 ppb, the

7  lowest end of the benchmark, there are only two studies

8  that are identified in this supporting table, and

9  neither of them found a statistically significant

10   effect.

11	What this suggests also is you see not a

12   lot of appearance that the significance is more

13   frequently occurring as you go to higher and higher

14   doses.  This table goes all the way...to exposures all

15   the way up to 600 ppb exposures.

16	So, while there certainly are some

17   effects being observed for airways responsiveness down

18   in this range, it doesn't appear that there's much

19   justification to go below about the 250 ppb, at least

20   as the data have been presented and justified for these

21   benchmarks.

22	Now, I'm not going to try and argue what

23   the right benchmark level should be, but what my point

24   is is that whatever benchmarks are in the REA, they

25   need to be given some grounding in the underlying



235

237



1  of risk.  This is particularly true if you're going to

2  be relying on concentration response functions to

3  quantify the estimates that...on benchmark levels.  EPA

4  has selected three benchmark levels.  We've talked

5  about that already, and the real question is,

6  what...what do these really mean in terms of risk?

7  What is their grounding in risk?

8	And there are some statements made in

9  the document.  The quote here comes straight from the

10   REA, that these represent the lower, middle, and upper

11   end of the range identified in the ISA as the lowest

12   levels at which controlled human exposure studies have

13   provided sufficient evidence for the occurrence of NO2

14   related airway responsiveness.

15	So, that's what it says about them, but

16   to me, the question is, well, how is that grounded in

17   what the science is saying?  And the REA actually does

18   provide Table 1 which is what shows on the next slide.

19   I've just pasted it in here to show you.

20	First, I'll point out that this table,

21   there's nothing quite like it in the ISA.  This didn't

22   come from the ISA, but this is what the REA is

23   providing in the way of backup for the justification.

24	If we go to the next slide, I've just

25   taken the data in that table and reordered it according

1  science and some explanation of just what you can take

2  them to mean from the point of view of risk.

3  Otherwise, to know that there are so many exposures or

4  so many individuals being exposed above a benchmark

5  level doesn't really tell people anything that's

6  relevant for policy making purposes.  And, ultimately,

7  the risk assessment has to be relevant for that.

8	Next slide.

9	The second point I want to turn to is

10   the question of the choice of the cities, and there's

11   no question that what is being selected in the way of

12   cities in both parts of the REA are worst case

13   conditions for exposure, and it's my belief that in a

14   good risk assessment, you really want to provide some

15   representation across the whole board of types of

16   exposures and information on how many people exposed to

17   the different levels of exposure.

18	And here you can see that the

19   exposure...the criteria in the exposure modeling is so

20   selective that only two cities in the whole country

21   actually meet the selection criteria.  EPA put in a

22   couple more criteria by reaching back in time to find

23   some more worst case cities.

24	The key point here is we're getting in

25   the exposure modeling five cities that are clearly



240



1  among the worst in the country, and we have no

2  information that will come out of the REA if we use

3  only these cities to indicate what's happening in a

4  broad representative sense across the country for risk,

5  and I think that should be relevant for policy making

6  purposes in addition to what's happening in the worst

7  case.

8	Furthermore, in the air quality

9  characterization case, there are 18 cities that meet

10   the screen.  That's a less selective screen, but

11   they're all worst case.  They're all in the worst

12   90...10th percentile of exposures for the annual

13   average.  And when you have 18 cities and you're

14   showing exposures and exposure rates above benchmarks

15   in all 18 cities, one would, as an audience for this

16   REA, start to get the feeling that this is what's

17   happening across the whole U.S., that it ought to be

18   representative, but by construction, it's not

19   representative.  And so, more other cities ought to be

20   included and properly characterized as to what

21   proportion of the population in the U.S. faces those

22   additional exposures.

23	So, just to conclude on these first two

24   points alone, both the fact that the benchmark levels

25   are at levels at which effects have not been

1  representativeness within the cities based on your

2  evaluation of this document?

3	DR. SMITH:    Could you explain what you

4  mean within cities?

5	DR. SHEPPARD:    Well, you know, the

6  city...

7	MR. SMITH:    Oh, the monitors, for the

8  monitors not being in representative locations?  I

9  think the exposure modeling piece is supposed to try to

10   get at that, because what that's doing is trying to

11   monitor...come up with exposure estimates on a smooth

12   spatial area and then move people in and out of the

13   spatial orientation.  So, the exposure modeling comes a

14   lot closer to an attempt to understand how people are

15   being exposed above some benchmark level, whereas I do

16   agree that the air quality characterization, I'm not

17   sure it's very useful to know how many monitors go

18   above a benchmark level.  To me, that's really air

19   quality characterization and not risk exposure

20   characterization at all.

21	DR. SHEPPARD:    But, actually, monitored

22   air quality characterization as opposed to air quality

23   characterization.

24	DR. SMITH:    And...and...yes, and that

25   gets to my second point, or my fourth point, actually,



239

241



1  documented, at least for what they're stated to be, and

2  the fact that the analysis is focusing only on worst

3  case cities implies that this particular REA is going

4  to characterize NO2 exposures in a way that will lead

5  to an overstatement of the nature of the problem,

6  particularly at as-is conditions.

7	And this concern is really just

8  exacerbated by the other two points that I don't have

9  time to go into here, the fact that the roll-up to...to

10   simulate exposure at the current NAAQS is so extreme

11   and also that there do appear to be inconsistencies in

12   the data between the two parts, that the air quality

13   data that's being used to assess exposures above

14   benchmarks may be highly inconsistent between the air

15   quality characterization portion and the portion that

16   does the exposure modeling.  That really needs to be

17   explored in order to make sure that both parts of the

18   REA are telling a consistent story when they're all

19   said and done.

20	Thank you.

21	DR. HENDERSON:    Thank you very much.

22   Are there questions?  Yes, Lianne?

23	DR. SHEPPARD:    Given...given your very

24   important point about representativeness of the cities

25   across the U.S., do you have any comments about

1  in the...in the comments, which is that when you get to

2  the exposure modeling, that's where you do have more

3  complete information on what exposure may look like

4  that needs to be consistent with the monitored

5  information, and the evidence that is available in the

6  tidbits of evidence that are available in the technical

7  support documents and the REA suggest that maybe

8  they're not, and until they are, you can't really trust

9  the exposure modeling, because that needs to be

10   demonstrated that they're consistent.

11	DR. HENDERSON:    Okay, are there other

12   questions?  Yes, Frank?

13	DR. SPEIZER:    Given you are being

14   critical of these sort of what you call worst case

15   cities, what proportion of the population is

16   represented by these cities?

17	DR. SMITH:    I think it should be

18   represented in the...in the document.

19	DR. SPEIZER:    Well...

20	DR. SMITH:    That information isn't

21   provided.

22	DR. SPEIZER:    Has not been.  Okay, well,

23   maybe the question is really to EPA to tell us what

24   proportion is represented by the total selected cities.

25	DR. GRAHAM:    I'm sorry, I can't answer



244



1  that right now.  No, I don't mean to be like that.  I

2  don't have the populations in front of me.

3	DR. SPEIZER:    Yeah, but you probably

4  ought to be able to get that you're representing...

5	DR. GRAHAM:    Right, yes.

6	DR. SPEIZER:    ...a third of the U.S.

7  population or whatever it is.

8	DR. GRAHAM:    As I say, with

9  Philadelphia, it was 1.5 million.

10	DR. HENDERSON:    Are there other

11   questions?

12   (No response.)

13	DR. HENDERSON:    Okay, I think you close

14   out the...

15	DR. NUGENT:    Thank you.  Thank you for

16   the presentation, Dr. Smith.

17	DR. HENDERSON:    Okay, we'll go now into

18   the CASAC review.  There are three areas covered by

19   this document.  One is air quality, one is exposure,

20   and one is characteristics of health risk, and we're

21   going to start out with the air quality information and

22   analyses, and Christian Seigneur is our lead

23   discussant.

24	DR. SEIGNEUR:    Okay.  Well, I will start

25   on that.  I have several comments which I've given

1  reproducing the spatial variation that you have in the

2  measurements, and if you proceed to use the modeling

3  results after that to get the distribution exposure,

4  you simply have the wrong representation of the spatial

5  distributions over the two concentrations from the

6  model even after the adjustment.

7	So, there are two things which could be

8  done.  One would be to identify the cause for that

9  significant underprediction, whether some initial

10   inventory is incorrect and can be correct, or whether

11   the model, when applied in that region, is missing

12   something.  One possibility is that lack of NO2

13   chemistry is affecting emissions from the point source.

14   I don't know.

15	The other possibility, of course, is

16   to...if we can't identify the cause, is to use the

17   model in combination with the data and to do some data

18   fusion so that you can then keep the spatial gradients

19   that you have in the measurements.  I don't know if you

20   want to answer that right now?  Okay.

21	DR. ROSENBAUM:    Yeah, I just want to

22   point out when we did this model...the AIRMOD leveling,

23   we didn't actually try to model all the sources.  In

24   particular, we only tried to model the major roads

25   which we defined as roads that had annual average daily



243

245



1  already in writing, and I don't have concerns about the

2  use of the air quality data but most...more the

3  implications of the air quality model which was the

4  second part of the analysis that you did for

5  Philadelphia.

6	My main concern was that when I looked

7  at Table 26 which shows the performance of the model

8  compared to the data at the monitors, and you showed

9  only three monitors in that table.  Performance, I

10   think, was pretty good for the two, the third and

11   the...the second and third monitors, but it was pretty

12   bad for the first.  I mean, second and third monitors

13   appear within 30, 40 percent of most of the

14   measurements.

15	The first monitor, there is a difference

16   of a ratio of about 3 between what was measured and

17   what the model did, and to me, that...that performance

18   is inadequate, because then you proceed to make an

19   adjustment by increasing the modeling values by the

20   same amount across the entire area.  As a result of

21   that, you move from a slight underprediction to a

22   slight overprediction at monitors 2 and 3, but at the

23   first monitor, you still have a significant

24   underestimation.

25	And my concern is that the model is not

1  traffic of 15,000 or greater.  So, that encompassed

2  actually only about 40 percent of the traffic

3  emissions.

4	So, there's...there's a...there's a lot

5  of emissions that are't...that weren't modeled.  The

6  reason we did it that way is because we're actually

7  looking for just the hot spots which we assumed would

8  be near the major roadways and that the...the other

9  traffic would be more evenly dispersed and wouldn't

10   create hot spots.  So, I mean, the...so that's partly

11   why you're seeing that, the discrepancy between the

12   monitor and the modeling, is that we weren't...in the

13   modeling, we weren't trying to be comprehensive in

14   giving all the sources, just the big sources where we

15   thought there might be hot spots.

16	So, on the one monitor where there's an

17   even larger discrepancy, it just so happens that that

18   monitor is not near any major roadways or point

19   sources.  So, it's...the monitor value is coming

20   primarily from the local traffic that we didn't model.

21	DR. SEIGNEUR:    Okay.

22	DR. GRAHAM:    Right.  I'm

23   sorry...and...and that's why it was corrected for a

24   factor.  The factor is attributed to sources not

25   modeled.



248



1	DR. SEIGNEUR:    Right, but I'm sure

2  you...you've done it by a factor.  You increase the

3  modeling by a constant value, and that's...that's my

4  contention, because by doing that, you increase the

5  values at all the monitors whether the underestimation

6  was very small or very large, and when you look at the

7  results of the model after integration, they still look

8  pretty bad.

9	So, my recommendation is that your

10   adjustment should be receptor specific.  You need to

11   increase the modeling results at the location where you

12   have a gross underprediction by a larger value than you

13   would adjust them at the other locations.  So, that's

14   why I was...why I was referring to data fusion.  You

15   need to do something a bit more complicated than what

16   you have done when you address the modeling results.

17   So, that was my...my main comment.

18	The other comment I have, I was a bit

19   confused on why...I understand why you roll back the

20   benchmark when you are looking at the results when you

21   only have outdoor sources so that you don't have to

22   redo all the calculations.  You simply change your

23   benchmark and scale everything.

24	I got confused when you did that when

25   you have outdoor and indoor sources, because

1  were...were massive.

2	So, rather than adjust the factor...I

3  mean, adjust all of those concentrations to have it run

4  in the model which would also take additional time, we

5  decided that it was mathematically equivalent to just

6  roll back that benchmark concentration by an

7  appropriate factor, and if you take a look at the

8  equation that's listed there, the exposure is the

9  combination, proportional combination, of the ambient

10   concentration as well as the contribution from the

11   indoor sources, and we are estimating that exposure and

12   then determining whether or not it exceeds this given

13   benchmark.

14	Well, if we adjust the air quality by a

15   given factor, you can see we multiply it by the ambient

16   concentration as well as these different proportions

17   that would be assigned to the exposure

18   concentrations...I mean, that would go into the given

19   exposure concentrations, and then you'd determine your

20   benchmark.

21	If we just divide both sides of that

22   inequality by the factor, you can see if we divide the

23   indoor source contribution by the given factor and the

24   benchmark by the factor, we will get the same answer as

25   if we rolled up the concentrations.



247

249



1  mathematically, that did not make sense to me.

2	DR. GRAHAM:    Angela, the slide, if you

3  could pass that out?  Oh, you have it.  I'm sorry,

4  that's right.

5	So, for the air quality

6  characterization, I think I mentioned earlier that we

7  had raised the concentrations at each location by a

8  given factor, and that factor was determined by the

9  highest monitor, the highest annual average for each

10   year.

11	No, no, we don't need those.

12	SPEAKER:    We don't have it on the

13   slides.

14	DR. GRAHAM:    Yeah, I'm sorry.  It's the

15   slide that was passed out.  Sorry.  The hard copy.

16	So, that was the air quality, but what

17   you're referring to is what we did for the exposure

18   modeling where we actually...

19	DR. SEIGNEUR:    Right.

20	DR. GRAHAM:    ...had to roll back the

21   benchmark, and that's explained in the second section.

22   If we...I guess to make it a more efficient process,

23   instead of...because of the size of the files.  I mean,

24   we had 17,000 receptors for three years across, you

25   know, every single hour.  I mean, the size of the files

1	DR. SEIGNEUR:    Okay.  So, you mean that

2  you divided indoor sources by the same factor?

3	DR. GRAHAM:    Right.

4	DR. SEIGNEUR:    Okay.  All right.

5	DR. HATTIS:    Well, I guess I have a

6  problem, but I'll wait until later, but it seems to me

7  that you...your response doesn't completely disclose

8  that that implicitly says that the indoor sources

9  are...are changing by the same factor as the outdoor

10   sources, that your assumption of doing that must mean

11   that your indoor sources are changing by the same

12   factor as your outdoor sources, and you...that's

13   clearly not right.

14	DR. GRAHAM:    No, no, that's not what's

15   implied.  If...do you have the slide?

16	DR. HATTIS:    I've got the slide.

17	DR. GRAHAM:    Okay.

18	DR. HATTIS:    And I also read exactly

19   what you said you did, and it's not...I mean, clearly,

20   the indoor source...if you...if you increase the

21   outdoor concentrations, they will make up a total...a

22   different fraction, a larger fraction, of the total

23   exposures than indoor sources.  So, you must have

24   assumed that that's a constant fraction.

25	DR. ROSENBAUM:    I think what we did was



252



1  we scaled back the indoor sources by the same amount

2  that we scaled back the benchmark threshold so that

3  we...so that we...we made that adjustment...

4	DR. GRAHAM:    Okay, yeah, that's what

5  the...

6	DR. ROSENBAUM:    ...to...to the problem

7  that you're talking about.  If you look at

8  the...the...the final equation on the slide, you can

9  see we divided the indoor source by the same factor

10   that we divided the benchmark concentration by.

11	DR. GRAHAM:    So, it's contribution would

12   be the same in the end.

13	DR. SEIGNEUR:    Yeah, it was not clear

14   whether you had corrected the indoor sources, but now I

15   understand.

16	DR. GRAHAM:    Right, it wasn't...it

17   wasn't in the document, and I appreciate the comments.

18	DR. SEIGNEUR:    All right, good, that

19   makes sense.

20	The last major comment I have is

21   regarding the treatment of uncertainty and variability

22   in the document.  When I read the document, my

23   impression was it was a catalog of different sources of

24   uncertainty and variability, and I have two points to

25   make.

1  just...I just had two...two issues, and I guess one was

2  with that whole roll up/roll down thing, but this

3  correction makes me feel better.  So, I...I mean, I

4  think that addresses my...my issue with it.  So, thank

5  you for providing that.

6	The other issue I had here was...and

7  maybe this is my foggy brain or maybe that it wasn't

8  very clearly laid out, but it's this whole concept of

9  the on-road estimation and...and the model that

10   developed...developed the distribution for on-road

11   concentrations.  You know, I...I guess I thought it was

12   a reasonable model, but I didn't quite see why...I

13   mean, you...you generated this wonderful distribution,

14   but I didn't...I was not convinced that that

15   distribution was...was an accurate depiction of what's

16   really happening in on-road concentrations, because I

17   didn't see any comparison to real on-road data, just

18   that it, you know, you had applied this model to come

19   up with that.

20	So...and...and in this, there are some,

21   you know, a handful of monitors in the U.S. that are,

22   if not on-road, at least very near road.  I guess I

23   would have...really would have liked to see some kind

24   of sort of real-world comparisons of your...your model

25   distribution with...with the real...the real data.  It



251

253



1	One is that somewhere missing are the

2  uncertainties due to the model formulation, for

3  instance.  It would be nice to add that and what

4  uncertainty can you expect from a model like AIRMOD,

5  what uncertainty can you expect from your main source

6  models which is the formatting which AIRMOD is applied

7  here.

8	And also what I'd like to see is for the

9  document to go one step further and to give the reader

10   some information on which of the sources of uncertainty

11   you think would be the most important.  Some are going

12   to be more important than others, and I think it would

13   be helpful for the document to...to give some

14   information there.

15	I don't know if you have any plans to do

16   a quantitative uncertainty analysis.  Maybe not, but at

17   least, if it was semi-quantitative that you would say,

18   you know, we expect the results, you know, to be good

19   within a factor of 2 or a factor of 10 and so on.  I

20   think that would be useful.

21	That's the end of my comments.

22	DR. HENDERSON:    Is that?  Okay, thank

23   you very much, Christian.  And Donna?

24	DR. KENSKI:    Thanks.  Mostly, I thought

25   the air quality end of this was...was well done.  I

1  would have made me feel more confident that this was

2  really useful.

3	DR. GRAHAM:    As in the modeling world

4  always, we derive the data from the literature studies,

5  and then, what are we left with after we use all of

6  those studies to derive our little factors.  It's not

7  like we can go back to those same studies and say hey,

8  look, we reproduced those concentrations.

9	Because the factors were derived

10   from...well, I think I was able to obtain 20 different

11   literature sources that contained data on

12   concentrations on-road and at distances from the

13   roadway, and that's what I used to derive these...these

14   factors and, ultimately, you see represented by

15   the...the distributions provided in the TSD.

16	So, to...to compare, I'm at a loss

17   for...for data outside of the papers that I derived the

18   information from, but as a rough cut, I mean, we could

19   see that the on-road concentrations are approximately 2

20   plus or minus...

21	DR. KENSKI:    A factor...

22	DR. GRAHAM:    ...I'm not sure how much

23   higher than the ambient monitor concentrations which is

24   consistent with what is reported in the literature.

25	DR. KENSKI:    But why not...why not



256



1  compare some real ambient measurements?  I mean, yes,

2  you took your literature models, but...but there are,

3  like I said, a handful of very close to the road

4  monitors.  I would have liked to have seen them.

5	DR. GRAHAM:    Right.

6	DR. KENSKI:    Okay.  And...and I...I

7  guess I was then struggling to figure out how those

8  on-road factors then got used in the model.  I think

9  that...that was a little fuzzy about how they got

10   applied later on.

11	DR. GRAHAM:    Right.  In the first draft

12   that's reported in the first draft document, it was

13   applied to all the monitors regardless of...of where

14   they were, and then we pulled out the monitors that

15   were within 100 meters of a major roadway and redid

16   that analysis, and that's what's reported in the

17   technical memorandum.

18	DR. KENSKI:    Right.  That's all.

19	DR. HENDERSON:    So, that's all, Donna?

20   Okay, Tim Larson, are you there on the phone?

21	DR. LARSON:    Yes, I am.

22	DR. HENDERSON:    Okay, it's...it's your

23   turn for comments.

24	DR. LARSON:    Thank you.  I would second

25   Christian's comment about the...the fact that making an

1  offset, you had in there...and I wasn't sure, but a

2  sort of a characteristic length of 10 meters from the

3  roadway, and that seemed low to me, even though

4  it...because that seems like it's in contradiction with

5  the concept of a factor of 2 difference between the

6  monitors and the on-road level.

7	If the gradients are that steep, then

8  I'm surprised, because I've seen...most of the

9  literature I've seen is on...I mean, these sort of

10   characteristic drop-off scales are on the order of 10s

11   of meters or 100 meters or something like that, and it

12   clearly depends on the meteorology at a given time,

13   but, I mean, on average, it seems...it seems like

14   that's a fairly...a fairly steep gradient right near

15   the road, and it would make the estimates on-road

16   fairly unstable...or very sensitive...excuse me...to

17   the actual values that you use at the monitor or where

18   that monitor is within the real gradient, and it seems

19   like that's probably borne out by the technical

20   memorandum where you've taken away those monitors

21   within 100 meters of the road and you get quite lower

22   exceedance values.

23	But I...I don't know whether it's...it's

24   real because...or it's because of that 10 meter

25   assumption, if that's the...unless that's a typo.  It



255

257



1  adjustment that's a constant adjustment across space to

2  bring the model, the  AIRMOD predictions, and the

3  measurements in line is probably inappropriate.  I

4  mean, there's a number of...of literature...there's a

5  lot of literature on this, but...and there are models

6  that are invoked that are somewhere in scale between

7  the AIRMOD and the...the larger chemical transport

8  models.  Urban background model is one of them in which

9  there's at least some attempt made to grid and

10   aggregate those emissions, those...the other 60 percent

11   of the emissions that you're missing from smaller roads

12   and move them around with some simple...simplified

13   photochemistry.

14	And then you see from those kinds of

15   models that there's quite a difference spatially, as

16   you might expect, from, you know, the center of the

17   city versus the surrounding outer edges of the city.

18   And so, a simple adjustment across space is probably

19   underestimating the concentrations in the center of

20   town where most of the...the traffic emission densities

21   are greatest.

22	So, I would...I would second that, and

23   that's one...one comment.

24	I was a little...I wasn't...on the

25   method used for screening the exponential model with an

1  seems small to me.

2	But anyway, moving on, I had a...well, I

3  guess one of the questions had to do with the

4  appropriateness of the locations, and in looking at the

5  Philadelphia data, where people actually live, it's

6  true.

7	Philadelphia is a very...it's a very

8  residential city, and the average census centroid was

9  like 450 meters from a major road which, when you

10   look...fly over Philadelphia makes sense.  There's a

11   sort of central...a fairly small central downtown area.

12	But it seems like that's...that's not

13   representative of a lot of cities.  I mean, it's a very

14   residential single-family city compared to a lot, and,

15   you know, 90 percent of the people in New York City

16   live within 100 meters of a major road, and you're

17   saying...so, I mean, Los Angeles has its own issues

18   with respect to just the large amounts of traffic as

19   well, I mean, but Philadelphia seemed like a curious

20   choice, to me, if you're looking at what you classify

21   as worst case, because when you run the AIRMOD model

22   later on in your...in your APEX model, you're running a

23   fairly...I mean, your centroids are so far from these

24   major roads that it's not clear to me...there have got

25   to be a large number of people that you're...that are



1  living near those roads that you're just

2  underrepresenting completely by this...this kind of

3  a...an approach.

4	I don't know.  I just...I think

5  it's...it would be nice to know something...the

6  statistics...that particular plot, figure 4, I think it

7  is, would be a...would be a...an interesting plot to

8  look at as a screening tool for choice of other cities,

9  because, you know, basically, you do it in the

10   chapter...Chapter 4 or whatever it is in the ISA where

11   you look at percent of people living near roads, but I

12   think that's an important criterion, and I think

13   Philadelphia is...you could prove me wrong, but I think

14   it's a little bit on the end of the distribution where

15   people don't live near major roads.

16	So, that's...

17	DR. HENDERSON:    Tim, you are asking

18   questions, and I think the...the EPA folks here can

19   answer.  Would you like to get a reply from the...

20	DR. LARSON:    Sure.

21	DR. HENDERSON:    Okay.

22	DR. GRAHAM:    Yeah, I just wanted to

23   respond to the...the first question about the 10

24   meters, and that is, I believe, what was reported in

25   the literature that I derived the data from, and if you

260

1  just the representa...representativeness of...of

2  Philadelphia as a test case here from an air quality

3  perspective.  I think it...it's surprise...it surprised

4  me as being labeled worst case city even though it has

5  high monitor values.  Most people don't live near the

6  road.  That's my only other comment.

7	Thanks.

8	DR. HENDERSON:    Okay, and they are going

9  to be doing more cities later.  Within the next

10   version, what cities will be done, all of those on the

11   list or just...

12	DR. GRAHAM:    Do you want to respond to

13   that?

14	DR. JENKINS:    I think we'd be looking to

15   input from the panel on what cities might be

16   appropriate to expand our analysis.

17	DR. HENDERSON:    But how many more do you

18   think you...you can do?

19	DR. JENKINS:    Good question.

20	DR. GRAHAM:    Well, I think it comes down

21   to the value that's added from this analysis.  I think

22   that needs to be reviewed and...and elaborated on

23   before we would proceed, and I think that is one of the

24   charge questions.  Is it charge question 4, I think?

25	DR. LARSON:    I think I agree with you



259

261



1  look at the exponential equation and did calculations

2  at various distances, you probably would see the decay

3  drop off very rapidly, but you bring up an interesting

4  point in that there are studies out there directed to

5  me as well by others, there are studies that show that

6  we do have some of these peak occurrences in the

7  concentrations that occur at greater distances from the

8  roadway, sometimes upwards to 50 or 100 meters.

9	So, we'll take that into consideration,

10   and, actually, in a sense, we had taken that into

11   consideration by removing those monitors, as reported

12   in the technical memorandum.

13	DR. HENDERSON:    Okay, thank you very

14   much.  Continue, Tim.  I just thought you might want to

15   get some answers as you go along.

16	DR. LARSON:    Yeah, no, I...I look

17   forward to the Los Angeles results, because I think it

18   will be an interesting comparison.

19	But I...I think there's this...that's

20   not me, I don't think.  Hopefully not.  Hello?

21	DR. HENDERSON:    Yeah, we can hear you.

22   We heard some static, but...

23	DR. LARSON:    Well, it wasn't me, but

24   I...I don't know.

25	Yeah, so I...my only other concern was

1  that Los Angeles, if you're doing it, would be on the

2  high end of the exposure range.  I just don't know

3  about Philadelphia.  That's my...I mean, if you...

4	DR. GRAHAM:    Well, can I respond to

5  Philadelphia?

6	DR. LARSON:    Sure.

7	DR. GRAHAM:    Yeah, well, because it came

8  out from the air quality characterization as a city

9  that had these...an exceedance or two of 200 ppb and

10   there was significant variability in the hourly

11   concentrations, and it was one of the five that we had

12   selected in addition with made that the first city to

13   be done was that we were able to get the data quickly,

14   the data required for inputs, including the travel

15   demand modeling information which is critical for this

16   analysis, and...

17	DR. LARSON:    Right.

18	DR. GRAHAM:    ...time is...is of the

19   essence here.

20	DR. LARSON:    Yeah, I...and I think it's

21   a great city to do.  I just...I just was providing my

22   own perspective on that.

23	So...and the problem is if you get...if

24   you get cities where people live closer to the road,

25   then you have a more densely urban geometry, and then



264



1  you bring up a whole set of other issues, and it makes

2  this modeling more difficult and...and brings this

3  approach into question, I mean, the AIRMOD approach.

4	DR. HENDERSON:    Okay, I think Ted is

5  next anyway.

6	DR. RUSSELL:    No, I just had...had a

7  question, too.  I'm not sure if Tim is done.  What

8  cities do you actually have the data to do, any of the

9  18 or whatever?

10	DR. GRAHAM:    No, the five that were

11   identified for the exposure analysis.

12	DR. RUSSELL:    So, that's really, those

13   are the...

14	DR. GRAHAM:    Los Angeles...

15	DR. RUSSELL:    ...five...

16	DR. GRAHAM:    Yeah.

17	DR. RUSSELL:    ...that you really have

18   the data to do.

19	DR. GRAHAM:    Right.

20	DR. RUSSELL:    Because it...I mean,

21   as...one of my comments is that it takes an impressive

22   amount of data, and I was, you know, I thought you did

23   a great job on Philadelphia save a few things.  So,

24   I'm...but, really, we're limited to those five unless

25   we really push?

1  needed 17,000?

2	DR. ROSENBAUM:    Well, we selected...it

3  turns out...well, we did Philadelphia County, and it

4  turns out that was like almost all the census blocks in

5  the county, but what...we selected them originally on

6  proximity to major roadways and then being within 10 km

7  of the major point source.

8	So, like I said, we were looking for hot

9  spots.  So, that's how we selected the receptors, and

10   we wanted to do it at the block level to make sure we

11   got the spatial variance the best we could.

12	DR. HENDERSON:    Okay, Tim, are you done,

13   or do you have more comments?

14	DR. LARSON:    Well, I mean, I...yeah,

15   I...I guess that we're too far down the road to change

16   the approach here, but I wouldn't look for hot spots

17   that way myself.  I would...I would look at the traffic

18   and where it's confined by...by the building geometries

19   in the city.  Every city has got...major city has got a

20   central area, and...and that's what I would use as a

21   criteria rather than the proximity of the census block

22   centroid to the...to the road, but, you know...because

23   we're only talking about the extremes in the

24   distribution here.  If we were talking about the means

25   in the distribution, then...then what I just said



263

265



1	SPEAKER:    Yeah.

2	DR. HENDERSON:    Okay, yeah, that's...I

3  wasn't sure what you were asking, whether you wanted

4  more than the five or whether you can only do two more

5  and you wanted to know which if the five to do, but if

6  you can do those five, I think that would be good.

7  That's my opinion on it.

8	DR. LARSON:    I think that's right.

9	DR. GRAHAM:    Well, because we have the

10   input data, again, the issue now is...is about the time

11   when you think about the size of the...the domain.  The

12   modeling domain that we did for Philadelphia was 17,000

13   receptors.  For something like Los Angeles, it's going

14   to be at least twice that, probably longer, and I don't

15   think the computational resources used is going to be a

16   linear function.

17	DR. RUSSELL:    Of the AIRMOD computations

18   or APEX or what are you...

19	DR. GRAHAM:    I think all of it combined.

20   I...I think it's just going to take longer, much

21   longer.

22	DR. ROSENBAUM:    Yes, I think it would be

23   good to get your input on like prioritizing among that

24   list.

25	DR. HATTIS:    Why did you decide you

1  probably doesn't matter, but if you're only talking

2  about 1 percent of the population getting high

3  exposures, then these extreme cases do matter.

4	And so, it's just...it's an uncertainty,

5  because you're not going to be able to do that in this

6  current approach, and you don't have enough data to use

7  a model...a measurement based approach to get a

8  different hot spot identification procedure.  So, I

9  would...I guess I...I guess I'm just glad you're going

10   to do more cities, but I don't know how you...well,

11   anyway, we'll wait and see.

12	DR. HENDERSON:    This is the time to give

13   input.  I guess let's hear what Ted Russell has to say.

14   I mean, I want to hear what everybody has to say, and

15   then we can discuss, you know, is this the right

16   approach, or do...do others think that there's a better

17   way to do it?

18	So, Ted, why don't you go ahead?

19	DR. RUSSELL:    First, I definitely want

20   to second what Christian, Donna, and Tim have brought

21   up.  If you want my choice, it would be...since you've

22   already done Philadelphia, it's in the bag.  Then, Los

23   Angeles is the ob...obvious second choice.  I'm from

24   Atlanta, so that must be third, but also, I think

25   geographically, it's...it's different.



268



1	I'm not so worried about Phoenix, since

2  it's, you know, not that far from L.A.  It's a similar

3  kind, but it's good to know that, actually, it's really

4  the data that's limiting it, and so, that is our

5  choice.

6	And I will say I was very impressed with

7  the detail that went into this, and I recognize that it

8  is a...a significant effort to...to get that amount of

9  data and to pull this off.  So, definitely, kudos on

10   that part of it.  You know, I'd pointed out that maybe

11   there's...there's some questions on the data.

12	And then, the other point I'll just

13   bring up is I'm actually quite uncomfortable, the more

14   I think about it, with how you did the on-roadway

15   concentration adjustment with the exponential decay.  I

16   mean, when I was reading it, I was going it would

17   strike me that this should be done in some sort of a

18   Gaussian modeling framework, even if it's a very simple

19   one, I mean, just a back of the envelope Gaussian

20   modeling, one where you also include a chemical

21   conversion.

22	With your current formulation, NO2

23   concentrations can only go down from the roadway even

24   though, in fact, looking at some data that Joe...Joe

25   had at break, you know, the NO levels are much, much

1  on road or near road.

2	DR. RUSSELL:    Okay, so that...I would

3  have to look in more detail at the studies, but what

4  you're saying is that in those studies, they measured

5  where the peak was anywhere near a road?

6	DR. GRAHAM:    Well, they...that...from

7  the studies, it was on road, but I'm suggesting, based

8  on your comments as well as others, that okay, if this

9  peak is occurring somewhere at a distance from the

10   source, it's going to have that decline, and I've seen

11   that reported as well.  So, what we're capturing with

12   the calculation is basically the peak.  Where it

13   actually occurs, that's really a question here.  Is it

14   on the road, or is it near the road?

15	So, rather than characterizing what I'm

16   saying as on road, maybe I should characterize it as on

17   or near roadway.

18	DR. RUSSELL:    And from what you've done,

19   have you actually seen, essentially, a Gaussian

20   dispersion with some conversion rate gives you about

21   the same thing for a busy roadway, or is that even

22   pertinent to what you're trying to do?

23	DR. GRAHAM:    I...I haven't seen anything

24   like that, so if you had something to share with me...

25	DR. RUSSELL:    Oh, I was just hoping you



267

269



1  higher on the roadway, right on the roadway, than the

2  NO2, and there's a variety of data that shows that the

3  peak is actually somewhat further away from the

4  roadway.  So, I'm...if I were to go and really look at

5  one part to improve it on the air quality/modeling

6  characterization, that's where I would probably spend

7  my effort.

8	I'm not as worried about it, given that

9  I'm not sure how that data's going to be used in the

10   end, except it sort of gives you a characterization of

11   it's a factor of 2 or whatever, but it would be of

12   interest to see...get a little bit better support for

13   that and to also show that it...it may not be on

14   roadway.  It may be someone sitting 50 meters away or

15   something like that, and this could be particularly

16   important in a place like Los Angeles.

17	So, I might...I would hope or at least

18   like to see that that might be one of the changes

19   between now and the next draft of this.

20	DR. GRAHAM:    Well, in a sense, you could

21   look at the concentrations that have been estimated and

22   characterized as on road.  That's really the peak

23   concentration that's being estimated, whether it

24   occurred on the road or somewhere at a distance from

25   the road.  I mean, you could characterize it as either

1  had done it.

2	DR. GRAHAM:    Oh, no.

3	DR. ROSENBAUM:    Actually, we did do that

4  with the dispersion modeling when we did do an analysis

5  of the ratios from the...the concentrations at the...at

6  the blocks to the concentrations on the roadways,

7  because we put receptors on the roadways as well.  So,

8  we did get a distribution there of the...of those

9  ratios.

10	So, it's somewhat of an independent or a

11   Gaussian...

12	DR. RUSSELL:    And what did you find?

13	DR. ROSENBAUM:    ...estimate.  I can't

14   remember off the top of my head, but I can tell you

15   that tomorrow.  I can look it up tonight.

16	And that did...that did include the

17   chemistry, too, the ozone limiting chemistry.

18	DR. HENDERSON:    Okay.  Is that yours,

19   Ted?  Okay, Jim Ultman, do you have something to add?

20	DR. ULTMAN:    Several of my comments have

21   already been covered.  I guess the mo...the thing that

22   was most on my mind was the...or I had most of the

23   problem with was this extrapolation method for ongoing

24   conditions.  I really couldn't follow it in the tech.

25   This is out of my field, so I'm probably a good...I may



272



1  or may not be a good test case.  Depends on how you

2  view the document who your readership is going to be.

3	I had a really hard time understanding

4  something which I think is pretty simple, but I think

5  it needs to be cleaned up somehow, and I...we can talk

6  about this more privately if you want.  And, also, the

7  methodology that was applied was hard to follow.  I

8  mean, the equations themselves and what the variables

9  exactly meant and what you were getting at I didn't

10   understand, but then, when it came to the methodology,

11   I had a hard time, although going to the...going to the

12   annex helped a little bit, because it has some of the

13   distribution information that the...that's not present

14   in the main part of the document...not the annexes, the

15   PSD.

16	DR. GRAHAM:    Sure.

17	DR. ULTMAN:    So, one of the suggests I

18   have is...is...I made it this morning, was that it

19   would be good to move some of this up, possibly, into

20   the ISA, and I know that the two groups communicate on

21   this, but I think if you could some...put some of the

22   basic data from which the onward extrapolations are

23   derived in the main body and some of these

24   distributions in the main body, it might...of the ISA,

25   it might help when you get to this point.

1  extrapolation data versus what you get in the exposure

2  analysis.

3	So, have you looked at putting those two

4  together?  In other words, you've got on-road

5  extrapolations that you've done for the air quality,

6  but you also have, you know, you've simulated spatial

7  distributions in your exposure.  Now, can't those come

8  together?  Can't you validate one with the other

9  and...and lay some of these things to rest?

10	DR. GRAHAM:    Well, I think it's close.

11   It's probably one of these apple and oranges things,

12   because I think it's going to be a...a function of the

13   distance the receptors are to the roadway that are used

14   to generate the distribution.

15	DR. ROSENBAUM:    And we did get a

16   distribution.  I mean, I think...I think your point is

17   well taken that it would be good to compare them.  They

18   aren't an exact comparison, but just to see if we're in

19   the same ball park.

20	DR. ULTMAN:    Give more credibility to

21   both methods if they...if you could show some

22   consistency.

23	The other thing I was...this is more of

24   a, I'd say, an academic curiosity than...that something

25   that you can do something about at this point, but



271

273



1	And there's a, I think, a logical place

2  for it in Chapter 2 when they're talking about spatial

3  variations, that a section could be put in on that.

4	And maybe even put the equations in and

5  explain the basis of the equations.  I don't know if

6  they're theoretically based or purely empirical or what

7  the assumptions are, but, certainly, there are exposure

8  equations in Chapter 2 of the ISA.  I don't see why

9  there couldn't be equations that have to do with the

10   air quality in there as well.

11	One thing I was wondering, I was just

12   curious about, and I think you were...you actually kind

13   of touched on it.  The...I'm saying you, because I

14   don't remember your first name.  I'm sorry.

15	DR. ROSENBAUM:    Arlene.

16	DR. ULTMAN:    Arlene.  One thing that

17   Arlene touched on was that you had actually gotten some

18   information from ratios of off-road to on-road

19   concentrations from your simulations, exposure

20   simulations, and I was thinking that, you know,

21   somebody else had made the comment that well, you know,

22   isn't there a way...can't we use some kind of data to

23   validate the extrapolation method, and somebody else

24   made a comment, is there really consistency between the

25   extrap...between the values you get from the

1  the...does the CHAD take...have...take account at all

2  in terms of avoidance behavior of people that have

3  compromised lung function?  In other words, if it's a

4  bad...if it's a bad ozone day, in my field, or if it's

5  a bad NO2 day in your field, they may choose to change

6  their activity patterns.  Is that in the CHAD database

7  at all?

8	MR. RICHMOND:    Yeah, this is Harvey

9  Richmond.  Not that information.  It is...they didn't

10   collect what the ozone or NO2 levels.  They caught

11   the...we matched up later temperature with it, but so,

12   to the extent in the whole large database of thousands

13   of activity days, there may be some...there certainly

14   would be some days that included high ozone and others

15   that didn't, but we don't have access to that

16   information.  It's not a variable that's captured

17   within CHAD.

18	DR. ULTMAN:    Because we did...we

19   did...there was some discussion about the...the

20   validity of doing, you know, the roll back in the...in

21   the ambient air levels as a way of dealing with the,

22   you know, trying to make your computations more

23   efficient so you don't have to repeat your computations

24   as often, and that relied on linearity of all the

25   models for that to be true.  I mean, you gave us the



276



1  equation here, and it's a linear equation, but if there

2  are parameters within that equation which depend on the

3  air quality, then you can't really do what you're

4  doing.

5	You see what I'm saying?  So, in other

6  words, if there are parameters in the...if there are

7  parameters in the time activity, sub-routines that you

8  use in the CHAD, that were sensitive somehow to

9  concentration, that would make the models non-linear,

10   and then you couldn't do that.  There wouldn't be thus

11   equivalence that you're talking about.

12	So, as I said, if it's not in the CHAD

13   model, there's probably nothing you can do about it at

14   this point.

15	MR. RICHMOND:    Right.  In fact, we are

16   supporting additional research at this point to

17   actually carry out those kind of surveys to better, in

18   the future, get information on how much individuals do

19   avert any mitigating behavior on high ozone days for

20   sensitive populations, but it's research that doesn't

21   exist, for the most part, out there right now.

22	DR. HENDERSON:    Okay, thank you, Jim.

23   Now, it's open for other people who may have comments

24   on the air quality part of the risk assessment

25   document.  Whew, we've gone all the way down the line,

1  me ask the question was the population you gave

2  for...for...or somebody gave for Philadelphia was the

3  residence population, but the real population is the

4  residents plus the people who come in.  In the case of

5  Atlanta, I was just looking on the web, and it's like

6  the population of Atlanta grows by 60 percent during

7  the day.  So, that...that wasn't considered, and, of

8  course, those commuters are getting the...the most

9  exposure to ambient air pollution, so they're...they're

10   probably a big chunk of the problem.  Right?

11	DR. GRAHAM:    Yeah, in this model,

12   we're...we're not including people that come from

13   outside the domain as well as individuals that leave

14   the domain.  They end up with average concentrations

15   when they commute outside the domain.

16	DR. THURSTON:    So, you're not counting

17   people who live in the suburbs and drive into the city?

18	DR. GRAHAM:    Well, it's the whole

19   county.

20	DR. THURSTON:    It's the whole county.

21	DR. GRAHAM:    So, if there are suburbs in

22   Philadelphia and they're traveling to the inner city

23   region, then they are counted.

24	DR. ROSENBAUM:    If they live in

25   Philadelphia County.  We just...we only modeled



275

277



1  so, George, we'll come up this way, George and then

2  Lianne.

3	DR. THURSTON:    I'm just wondering if the

4  model considers journey to workinformation?

5	DR. GRAHAM:    Yes, it does.  It has a

6  commuting database that's derived from the U.S. Census,

7  and people have a home tract...I'm sorry...a home

8  block, and they also have port blocks.

9	DR. THURSTON:    And does that apply just

10   to the people who live in the city or the commuters

11   who...

12	DR. GRAHAM:    It's the entire modeling

13   domain.

14	DR. THURSTON:    The entire, so...oh.

15	DR. GRAHAM:    All the simulated

16   individuals...

17	DR. THURSTON:    I'm sure it's an answer.

18   I don't understand it, though.

19	DR. GRAHAM:    If they were characterized

20   as...as workers, they link to different blocks.

21	DR. THURSTON:    So, but workers in there,

22   they are counted.

23	DR. GRAHAM:    Yeah.

24	DR. THURSTON:    Oh, okay.  Because, I

25   mean, like Atlanta's population...the thing that made

1  Philadelphia County.

2	DR. THURSTON:    But don't you think that

3  you might be missing a lot of people who are affected

4  by the pollutions...the pollution in Philadelphia?

5  You're missing a lot of people who might be exposed to

6  that, because you're not counting people who live

7  outside and come in for the day, because the maximum

8  concentrations are in the daytime, and the...and along

9  the roadways where they're driving, and I don't know,.

10   It seems like...

11	DR. GRAHAM:    Then you expand...you have

12   to expand the model domain, and then you have to

13   include the sources where those individuals are living.

14	DR. THURSTON:    Well, yeah.

15	DR. GRAHAM:    Right, but, I mean, you

16   have to draw the line somewhere, I guess is what I'm

17   trying to say.

18	DR. THURSTON:    Well, okay, but if you do

19   Atlanta, then you're drawing a line...you're

20   eliminating a lot of people.  Philadelphia's probably

21   not...based on these numbers, Philadelphia's not too

22   bad, because only like 10 or 15 percent of the people

23   in the city come in from outside, but you go to

24   Atlanta, there's a big chunk of people you'll be

25   missing.  Anyway, it's just a thought.



280



1	DR. GRAHAM:    Well, again, it depends on

2  the boundaries for the domain.  If we run the model and

3  we see that there are so many people that are leaving

4  and so many people that are not being counted that are

5  coming in, then that could be a problem.

6	DR. HENDERSON:    Jon, did you have your

7  hand up?  I...okay, Jon has...I know you have to leave

8  early, so let's get your comment while you're still

9  here.

10	DR. SAMET:    I mean, I had this one

11   general comment in my written remarks that it would be

12   really helpful if there was some sort of overview of

13   this cascade of models and assumptions so that readers

14   could approach this.  I read this several times, and

15   really struggled to kind of get the big picture,

16   because you immediately sort of leap into the details

17   of this model and that model, and this is all

18   strung...strung together, and I think you need a...sort

19   of a start-to-finish picture somewhere up front to help

20   the reader struggle with this...or not struggle so

21   much.  The readers will still struggle, but they won't

22   struggle quite as much but something up front.

23	DR. HENDERSON:    Thank you, Jon.  Now,

24   we're coming up the road.  Did Ed have his hand up?

25   No.

1  five, or you've done work on that, but does that

2  selection of city site, especially for California, do

3  you have problems with actually measuring accurately

4  NOx concentrations due to, perhaps, higher levels of

5  nitric acid in those areas?  Is that something that has

6  to be taken into account in...in this modeling, or is

7  that just not relevant?

8	DR. GRAHAM:    I...I'm not prepared to

9  respond to that.  I apologize.  I'm not an atmospheric

10   chemist.

11	DR. LARSON:    This is Tim Larson again.

12   If you...I mean, Christian probably...Ted knows this

13   better than I, but if you look at the distribution

14   large scale of NO2 across the Los Angeles Basin,

15   there's a considerable gradient from the coast at low

16   levels eastward towards Riverside on a very large

17   scale, and it's...it's...it's not easily represented by

18   even aggregating the traffic emissions.  You have to

19   run a more sophisticated analysis.

20	And in that particular case, if you

21   don't account for the...those large-scale gradients in

22   NO2, you can be spectacularly wrong, because

23   they...they overwhelm the near-road gradients in the

24   eastern part of the basin.

25	DR. ROSENBAUM:    So, that may be a



279

281



1	DR. POSTLETHWAIT:    For once, no.

2	DR. HENDERSON:    Okay.  Lianne, then.

3  I'm just kind of coming up the road here.

4	DR. SHEPPARD:    Yeah, I'll make more

5  comments tomorrow, but I wanted to emphasize what has

6  already been said, is that you're focusing on the

7  extremes of the distribution in counting the

8  exceedances.  So, in all of this modeling and all of

9  this data analysis, you have to get the extremes right,

10   or you're not going to be counting it right.  So, it's

11   not anywhere near as easy a problem as it typically is

12   where we're trying to get an average.  We need to get

13   the extremes right.

14	And...and there's really no information

15   in the document to assess that, and there's lots of

16   reasons to believe that they're not right, and,

17   therefore, we're undercounting.  So, I'll make this

18   point again tomorrow, because that's basically what my

19   comments are, but I think that's probably the key point

20   about predicting exceedances, is that you've got to get

21   the extremes right.

22	DR. HENDERSON:    And Kent?

23	DR. PINKERTON:    This is just a point

24   that I'm just curious about.  You'd mentioned that in

25   the selection of cities that you've identified, perhaps

1  particular problem of that city.

2	DR. LARSON:    I think it's more acute

3  there than in most areas.

4	DR. HENDERSON:    Okay.  I don't...oops.

5  Dale, did you have your hand up?  I didn't remember.

6	DR. HATTIS:    I...I have a...well, I'm

7  going to have a lot to say tomorrow, but just to...

8	DR. HENDERSON:    Now, when you say

9  tomorrow, are you talking about for air quality?

10   Because tomorrow, I'm thinking we'll go to exposure and

11   health effects.

12	DR. HATTIS:    Yeah.

13	DR. HENDERSON:    Okay, okay.

14	DR. HATTIS:    But because of the linkage

15   of...I got this in your notice that the table 26 for

16   the comparison of the three monitors was not

17   entirely...

18	DR. HENDERSON:    Get close to the mike.

19	DR. HATTIS:    ...was not entirely

20   unambiguous, let's say, first, are those the only three

21   monitors for which you've done comparisons?

22	DR. GRAHAM:    Yeah, those...those are the

23   only three monitors that are within Philadelphia

24   County.

25	DR. HATTIS:    Oh, is that right?  So,



284



1  those were the only three monitors that were available

2  to you, in fact.

3	DR. GRAHAM:    In Philadelphia County,

4  right.  There was another table in there that was in

5  error.  It had like ten monitors or something like

6  that.

7	DR. ROSENBAUM:    The ten monitors

8  were...were for the whole CMSA, but we didn't model the

9  whole CMSA.

10	DR. HATTIS:    Okay, and what I saw was

11   just one number per year per monitor which suggests

12   that it's...what you're comparing is long-term

13   averages.  Is that right?

14	DR. ROSENBAUM:    Yes.

15	DR. HATTIS:    I mean, it seems to me you

16   could have compared the distribution of hourly

17   concentrations that was observed at the monitor with

18   the distribution of hour...of hourly concentrations

19   that was predicted for that monitor.  Then, that would

20   be a different kind of comparison.

21	So, I mean, I think that would be

22   something that would be, I think, of interest, but I'll

23   say that later.

24	DR. HENDERSON:    And Ed?

25	MR. AVOL:    So, I have a couple of

1  that if you don't have some credibility and confidence

2  that they relate well at the outset, then all these

3  adjustments and everything are...sort of put you

4  further away from the...from the truth, and it's hard

5  to believe where you're going.

6	DR. ROSENBAUM:    Well, as far as the off

7  road, I guess we assumed that that would be more

8  dispersed, just like the local traffic, so that it

9  wouldn't be necessarily creating hot spots, and we

10   were...we were really trying to just focus on that.

11	As far as the shipping is concerned, we

12   did look into doing that, but we didn't really have a

13   good emission inventory for the port there.  So, it

14   was...that was like a data limitation to looking at the

15   port.

16	MR. AVOL:    And rail is not part of the

17   equation?

18	DR. ROSENBAUM:    The rail...the rail was

19   also a data limitation.  We did have the...the NEI has

20   the county level estimate for rail...for rail yards,

21   but there were several rail yards, and then we didn't

22   know how to...we didn't have the information to

23   allocate, spatially allocate among the rail yards, so

24   that was another data limitation issue.

25	MR. AVOL:    And I thank you for the



283

285



1  questions.  In fairness and disclosure, I should say

2  that conversations I have with air quality modelers are

3  sort of like an audience with the great and powerful Oz

4  in that it seems like a lot's going on behind the

5  curtain that I don't quite see or understand.  So, I

6  just...let me ask you some questions about what's going

7  on in front of the curtain, and this gets at some of

8  the questions that I think Jon Samet raised about

9  visibility and credibility and confidence about a

10   bigger picture about what's going on.

11	And so, one question or...or statement

12   that you could consider is in terms of the inventory

13   comparison, you talked about using some fugitive source

14   information and...and on-vehicle...on-road vehicle and

15   stationary sources and airports, I guess, the airport

16   information.

17	And so...so, relating from Los Angeles

18   and the California experience, it seems like the other

19   large sources such as off-road construction, chipping,

20   et cetera, I don't know how important that is in

21   Philadelphia per se, but the question would be how

22   the...your inventory compared, when you just used that

23   subcategory that you used, how that compared with the

24   AQMP for the area or the inventory, the best emission

25   inventory, that exists, and if that...it seems to me

1  comments about the major road assignments and this

2  issue about, you know, not being able to deal with

3  smaller category roads does raise some issues, and,

4  again, when you get to Los Angeles, it will be an

5  issue, although if you go down to 15,000 vehicles, it

6  will cover a lot of roads in Los Angeles.

7	But I do have a question with regard to

8  your model, and this gets some...some at what Christian

9  raised which is how well it fits and...and getting to

10   your concern about trying to identify hot spots as to

11   with air modeling and the adjustments you make, whether

12   you take into account topology or land use or other

13   issues like this which may well change the character of

14   where those hot spots show up.

15	DR. ROSENBAUM:    We did for the...when we

16   modeled the elevated sources, we did use terrain

17   elevation.  We didn't for the roadways, because

18   it's...for ground level sources, it's terrain

19   following, so it didn't really make a difference.  It

20   wouldn't...it shouldn't make a difference there.

21	MR. AVOL:    Okay.  And then, one comment

22   with regard to boundary domains, talking about

23   Philadelphia County.  I think Los Angeles is an

24   important area to take, but...but because of this issue

25   that George raises about commuter traffic, in Los



288



1  Angeles, that's a big issue, and I'm not sure what the

2  domain is going to be when you characterize that, but

3  it's...it's a...a key ingredient to that is, you know,

4  commuter traffic daily into and out of the...whatever

5  the central area is that you're going to talk about

6  defines Los Angeles.

7	Finally, I just want to comment on...on

8  the table in terms of of...there was some comment about

9  uncertainty and analysis, and I appreciated table 16 on

10   page 65 in terms of talking about the different sources

11   of uncertainty in the air quality measurements and how

12   to represent.  I think it's helpful to identify that,

13   but I found it sort of frustrating that it sort of said

14   none, a little, or a lot, sort of, and the nature of

15   the magnitude of the errors.

16	And so, I was sort of left well, I

17   appreciated it identifying the sources, but I wasn't

18   quite sure what to do with this.

19	DR. GRAHAM:    Right, I...I appreciate

20   that comment, and I would say that there needs to be a

21   little more rigor there.

22	DR. HENDERSON:    Okay, are there people

23   on this...on my blind side?  Yes, Ron?  We'll go down

24   this whole...oh, I'm so sorry.  Terry?

25	DR. GORDON:    I guess the modeling by Oz

1	DR. WYZGA:    I have three comments.

2  First of all, I do want to commend you.  I think this

3  is an awful lot of work here, and I was very impressed

4  with the quality of the work.  I think you...I think

5  it's...I really appreciate it.

6	First comment was when I think of just

7  meeting the standard, it appears that what you've done

8  is, in a sense, created an upper bound with a number of

9  exceedances with rollup never being used, and I

10   wondered if it might make sense to come up with a range

11   estimate, and that would be the upper end, and the

12   lower end might be...it essentially would equate to the

13   as-is scenario, because it essentially would say that

14   where you've exceeded the standard, you would lower it

15   so that the standard is satisfied otherwise as is, and

16   if a range estimate might be, in fact, more realistic

17   than what you presented.  That...that was my comment.

18	And then, the second question was, I

19   guess, motivated by what Lianne said.  Are you really

20   trying to portray an extreme case, or are you trying to

21   portray something that's representative?  And...and

22   it's unclear in my mind.

23	DR. GRAHAM:    For which analysis?  I

24   mean, if you're considering we're estimating exposures

25   using exposure modeling under situations where



287

289



1  has...has confused me, but not it's enlightened it.

2  The figure we have in the pie chart when you have in

3  the benchmarks of 200, 250, and 300 on page 117, to get

4  up to 300 ppm, the person is barely sleeping and

5  spending almost as much time outside in the parking lot

6  to get the 300 ppm.  So, given all the questions from

7  this table, I wonder about the modeling.

8	DR. GRAHAM:    Right, no, I...I think it's

9  maybe not explained clearly in the legend.  This...this

10   has to do where those peaks occurred.  It's not

11   suggesting that they spent more time there.  That's

12   where they ended up with the peak occurrences.

13   That...that's what it's speaking to.

14	DR. GORDON:    Okay.

15	DR. HATTIS:    The peak hour was located

16   differently.

17	DR. GRAHAM:    Right.

18	DR. GORDON:    But it's this fraction of

19   time all simulated persons in Philadelphia County spend

20   in these 12 environments.

21	DR. GRAHAM:    And receive that exposure

22   above the particular health benchmark.

23	DR. HENDERSON:    So, you might clarify

24   the legend if...not clear.  Okay, then we go down.

25   Ron?

1  conditions exist as is, I think we're trying to

2  represent exposures as they exist for the year 2001

3  through 2003 in...in that area, but then when we

4  simulate the standard, that's where it changes.  I...I

5  guess I'm not clear about what the question.

6	DR. JENKINS:    Well, one issue that you

7  might...that might not be totally clear is the purpose

8  for simulating just meeting the current standard, and

9  one of the things that we try to do in the review from

10   the big picture perspective is we try to evaluate the

11   level of public health protection that's associated

12   with an area that would be just meeting that standard.

13	So, you know, when we get into talking

14   to our management, one of the things they're going to

15   ask us is well, what happens at the current level of

16   the standard?  And so, we do these types of simulations

17   to...to look at what might happen if an area just meets

18   that standard.

19	So, we're really just looking at the

20   level of public health protection.  It wasn't...and I'm

21   not sure.  I think this could be sort of a source of

22   confusion for you which...is that...does that offer any

23   clarification for...

24	DR. WYZGA:    It...it does, but it

25   just...it's...it just seems very unrealistic, and you,



292



1  in fact, say so in the document.

2	MR. RICHMOND:    Right, we certainly have

3  said we will make it clear in the document that it is a

4  hypothetical situation for what protection there would

5  be if you just met the standard, and there's no

6  intention to try to do risk reductions from that.  In

7  other words, once we get to the second phase of

8  alternative standards, we're not going to look at risk

9  reductions from the current standard to these

10   alternative standards, but it is what protection you

11   would have if you just met the standard.

12	Doing an as is which is well below all

13   the areas we've looked at and...and in the country are

14   well below, fortunately.  We've had a lot of progress

15   on NO2 reduction...NOx emission reductions.  To say the

16   lower bound is the as is situation, that would not be a

17   lower bound for air quality that would be allowed where

18   you'd just meet the standard.

19	And so, I didn't agree quite with that

20   being a lower bound.  We're certainly presenting as is

21   and will characterize those levels are considerably

22   below the current annual standard.

23	DR. SHEPPARD:    Can I speak up for a

24   second?

25	DR. HENDERSON:    Sure, go ahead, Lianne.

1  difficult, we agree.

2	DR. SHEPPARD:    Extremely.

3	DR. HENDERSON:    Okay.  Jim?  James?

4	DR. CRAPO:    I still have a question that

5  I just don't...or maybe a comment, but it seems like

6  the key thing we're trying to get at in this modeling

7  is sort of illustrated in figures 9, 10, and 11 where

8  you're looking at number of people at risk for a

9  certain exceedance, and as I read this and as I listen

10   to all the comments, one of my questions that I'd love

11   to see estimated in here is if you can find any way to

12   express the confidence interval we have around that.  I

13   know you can't calculate a real confidence interval,

14   but the...but it's the kind of thing Ed was bringing

15   up, that the level of accuracy needs to be somehow

16   incorporated into this more than just in the paragraph

17   that says we had all these assumptions.

18	And it's important for a couple of

19   reasons, because I think when people see these tables,

20   there's a tendency to say well, that's what it is, and

21   now you multiply that and count the number of people

22   that I can do something with health on, and...and some

23   expression of the confidence interval would be helpful.

24	And the other thing it might really do

25   for you is help you understand how many cities you want



291

293



1	DR. SHEPPARD:    I just wanted to follow

2  up, because the..the summarization of the data is based

3  on exceedances at the hourly level.  So, in that sense,

4  it's...it's definitely about the extremes of the

5  distribution, because you're looking at the number of

6  high values that you're simulating.

7	So, you know, while...while you may be

8  focusing on meeting the stand...or, excuse me, as is

9  that's well below the standard, that's an annual

10   average, but everything you're summarizing is based on

11   hourly data and tallies of exceedances, and when you're

12   getting, you know, for thousands of measurements,

13   you're getting one and two numbers, you know, numbers

14   of that size, you're talking about the extreme of a

15   distribution very far out on the tails.

16	MR. RICHMOND:    Right, because we're

17   trying to relate it to...this is just relating for the

18   clinical based health endpoints where the effects are

19   on that scale.  So, I mean, that...that's why.  That's

20   the motivation for that.

21	DR. SHEPPARD:    Right, but it

22   becomes...the modeling exercise changes dramatically

23   when you focus on the tails than when you're focusing

24   on the mean.

25	MR. RICHMOND:    Sure, much more

1  to...want to study, because you've got to estimate

2  whether you could really improve that with ten cities

3  versus two or three.  I mean, once you've done

4  Philadelphia and Los Angeles, if our confidence

5  intervals are just huge around these things, you

6  maybe...you may have made your point, and there may not

7  be any value in doing more if your air margins are too

8  large, and it kind of sounds like they might be from

9  what I'm hearing on all the assumptions.

10	I just throw that out, as I'm curious

11   how you...how you respond to that or could you figure a

12   way to express something of that nature.

13	DR. GRAHAM:    It's something we are going

14   to investigate how, because we do take those comments

15   to heart, and it's not something we've finalized right

16   now.

17	DR. CRAPO:    Yeah, I don't have any

18   concrete way to do it myself, but I...but I think it

19   would probably lead me to decide you want to do...it's

20   really valuable to see what this tells you, but you

21   might not get a lot more information after you've done

22   it a couple of times.

23	DR. GRAHAM:    Yeah, we've had some

24   discussions recently about some sensitivity analyses

25   that we can perform to try and get at that number



296



1  you're looking for, the...these intervals.  So, we're

2  going to look into that.

3	DR. CRAPO:    Okay.

4	DR. HENDERSON:    Okay.  I don't see any

5  more hands.

6	One thing I want to ask the people from

7  the Air Office.  I chose, because of the number of

8  charge questions you had listed under three different

9  areas, to just look at one area at a time.  If...in

10   looking through these charge questions for the air

11   quality information, it seems to me we've addressed

12   most of them, but I'm going to ask you, is there

13   anything in here that you wanted to know from this

14   committee that we have not addressed?

15	DR. JENKINS:    I don't think so.

16	DR. HENDERSON:    One thing, Christian,

17   you're going to be the one who's summarizing all this,

18   and there are six charge questions.  I listened, and I

19   think they were all addressed in some way, but because

20   we had so many people commenting on this other than

21   just the...the few that were assigned to look at it, I

22   think Christian would probably enjoy your sending, you

23   know, your comments, what you think is really critical

24   to him so he can develop a paragraph.

25	Now, you tell me, Christian, would that

1  part of this document.

2	DR. NUGENT:    And, Rogene, if I might

3  add, could everyone also cc me on that so I can have

4  a...be a switchboard?

5	DR. HENDERSON:    Definitely, cc it to

6  Angela.  I mean, you want to send it to Angela and cc

7  me, because she is the one who will be pulling them

8  together.

9	As far as the...your pre-meeting

10   comments, everyone should know that whatever letter

11   goes to the Administrator, we attach your comments

12   unedited, too.  So, if you want to change anything that

13   you've written in the pre-meeting comments, just change

14   them and send them to Angela, and when we do send it

15   in, that's what will be attached.

16	You have quite a bit of time for that,

17   because it will take us a while to come to...I mean,

18   the last date on this, is what, May something.  May

19   what?  You all have the thing, but I just want to know

20   if you...your pre-meeting comments, if you want to

21   modify them, modify them, and your individual comments

22   always go with the letter, and they're very important.

23   They contain a lot of detail that we don't have the

24   time to discuss in the meeting, so be sure you do that.

25	And, okay, we've got something from...



295

297



1  be helpful to you?

2	DR. SEIGNEUR:    Yeah, that would be

3  useful if people send me an email with their comments.

4	DR. HENDERSON:    Okay.

5	DR. SEIGNEUR:    My email address is my

6  last name at aol.com.  You can do that tonight.

7	DR. HENDERSON:    Yeah, Ted?

8	DR. RUSSELL:    If I might, actually, I

9  think all of us who are writing up any part, I mean,

10   certainly, I would like the same for the participants

11   on my ques...my part on the ISA.  So, I think just...I

12   don't know.  I wouldn't mind getting them tonight, if

13   possible.

14	DR. HENDERSON:    That has happened in the

15   past where we've done it overnight, and then we...and

16   it really facilitates things.  So, if you have

17   something for a lead discussant on any of these issues

18   that would be helpful in them developing the paragraph,

19   go ahead and get it to them like tonight, if you can,

20   because it can...now is the time when they can be

21   working on their laptop and...and doing their thing.

22	So...and I really like the way we did it

23   when we just did it overnight.  So, why don't we try to

24   do that, and we will develop as much of our writing as

25   we can, covering both the first document and the first

1	DR. GRAHAM:    Thank you, Rogene.  There

2  was one thing on the air quality characterization.  In

3  the comments, I think it was Tim Larson had suggested

4  that we review some of the literature on street canyons

5  to improve the exposure estimations.  I mean, in a

6  sense, what we've done has not necessarily taken those

7  into account.  So, to derive some sort of a

8  relationship between the street canyon exposures and

9  that would...would occur on a typical roadway, I...I

10   would be interested if he could refer to me to some of

11   those...those citations, that would be wonderful.

12	DR. LARSON:    I can't hear.  You're

13   breaking up.

14	DR. HENDERSON:    Tim, did you hear that?

15	DR. LARSON:    I didn't hear the last

16   part.  Sorry.

17	DR. GRAHAM:    Yes, I was sug...yeah,

18   there you go.  I was requesting if you had some

19   specific citations for the street canyon information as

20   far as a relationship, that would be wonderful.

21	DR. LARSON:    Yeah, I think there are

22   four of them in my written comments, some recent

23   summaries, some recent papers.  Look at those as well,

24   and I can...I can send something on to a Angela for Ted

25   as well.



1	I mean, there's this really nice...I may

2  have mentioned it before, but there's this really nice

3  Danish THOR system, it's called, air GIS that has a lot

4  of supplemental information but is an operational model

5  that's already implemented in numerous cities

6  throughout Europe.

7	DR. HENDERSON:    Okay, and as I said, try

8  to get your...your comments in to the lead discussants

9  tonight, if at all possible, and we meet at 6:15 if you

10   want to go eat with us.  And we start tomorrow morning

11   at what time?  8:30, and we do have our breakfast.

12   Right?

13	SPEAKER:    Can we leave things in this

14   room overnight?

15	DR. NUGENT:    Yes, it will be locked up.

16   I wouldn't leave your laptop materials.

17	DR. HENDERSON:    And thank you for a good

18   day's work.  I think we did a good job today.

19   (WHEREUPON,   the SESSION   was concluded at 4:47 p.m.)

20

21

22

23

24

25

1	CERTIFICATE OF REPORTER

2  COMMONWEALTH OF VIRGINIA

3  AT LARGE:

4  I do hereby certify that the witness in the foregoing

5  transcript was taken on the date, and at the time and

6  place set out on the Title page hereof by me after

7  first being duly sworn to testify the truth, the whole

8  truth, and nothing but the truth; and that the said

9  matter was recorded stenographically and mechanically

10   by me and then reduced to typewritten form under my

11   direction, and constitutes a true record of the

12   transcript as taken, all to the best of my skill and

13   ability.

14   I further certify that the inspection, reading and

15   signing of said deposition were waived by counsel for

16   the respective parties and by the witness.

17   I certify that I am not a relative or employee of

18   either counsel, and that I am in no way interested

19   financially, directly or indirectly, in this action.

20

21

22

23

24   CHARLES DAVID HOFFMAN, COURT REPORTER / NOTARY

25   SUBMITTED ON MAY 01, 2008

300



299

1	CAPTION

2

3

4  The foregoing matter was taken on the date, and at the

5  time and place set out on the Title page hereof.

6  It was requested that the matter be taken by the

7  reporter and that the same be reduced to typewritten

8  form.

9  Further, as relates to depositions, it was agreed by

10   and between counsel and the parties that the reading

11   and signing of the transcript, be and the same is

12   hereby waived.

13

14

15

16

17

18

19

20

21

22

23

24

25



0

0.2 229:1,4,6

0.26 228:20

0.3 229:1,4,6

053 17:1 199:22

200:1

06 41:10

07 41:11

08 221:19

1

1 2:5 47:5,9 48:3

49:1 50:23 56:15

62:18 63:12

76:20 117:14

143:11 235:18

265:2

1.2 108:21

1.2...you 108:17

1.5 18:17 225:21,22

242:9

1.6-1 57:11

1:00 138:17,23

10 24:11

191:17,21 251:19

256:2,24 258:23

264:6 277:22 292:7

100 75:17 161:7

254:15 256:11,21

257:16 259:8

10s 256:10

11 147:5 292:7

113 192:22

117 287:3

11th 217:5 219:15

12 205:19 287:20

120 62:2

13 142:13

14 42:16

15 16:24 38:12,17

67:5,7 146:3

208:18 277:22

15,000 245:1 285:5

15-minute 219:7,25

16 286:9

16-2 192:21

16th 216:21

17 42:10 44:1

17,000 225:24

247:24 263:12

264:1

18 42:12 75:15

238:9,13,15 262:9

1964 58:19

1965 53:12

1976 117:13 142:10

1992 145:22 146:13

1993 142:11

1995 36:18

1998 38:12

1-hour 226:5

2

2 15:21,22,25

63:20,24 68:16

77:16,21 84:23

108:11,16,22 143:7

243:22 251:19

253:19 256:5

267:11 271:2,8

2.4.5 67:21

2.5 19:19 28:11

191:17,21 195:8

2.5.4 68:20

20 142:14 144:25

208:18 253:10

200 128:11 161:7

197:9 200:6 227:10

228:11 229:16,17

236:6 261:9 287:3

2001 225:23 231:1

289:2

2002 35:6

2003 16:23 35:6

225:24 231:1 289:3

2005 16:23

2006 152:25 227:4

2007 54:24

2008 2:5 54:25

70:20

21 173:23

23rd 216:22

24 39:24 179:10

241 19:20

24-hour 199:25

208:20

25 24:13

250 228:11 236:19

287:3

26 16:19 104:14

243:7 281:15

28 208:19

29 142:2,15

29th 43:8

3

3 20:5,16,17 35:6

68:13 77:18

143:8 173:18

174:15 190:23

243:16,22

3.11 96:15

3.1-1 108:5,23

112:1

3.12 96:10 141:3

142:22

3.1-2 104:17

105:2 112:13

3.4.5 115:14

3:00 219:25

30 142:2,16 229:1

243:13

300 197:9 228:11

287:3,4,6

30-minute 228:20

316 106:24

317 106:24

318 104:17 106:24

33 42:6

37,000 10:3

390 115:12,13

398 115:16,21

399 115:17 126:17

4

4 20:4,17 148:13,18

149:1,25 153:7

157:25 158:1



162:24 181:13

258:6,10 260:24

4:47 298:19

40 243:13 245:2

400 39:20

45 44:3

450 257:9

5

5 20:16,17 23:3

26:21,22

31:1,11,15 63:1

139:23 142:24

148:14 160:10

163:13,15

166:21,22,25

167:6,21,24 168:10

170:13,14 171:7

172:15 174:15

179:2 180:10 183:1

189:25 199:11

206:10 228:13

5.1 23:2 51:21

184:11

5.31 124:8

5.3-1 27:4 34:2

37:23

5.3-2 105:16 106:8

5.34 177:7

5.4 37:3

5.9 23:6

50 33:20 38:22 66:6

259:8 267:14

51 96:16

5-1 130:11

510 105:16

513 143:3

5-15 208:14

519 180:17

522 173:23

523 124:8

53 17:1 199:23

5A 178:5 199:9

6

6 18:17 20:5 137:17

6:15 218:19 298:9

6:30 218:17

60 44:3 255:10

276:6

600 236:15

64 53:25

65 286:10

7

7 20:4

76 117:18

77 117:18

 	8

8:30 60:15 298:11

80's 62:4

82 208:19

9

9 292:7

90 257:15

90...10th 238:12

90's 16:18

92 145:10

93 141:25

142:1,23 143:4

146:3 148:3

95 227:4

98th 227:25

A

a...a 215:4 266:8

272:12 286:3

a...an 227:19

258:3,7

a...be 296:4 a...sort 278:18 a...there's 245:4 a...well 257:2

281:6

a...would 258:7 abbreviated 41:23 abilities 85:15 ability 97:14

232:21

able 41:7,13 62:9

117:6 145:5,8

171:11,24 211:1

242:4 253:10

261:13 265:5 285:2

absence 25:4,14,16

absolute 229:14

230:11

abstracts 55:5,11

academic 272:24

Academy 152:18 accept 202:19 access 145:23

273:15

according 42:10

82:24 152:17

235:25

account 273:1

280:6,21 285:12

297:7

accuracy 292:15

accurate 55:14

252:15

accurately 40:5

112:16 280:3

acid 17:6 66:13

77:12 280:5

acid-aldehyde 29:1

acknowledge 35:13

41:24 46:11

acknowledging 30:19

acknowledgment

195:2

across 8:20 9:6

11:8 13:19 16:14

17:23

18:11,15,21

23:17 80:14

86:22 131:14

132:25 138:2

142:15 145:1

224:24 225:21

227:22 230:7

231:10,25 237:15

238:4,17 239:25

243:20 247:24

255:1,18 280:14

act 2:25 114:5





156:9 196:18

214:12,21

acting 27:13

28:19 36:9 203:5

action 60:11

61:13 127:10

active 50:8

activities 74:22

100:24

activity 101:8

155:7 273:6,13

274:7

actual 64:9 80:18

96:8 168:11,25

169:8 170:3 171:10

189:1 198:4 208:10

256:17

actuality 128:14

actually 14:23

17:22 31:11 32:7

38:24 44:21

47:15,24 48:8

53:19,21,24

58:19 59:11

61:12 64:6 65:17

73:11 74:17

76:4,17 77:3

78:1,20 85:15

95:13 103:13

104:16 110:4

111:25 114:3

115:19,24 118:3

140:1 144:23 145:8

146:2 150:12 151:7

152:14 155:9

156:20,23 157:9

158:2 160:9 162:14

166:21 167:22

169:21,22 170:2

171:22 172:17

177:22 180:16

181:2 182:6 185:17

186:6 187:13

200:9,13 202:6

207:8 208:9 209:21

210:17 213:22

222:24 223:1,11

224:6,16 225:20

229:13,18 230:13

235:17 237:21

240:21,25 244:23

245:2,6 247:18

257:5 259:10 262:8

266:3,13 267:3

268:13,19 269:3

271:12,17 274:17

280:3 295:8

acute 33:18

37:4,5 281:2

add 69:19 87:18

99:16 101:23

123:22 129:25

145:20 156:5,16

159:5 172:7 176:12

191:10 205:12

251:3 269:19 296:3

added 99:18

177:12 260:21

adding 20:20 70:1

154:8 155:15

addition 21:20

23:24 24:7,11,15

28:18 34:6

118:25 159:20

225:15 238:6

261:12

additional 33:23

67:23 68:4

222:15 229:13

238:22 248:4

274:16

additionally 68:21

address 20:18,24

21:25 35:1 36:10

73:12 88:7 97:18

99:19 103:16

121:19 124:23

154:13 179:13

180:5 185:10

196:1,18 222:12

246:16 295:5

addressed 25:19

30:25 51:8 64:4

73:16 74:18

122:8 124:6 125:12

126:5,16 130:16

170:21 173:21

179:25 186:13

294:11,14,19

addresses 191:3

252:4

addressing 46:23

65:1 109:23

adequate 113:1

149:2

adequately 34:7

51:7 138:4

adjust 246:13

248:2,3,14

adjusted 29:19,22

adjusting

29:18,20 120:4

adjustment 243:19

244:6 246:10 250:3

255:1,18 266:15

adjustments 29:21

196:24 284:3

285:11

administrator 4:7

5:8 41:10 42:6

63:16 191:18

215:5,6 296:11

admissions 22:16,21

23:20 37:10

90:17,18,19,21

208:13

admit 83:10 155:2

192:14

adolescence 149:18 adult 119:11 137:23 adults 22:22

149:8 150:7 165:12

advance 3:3

Advanced 221:18 advancing 60:9 adverse 21:8 104:25

109:4 201:5

adversity 158:18

advice 4:7 5:8,23

203:10

advise 194:5 195:3

advising 203:24





advisory

2:2,3,11,16,17,23,

25 3:21 4:5,25 5:1

affect 30:20

74:22 76:21 83:17

affected 277:3

affecting

85:14,18 244:13

affirm 160:4 afford 138:23 afternoon 3:10

5:4 25:20

66:12,17,19 103:25

against 97:4

105:6 168:6 174:22

203:6

age 38:19 47:19

99:18 149:17

152:22 164:23

age-based 149:14

agency 2:1 3:15 4:3

13:15 41:13 42:9

47:5 150:4

171:10 195:3,23

198:3,10

Agency's 3:23

12:8 42:13

agent 97:9,12 142:4

157:14 179:21

aggregate 255:10 aggregating 280:18 ago 29:12 112:3,6

127:2

agree...that 131:20

agreed 10:25 131:21

150:2

ahead 53:24 74:13

90:25 92:3

100:13 111:5

140:13 144:17

148:20 164:1

182:22 183:13

192:3 194:23 197:2

224:15 265:18

290:25 295:19

Ahmed 143:10

air 2:3,11,15,23

3:21 4:5,9,17,25

7:10 16:25

20:11,12 26:18

29:9 32:22 33:5

34:8,15 36:12

38:25 47:2 57:6,10

74:18

78:3,5,6,7,10,12

79:1,5 87:16 90:22

112:18,25 113:8,19

114:5 152:10 173:8

175:12 196:8,11

214:21 220:5 221:5

222:21 223:8

224:22 225:13,17

226:9,20,23,25

227:2,23 230:12

233:21 238:8

239:12,14

240:16,18,22

242:19,21

243:2,3 247:5,16

248:14 251:25

260:2 261:8

267:5 271:10 272:5

273:21 274:3,24

276:9 281:9

283:2 285:11

286:11 290:17

293:7 294:7,10

297:2 298:3

AIRMOD 226:2,25

244:22 251:4,6

255:2,7 257:21

262:3 263:17

airport 226:3

283:15

airports 64:17

283:15

airway 104:18

117:10 141:17

235:14

airways 96:10,17,23

132:24 141:8

142:18 144:7 152:3

155:18 228:15,16

236:17

al 27:21 28:21 29:6

30:2,3 35:6 57:3

Alabama 6:18

alarm 42:21

albeit 65:4 70:14 aldehydes 29:22 alike 195:18 allergen 104:18

109:4 111:2 112:13

141:10 142:6,21

143:2 145:2,3

147:15 158:11

228:18

Alliance 32:23 allocate 284:23 allocated 42:3 allow 44:5 allowed 120:3

132:15 290:17

allowed...have

120:3

alluded 124:12

159:21

alludes 159:18

alone 128:10 238:24

alongside 205:5

Alpha 19:12 20:2,4

already 40:7,19

43:16 44:12

47:11 51:7 60:11

81:13,25 83:6 84:6

116:19 151:5

155:13 170:22

186:3 235:5

243:1 265:22

269:21 279:6 298:5

also...I 230:20 alter 139:5 alteration 105:24 alternative

222:13 224:3,5,7

290:8,10

am 2:16 33:1 84:1

110:16 139:10

171:7 180:15

186:17 188:24

214:21 254:21

ambient 4:9





16:2,6,7,9,10,14,1

9,21,23,25

17:14,16

19:3,5,7,12,15

20:2,11

27:8,16,22,24

28:7,10,19,22,25

29:3 30:3,24 35:16

58:10 65:19

74:16,19 83:13

149:3 187:20

202:23 203:2 223:9

224:25

225:6,7,15 226:7

248:9,15 253:23

254:1 273:21 276:9

ambitious 41:19

American 39:14,20

among 22:22

123:15 135:3

190:25 193:24

238:1 263:23

284:23

amount 58:13 112:15

243:20 250:1

262:22 266:8

amounts 257:18

AMPI 220:8 amplify 93:17 analogy 49:13

100:22 101:7

137:13

analyses

196:20,24 198:13

230:8 242:22

293:24

analysis 26:23 27:1

35:5,11 68:22

72:10,24 73:17

105:15 106:10

109:16 135:8,23

137:7 181:22

198:20 204:15

221:6,7 222:21

223:11 224:22,23

225:10,20

226:19,21,22

227:3,8 229:23

230:4 231:13

234:11,12 239:2

243:4 251:16

254:16 260:16,21

261:16 262:11

269:4 272:2

279:9 280:19 286:9

288:23

analyze 109:20

111:16

analyzed 204:5,7

and...and 212:10

215:25 252:9 254:6

260:22 262:2

264:20 272:9

279:14 283:14

285:9 288:21

290:13 292:22

295:21

and...and...and

213:10 218:1

And...and...yes

240:24

and...because

110:10

and/or 127:9 183:21

Anderson's 38:11

Angela 2:9,16

3:19 4:23

5:11,12 8:8,19 9:6

26:4 216:19

217:4 218:3 219:21

247:2 296:6,14

297:24

Angeles 257:17

259:17 261:1

262:14 263:13

265:23 267:16

280:14 283:17

285:4,6,23 286:1,6

293:4

animal 21:17

81:8,15,18,23

82:1,4,11 83:1

95:14

98:1,2,4,10,18,19,

20 99:9 102:16

103:12 106:1

114:16 118:20,22

119:1 187:3 190:11

213:17,24 214:2

animals 80:15

82:8 98:14

114:20 118:2

125:20 126:9

207:24

Ann 100:7 233:19

annex 10:12 34:3

48:5 56:9,11

68:8 74:18,23

83:22,24 86:10

96:16 99:17,19

101:19 186:5

270:12

annexes 64:12

65:3 270:14

announced 41:9 42:6

annual 16:24 17:3

86:18 103:4,15

175:13,17

199:16,19,20 227:9

238:12 244:25

247:9 290:22 291:9

answer 17:17

44:16 130:14 131:8

136:8,24 138:9

139:19 167:25

171:4 180:3

184:15,21,23

185:13,25 189:11

191:3 202:25 203:9

210:15 241:25

244:20 248:24

258:19 275:17

answered 63:2

121:21 184:25

answering 131:5

139:3 185:18

186:10 209:11

answers 63:4,10

171:5 185:1,23

215:7 259:15

anthropogenic 36:11





anthropometric

98:21

anticipate 51:23

224:6

anybody 84:22 100:4

117:5 184:16

215:12

anyone 9:12 63:12

96:12 156:23 167:8

217:24 218:8

anything 98:11

133:25 134:14

146:15 195:15

228:5 237:5 268:23

294:13 296:12

anytime 155:4

anyway 11:15 74:4

132:15 164:8 257:2

262:5 265:11

277:25

anywhere 168:21

169:9 268:5 279:11

aol.com 295:6

APEX 223:11 226:8,9

257:22 263:18

API 39:20,25

40:1,3,10 41:8

API's 40:4 41:4 apologize 280:9 apparent 125:7,14 apparently 83:15

165:1

appear 43:12

126:4 236:18

239:11 243:13

appearance 236:12

appeared 55:13

79:24 147:13

154:25 162:4

appears 60:12

91:6 96:2 180:20

288:7

appended 234:2 appendices 134:5 appendix 10:12 77:6

78:2,6,18,25

79:7 80:8 178:4

199:8

apple 272:11

application

33:10,16 41:17

50:24 57:10 79:6

100:16

applications 49:23

applied 33:15

65:7 244:11

251:6 252:18

254:10,13 270:7

applies 57:4

132:9 165:20

apply 14:20 57:5

111:10,20 275:9

applying 15:1 appreciable 135:7 appreciate

5:1,12,18 39:22

44:14 46:21 127:13

250:17 286:19

288:5

appreciated 93:20

110:7 286:9,17

approach 35:24 36:1

43:10 48:22

137:3,5 221:3,21

222:19 224:2 225:6

258:3 262:3 264:16

265:6,7,16 278:14

approaches 43:13

65:16

appropriate 94:11

102:21 111:23

112:19 114:1

149:10 248:7

260:16

appropriately 108:5

185:21 206:5

appropriateness

257:4

approve 217:7 approved 196:14 approximate 208:25 approximately

46:7,8,13 253:19

April 41:11

AQMP 283:24

are...are 232:16

249:9

are...or 231:22 are...sort 284:3 area 16:17 33:20

57:7 66:16 70:7

119:20 136:25

147:23 161:5 205:6

240:12 243:20

257:11 264:20

283:24 285:24

286:5

289:3,12,17 294:9

areas 16:22 33:12

70:13 93:24 154:17

160:12,19 170:6

175:11,15 222:15

236:4 242:18 280:5

281:3 290:13 294:9

aren't 32:8 61:5

72:1 90:25

158:16 190:18

272:18

are't...that 245:5

argue 136:23 137:20

138:3 150:14

152:20 212:19

236:22

argued 54:9 argues 197:23 arguing 136:25

197:22

argument 134:13

136:18 153:9 198:3

201:3 212:16

arguments 197:6

198:5

Ark 172:8

Arlene 220:13

271:15,16,17

array 100:21 article 38:12 39:5 articles 37:25

38:5,8,13 55:12

106:9

articulate 213:6





articulated 10:22

13:15 14:10 172:16

articulation

87:22 122:11,16

artificial 212:20

as...as 222:4

275:20

as...one 262:21 aside 113:20 152:10 as-is 239:6 288:13 aspect 10:17
131:22

189:12

aspects 39:21

60:1,2,17 65:8

80:11,12 122:3

129:15 131:14

134:22 194:2

assess 11:19 239:13

279:15

assessing 67:1

182:7 221:22 224:3

assessment

3:23,25 4:15,22

13:16 20:7 34:21

42:4,8,14,15

43:4,23 45:9,18

47:2 63:6

68:6,19,25 69:5,17

70:2 72:7 78:4

79:9 80:19

81:2,6

83:9,11,21 84:4,14

85:4,19,22 98:6

104:1 128:23

131:12 148:5

149:10 150:2

158:15 159:10

166:3

167:9,16,17,23

168:11

169:9,16,17,23

170:3,9 171:12

173:2,3,21 174:5

178:25 179:5 183:6

185:11,22

186:12,19 187:11

188:8,14 189:25

197:8,15 198:18

205:18 206:7 210:7

211:6,8 221:12

222:9,15 223:8

226:23 229:4

234:23 237:7,14

274:24

assessments 46:10

110:9 169:24

170:2,14 200:4

203:12

assessor 169:7

170:21 171:2

assessors 171:10

assigned 248:17

294:21

assignments 285:1

associated

16:8,10 17:5

20:8 21:8 76:19,25

103:1 112:21

113:21 114:24

170:9 222:4,12

289:11

associating 76:6

association 19:2

27:7 34:20

36:17,20 40:17

60:2 105:22

107:8,20 137:12

193:7 202:12

associations

27:3,11 29:13

30:7,16 34:1

35:7,13,14,20,25

36:21 38:19,24

103:11 175:15

199:1 208:18

assume 78:13

assumed 245:7

249:24 284:7

assumption 37:16

83:15 180:22

249:10 256:25

assumptions

185:19 271:7

278:13 292:17

293:9

asthma 22:23 38:9

115:21,25 119:13

149:8,16,19

155:4

190:9,14,16 201:9

asthma's 115:25

asthmatic 109:5

119:12 126:6,19

138:8 155:13

157:16 206:3

231:15 232:10

asthmatics 97:7

104:22 106:6 111:2

135:5,6,10,11,15

137:2 140:9 150:22

151:17 231:15,17

Atlanta 265:24

276:5,6 277:19,24

Atlanta's 275:25 atmosphere 21:1 atmospheres 46:23 atmospheric 7:7

64:1 77:11 280:9

atmospheric/

chemistry 15:3

ATS 158:16 attach 296:11 attached 296:15 attempt 69:16

108:24 134:20

183:4 240:14 255:9

attempted 133:21

attention 31:5

33:12

attributable

21:11,12 23:10

204:19,21

attribute 192:11

208:11

attributed 245:24

audience 238:15

283:3

August 221:13

Australian 119:21

author 39:6 147:4

149:13 163:1





authors 37:15 38:24

55:7,15

authorship 48:10

Automobile 32:24

available 14:20

24:5 35:9 36:21

51:15 60:14 65:5

72:12 99:14

100:6 121:4 142:23

143:2 145:7,18

168:23 226:6

233:24 241:5,6

282:1

average 66:3

69:21 70:3,6 83:20

88:10,14,16,19

89:4,15 90:2

91:19,22,24 135:21

170:25 203:13

204:2,10,22 208:20

227:9 238:13

244:25 247:9

256:13 257:8

276:14 279:12

291:10

averages 66:22

135:2 209:15

282:13

averaging 168:5

170:23 179:8

avert 274:19 avoid 52:20 avoidance 273:2

avoided 36:4 193:4

Avol 6:9 93:15

110:3 159:6

188:7,13

195:16,20,21

204:16,25

211:15,19,25 212:4

282:25 284:16,25

285:21

aware 6:4 90:10

198:9

away 72:23 77:10

93:24 160:15 176:9

256:20 267:3,14

284:4

awful 288:3

awkward 52:22

axis 105:6 231:12

B

background 15:10

79:20 80:4

153:14 185:16

221:1 255:8

backup 235:23

bad 80:25 81:6

181:5 193:19

195:13 243:12

246:8 273:4,5

277:22

bad...if 273:4 bag 265:22 balance 107:4

117:25 118:6

balanced 109:9

ball 272:19

Balmes 7:25 8:19

9:6 110:14,16,19

111:6 112:4,7,11

114:12,14 115:11

124:12,20

147:2,4 148:23

165:4,8,11,15,18,2

1,25 189:19,22

194:21,24 212:25

213:3,4,5 215:13

barely 287:4

based 19:22 28:12

29:5,13 65:18 79:4

88:22 91:9 92:20

118:24 139:9

149:15 153:12

164:25 167:23

198:13 210:19

221:5 227:8

229:5 240:1

265:7 268:7

271:6 291:2,10,18

bases 203:13

basic 103:15

139:4 179:19

270:22

basically 19:20

31:13 41:25

54:22 87:12

98:22 162:17 177:9

179:8 190:23

223:7,18 258:9

268:12 279:18

basin 280:14,24 basing 110:25 basis 17:11 36:22

40:17 81:2 82:15

121:21 132:10

135:20,21 173:4

197:14 203:12

206:7 207:21,22

271:5

be...because 116:22 be...it 288:12 be...since 265:21 bear 59:4 61:16

153:19

bears 105:4

became 68:18 95:21 because...or 256:24 become 50:9

105:10 149:19

155:24 160:1 173:3

becomes 43:18 69:22

104:8 107:19 156:7

159:24

becomes...the

291:22

begin 102:19 103:22

199:5

beginning 24:24

41:5 59:3 88:5

130:12

behalf 4:3 26:17

32:23 233:20

behave 196:9

behavior 273:2

274:19

behind 206:17 283:4 belabor 43:15 belatedly 67:10 belief 237:13





believe

21:7,13,15,19

22:19 23:13,17

24:16,25 25:3

31:18 33:11

58:19 69:4 138:9

139:6 174:25

183:25 187:17

195:4 199:1

202:10,21 207:10

209:19 221:13

258:24 279:16

284:5

believed 187:15

201:14

benchmark 117:11

118:5 228:10

229:16 230:10

231:10,24 233:2

234:19 235:3,4

236:7,23 237:4

238:24 240:15,18

246:20,23 247:21

248:6,13,20,24

250:2,10 287:22

benchmarks 128:22

210:16 223:16,18

232:6,12 234:25

236:5,21,24 238:14

239:14 287:3

benefit 67:23

benzo(e)pyrene 29:4

Berge 127:1 besides 103:4 best 24:2,5

104:17 183:5,8

187:11 194:14

208:23 264:11

283:24

bet 183:8

better 13:10 16:4

38:1 42:25 44:8

53:10 57:13

62:13 65:24 70:9

78:21 85:20 86:5,8

87:5 88:20 89:20

105:2 114:17

116:14,23 121:20

127:5 144:16

145:12 165:2

172:12 186:11

224:17 252:3

265:16 267:12

274:17 280:13

bias 35:2 51:9,25

57:17 58:1 81:1

biased 34:22 106:20

bigger 64:25

92:11 177:14

283:10

billion 104:3,4

105:24 106:17

107:14,15,16

119:25 197:10

199:7 208:18,19

binary 146:25 biochemical 80:11 biological 21:18

35:15 36:5 61:13

94:7,9 119:3

125:23 127:6 138:6

155:24 162:14

163:4,5,8

biologically 153:13

Birmingham 6:18

birth 25:12

152:6,23

bit 10:6 14:15 45:8

50:12 55:24 56:6,7

76:18 78:9 79:5

80:2 92:16

93:16,24 94:22

95:18,22 99:16

113:21 116:19

117:1 121:1 133:13

154:22 155:24

158:5 159:3 173:15

181:12 191:6 221:1

222:6 226:20 227:2

246:15,18 258:14

267:12 270:12

296:16

bite 192:14

bits 144:4 158:23

172:20

black 22:3 79:18 blind 87:9 286:23 block 225:25

226:6 264:10,21

275:8

blocks 264:4

269:6 275:8,20

Bloomberg 6:22 blue 16:17 178:1 blur 120:24

BM 211:16

board 2:2,17

145:1 237:15

body 102:25 188:1

270:23,24

bold 146:10

bolded 221:10 bolded...the 221:9 borders 16:17

Boren 58:19

borne 125:18

129:8 256:19

bother 173:12 bothered 189:24 bottom 16:17

130:4 148:14 174:7

bound 288:8

290:16,17,20

boundaries 278:2 boundary 285:22 bounds 51:22

boys 149:16,17 brain 252:7 break 97:21

99:22,23,25

139:1 198:22

219:7,25 236:3

266:25

breakfast 298:11 breaking 297:13 breath 223:6 brief 2:10,18

56:3,18 64:9 93:15

100:1 116:20 220:2

221:3 222:18





223:22

briefly 127:17

bring 11:17 24:23

57:16 58:16

61:16 62:12,18

77:22 94:10 129:10

134:20 148:6

151:20 154:23

208:6 209:4,7

255:2 259:3

262:1 266:13

bringing 78:19 84:1

118:10 292:14

brings 113:24

168:12 262:2

broad 13:12 37:4

100:21 101:1,24

102:11 224:23

238:4

broader 107:13

174:16 193:18

Brock 105:12 106:9

bronchoalveolar

157:15

Brook 28:20 29:6

brought 78:5,11

95:11 97:1

111:25 116:19

121:18 125:3,15,22

126:20,22 129:17

154:11,16 158:14

159:1 160:5

174:3 183:19,22

184:4,8 209:12

265:20

Brown 95:9,10 99:11

140:11,13,14,18,20

,22 141:2,12

144:17,22 146:2,11

163:17 166:9

Browning 60:13

b-tex 29:1 bugs 117:3 build 185:12 building 95:1

264:18

bulk 206:9

bullet 143:3,9

190:6 192:14

199:11

bullets 139:24

142:25 175:22

bunch 107:7

burden 158:22 business 53:7 156:6 busy 268:21 but...and 255:5 but...but
254:2

285:24

buttons 215:19

by...by 264:18

by-and-large 22:19

byproducts 214:13

C

calculate 19:20

292:13

calculated 31:14

84:11

calculation 78:24

128:9 268:12

calculations

81:13 83:17

84:15,18 157:21

246:22 259:1

calendar 233:25

California

6:10,14 280:2

283:18

called-out 48:12

Cambridge 26:16 can...I 297:24 can...if 214:17 can...now 295:20
can...the 216:4

Canada 22:5 73:12

75:2

Canadian 22:4 29:7

cancer 12:9

13:16,19 25:12

50:9

canyon 297:8,19

canyons 297:4

capacity 2:19

caption 208:2

capture 17:15

232:22

captured 216:3

273:16

capturing 92:23

268:11

carbon 58:20

cardiovascular 25:2

213:19,23 214:1

care 86:22 87:3

careful 42:20

48:11,17 172:12

181:23 196:4

carefully 34:11

111:17 153:15

Carolinska 105:12

carried 35:5

50:21 72:4

carrier 58:20

carriers 59:13

carry 128:16 274:17

carrying 161:6

carry-over 68:11

CASAC 2:3 4:4

5:5,23 16:12

17:7,22 18:18

19:8,13 20:18

23:25 33:13 35:4

39:25 43:7

62:20,22 217:7

221:11 242:18

CASAC's 25:19 cascade 278:13 case 24:10 34:24

42:8,18 50:12 55:9

66:7 77:20 95:1

101:1 118:5

120:6 135:12

174:15 188:4

225:19 237:12,23

238:7,9,11 239:3

241:14 257:21

260:2,4 270:1

276:4 280:20

288:20





cases 49:24 78:17

82:7 89:2 101:10

134:8 265:3

cast 220:3 casual 206:20 catalog 176:22

177:4 250:23

cataloging 177:18

categories 38:19,20

49:10,20,22 131:15

228:6

categorization

146:25

category 12:23

13:4,5,14,17,20

285:3

caught 83:23 273:10

causal 10:11

12:1,24,25

13:2,7,11,22

25:1,4,7,10,14,17

33:10 37:6 50:8

111:11,19 112:24

120:21,22 190:4

202:12 203:15

causality

11:16,21

12:4,15,22,24

37:17 57:24 124:14

133:8,9 187:24

207:21

causation 54:11

60:4 120:15 122:1

cause 12:15,17 36:7

42:18,21 88:2

90:21 103:6,17

136:22 244:8,16

caused 17:6 34:24

causes 37:11 122:17

203:6

causing 36:12

213:22,23

cautious 193:2,23

CBFs 230:25

cc 296:3,5,6

Celsius 24:14

census 225:25 226:6

257:8 264:4,21

275:6

center 4:14 7:12

146:1,10 148:7

255:16,19

Centers 220:9

central 22:8,17

23:12 24:9,12

88:19,21 89:3

90:1,3,4 91:2 92:9

257:11 264:20

286:5

central...a 257:11

centroid 257:8

264:22

centroids 225:25

226:6 257:23

certain 44:11

50:6,7,10 101:14

121:24 168:1

170:19 208:24

292:9

certainly 44:15

47:13,19 48:21

51:24 53:15 61:3

70:21 91:18 92:8

114:19 152:11

155:17 159:17

171:7 200:16

202:10 211:3,21

212:6 236:16 271:7

273:13 290:2,20

295:10

certainty 50:2

121:12 122:17,25

cetera 38:16

64:18 94:20

142:4 154:13,14

155:24 283:20

CHAD 273:1,6,17

274:8,12

chair 2:20 99:22

100:8 217:11

challenge

69:19,23 97:9,12

103:19 104:19

109:5 138:24

141:15 142:3 145:2

147:15 156:16

158:11 193:15

challenged 111:2

141:9

challenges 99:1

141:18

142:3,6,8,19,21

143:2 189:11

challenging

156:17 206:1

chambers 61:25 chance 217:2,25 change 24:16

48:18 97:6

105:23 106:15

131:24 141:8

143:12 145:1,6

206:1 217:15

246:22 264:15

273:5 285:13

296:12,13

changed 24:22 47:18

changes 33:8

47:15,21 48:4

56:17 142:3 147:14

267:18 289:4

291:22

changing 145:1

249:9,11

chapter 15:21,25

23:3 26:21,22

31:1,11,15 49:1

50:23 56:15 63:1

68:13,16 76:20

79:13 80:4 84:9

94:16,20,22,25

97:2 110:5

111:9,12,22

116:8 123:5,6

139:23 142:24

146:2

149:1,12,23,25

150:2,13,22

153:7 154:13

157:25 160:10

162:23 163:13





165:5 166:22,25

167:6,21,24

168:9 169:6

170:13,14,24 171:7

172:15 174:9 177:2

178:20 179:2,4

180:10 181:13

183:1 184:11

189:25 190:23

199:11 206:9

207:19 228:13

271:2,8

chapter...Chapter

258:10

chapters 20:17,20

50:22

167:8,11,13,18,24

171:3 172:22,25

character 285:13

characteristic

256:2,10

characteristics

87:16 242:20

characterization

150:18 170:12

194:25 221:5

222:20 223:13

224:19 225:13

230:12 238:9

239:15

240:16,19,20,22,23

247:6 261:8

267:6,10 297:2

characterize

11:20 139:23

175:23 223:15

239:4 267:25

268:16 286:2

290:21

characterized 33:25

65:24 153:7 238:20

267:22 275:19

characterizing 10:9

181:10 268:15

characters 220:4

charge 10:5,6,8,16

15:22 20:16 47:5,9

48:3 62:17

63:12,19,24

84:2,23 100:3

102:15 120:13

121:22,23

122:4,5 123:4

138:18,21

148:12,14,18,21,24

150:16 157:25

158:3 159:3 163:15

166:21 173:6,18

174:15 181:2

189:24 194:1 215:7

260:24 294:8,10,18

chart 287:2

charter 62:19

217:6,22 218:2

chartered...for

218:7

check 54:2 checked 55:6 chemical 12:15

20:11 94:8

122:17 255:7

266:20

chemically 156:8 chemicals 100:21,23 chemist 280:10 chemistry 64:1

156:12 157:17

244:13 269:17

children 22:22 62:3

119:13 149:8

150:7,23 151:18

159:23 165:8 203:7

231:16,17

children's 113:4,16

191:1 193:1 194:25

195:6,15,22

chime 220:21 222:23 chipping 283:19 choice 237:10

257:20 258:8

265:21,23 266:5

choose 223:25 273:5

choosing 81:1,2

chose 223:16 294:7

chosen 80:19

Chris 26:15

Christian 7:6 76:10

79:10 94:18 242:22

251:23 265:20

280:12 285:8

294:16,22,25

Christian's 254:25

Christopher

26:10,13

31:10,18

32:7,12,16

chronic 190:13

195:12

chunk 276:10 277:24 circulate 218:8,15 circumstances 77:2 citation 48:9
citations 297:11,19 cited 29:23 57:4 cites 205:20

cities 18:1,4 27:25

209:16 227:24

237:10,12,20,23,25

238:3,9,13,15,19

239:3,24 240:1,4

241:15,16,24

257:13 258:8

260:9,10,15 261:24

262:8 265:10

279:25 292:25

293:2 298:5

city 66:8 89:13

113:5 162:4

191:1 240:6 255:17

257:8,14,15

260:4

261:8,12,21 264:19

275:10 276:17,22

277:23 280:2 281:1

city-wide 91:16 claim 34:23 clarification 52:11

95:7 102:13 162:23

207:3 289:23

clarified 88:4





93:4,10,13

180:20 181:1

clarify 46:2

163:9 206:9 287:23

clarifying 85:20

122:15

clarity 169:15

Clarke 6:11 class 156:15 classically

198:11,12

classification 50:5

71:2 72:1

classify 11:20

257:20

clean

2:3,11,15,23

3:21 4:5,25 81:6

114:5 214:21

cleaned 270:5

clear 8:7 34:4,24

57:8 68:18 95:7,16

96:22 118:20

122:11 160:20

163:1,10

165:10,21,22 166:1

167:1 174:9,11

183:8 186:5 207:12

250:13 257:24

287:24 289:5,7

290:3

clearer 14:6

89:24 122:16 174:8

clearly 8:20

57:16 77:18

83:13,15 88:5,24

89:21 97:4

102:25 104:20

105:19 108:6

112:24 125:8 137:6

174:17 183:3

197:16 206:5,8

209:3 232:4 237:25

249:13,19 252:8

256:12 287:9

clinic 32:2

clinical 21:18

30:8,13,15,23

31:12 32:4,5

80:20,21 87:18

95:15 97:18

99:10 102:17

105:18 106:6 111:1

118:19,22 119:6

124:24 125:5,21,25

139:9 140:4,8

168:7 175:20 179:7

187:5,12,23,24

188:2 197:19

201:6,22 205:17

206:2,11,15

207:10,23

208:5,9 210:16

213:16,17 223:17

224:13 228:14

291:18

clinically 99:8 clinicals 126:2 close 15:17 17:11

39:18 71:25

75:7,18 91:14

154:3 210:25

242:13 254:3

272:10 281:18

closer 75:12 182:20

240:14 261:24

closure 77:22

CMAC 78:4

CMSA 282:8,9

CMSA/MSA 228:7

CO 120:4 133:16 coalesce 161:9 coast 280:15

co-author 113:3

co-conspirators

59:12

coded 79:18

coefficient 18:9

19:15,21,24 51:19

coherence 23:17

36:2 49:14

123:15 125:7

coherency 126:15

coherent 21:16

28:14 114:21

cohort 88:11,16

92:20 152:23

colleagues 50:17

170:1 178:23

211:11 217:18

collect 273:10 color 79:18 colored 30:10 colors 22:4

Columbia 9:1 column 18:8,12 combination

215:17,21 223:9

244:17 248:9

combined 11:19

114:4 188:2 225:11

263:19

combustion 21:4

195:5,12 196:9

214:13,23

comes 16:6 64:17,21

103:19 104:15

105:22 106:5,7

151:13 159:9

206:20 235:9

240:13 260:20

comfortable 169:4

215:10,25

coming 8:6 50:17

127:13 151:24

152:8 221:11

234:11 245:19

278:5,24 279:3

commend 288:2 commendable 154:7 commending 33:9 comment 3:7,9

5:11 26:4 39:22

40:19 43:1,20,23

44:2 47:13,17 48:1

49:5,18 51:4,7

53:5,19,24 55:21

57:14 63:7 65:22

70:10 78:1

83:5,6 85:23,24

95:11 100:16 111:7





114:23 143:16

146:21 156:5 158:2

160:10 163:22

182:25 204:14

214:17 223:21,22

233:19 234:20

246:17,18 250:20

254:25 255:23

260:6 271:21,24

278:8,11 285:21

286:7,8,20

288:6,17 292:5

comment...this

95:10

commented 201:18

commenters 125:4

233:17

commenting 177:1

294:20

commentor 39:13 commentors 26:7 comments 3:8,11

5:2,3,6 16:12

17:7,22 18:18

20:18 25:19

26:3,17,20,22,25

33:7 35:19 36:22

37:12 38:4

39:15,25 40:3,23

44:10 46:22

47:10 48:4 50:18

55:18 57:18,21

58:18 61:24

62:17,22 63:12

67:18,19 68:18

69:14 70:11

76:11,14,16 79:4

83:4 84:20,22 86:2

88:5,23 93:16

94:3,4 95:6,7,12

97:1 104:11

110:1,4,7,14,21,24

117:21 120:6,10

121:9 122:9

124:4,5,7

130:2,9 131:1

133:12 148:21,24

149:1 150:1,10

153:22 157:25

158:14 162:20

166:6,8 169:13

172:5 173:7 176:10

182:18 183:18

184:8 190:20

191:12 212:10

215:24 216:22,24

233:17,20

234:5,10,17 239:25

241:1 242:25

250:17 251:21

254:23 262:21

264:13 268:8

269:20 274:23

279:5,19 285:1

288:1 292:10

293:14 294:23

295:3

296:10,11,13,20,21

297:3,22 298:8

committee

2:3,11,16,18,21,23

,25 3:1,13,21

4:5 5:1 10:23 47:5

55:20 191:18

193:25 218:7

222:17 294:14

committee's 47:22

common 12:7 20:10

34:8 155:5

196:8,11

communicate 45:1

210:6 270:20

communicated 110:12

community 12:19

14:6 91:9

community-based

90:14 91:23

commute 60:15

276:15

commuter 285:25

286:4

commuters 275:10

276:8

commuting 275:6

compare 86:18,23

98:9,13 124:10

168:6 223:18

253:16 254:1

272:17

compared 21:11 55:8

77:7 86:18 243:8

257:14 282:16

283:22,23

comparing 86:15

224:7 282:12

comparison 36:25

82:21 98:2 135:2

145:6 224:10

227:21 252:17

259:18 272:18

281:16 282:20

283:13

comparisons 87:1

252:24 281:21

compelling 28:19 competence 170:1 compilation 52:4 complement 101:21
complementary

224:20

complete 33:16

41:13,19 42:9 57:7

76:24 78:25

117:5 241:3

completely 85:10

89:6 102:22

103:8 105:22

109:22 197:4

210:14 213:13

249:7 258:2

completing 41:7 completion 219:15 complex 27:9 155:11

156:7 225:19

complicated 70:15

192:17 214:19

246:15

complication 97:10 complimenting 213:5 component 4:24

20:25 83:9 225:9

components 196:16





224:17,19

composition 87:19

compounds 29:1

87:20 156:15

comprehensive 15:24

51:2 245:13

compressed 41:25

43:17

compromise 82:13 compromised 273:3 computational 70:16

263:15

computations 263:17

273:22,23

conceivably 135:12

concentration 12:18

16:2,16,24

17:2,3

18:10,12,14,16,20

19:7,16 78:16

81:19 82:6 84:7

118:24 120:19

204:22 226:24

227:9 229:15 235:2

248:6,10,16 250:10

266:15 267:23

274:9

concentrations

12:21 16:7,20

17:14,25 19:4,10

27:2,8 28:11,22,23

29:3,19 30:12,24

31:14 32:3 36:7

65:13,18 66:4,16

67:22 68:20,23

69:7 75:16

77:5,7,15 82:9

83:7,17 99:4 119:4

154:14

208:15,16,17,19

209:1,2,24

225:1,3,7,8,13,17

226:5,9,12,25

227:10,15,16

230:2,7 231:4

244:5 247:7

248:3,19,25 249:21

252:11,16

253:8,12,19,23

255:19 259:7

261:11 266:23

267:21 269:5,6

271:19 276:14

277:8 280:4

282:17,18

concentrations...

I 248:18

concept 12:18

103:15 132:12

180:20 252:8 256:5

concepts 56:3 102:4

197:3

concern 41:1

42:18 43:22

51:10 61:6 66:2

106:4 111:4 112:19

119:17,20 130:25

136:9 137:1 173:20

174:16 178:12

239:7 243:6,25

259:25 285:10

concerned 41:16

46:24 51:16 107:21

136:11 170:16

284:11

concerning 11:16

concerns 40:10

177:7 243:1

concise 140:3

conclude 233:14

238:23

concluded 29:6

31:20

38:13,18,25 43:7

209:21 228:19,25

298:19

concludes

25:23,25 53:21

60:6 233:12

conclusion 29:13

30:3 36:16 94:21

107:22 109:17

112:23 126:14

133:4 175:2,18

176:22 228:22

conclusions 14:21

24:20,21 37:3

40:13,17 56:14

109:22 116:23

130:19 147:1

167:7,18 168:20

172:19 173:15

176:20 177:2

178:14 201:15

202:8 207:13

210:3,5 228:14

229:5

concrete 293:18

condition 119:12

152:5

conditions 56:25

80:14,23 151:21

155:3 163:5 237:13

239:6 269:24 289:1

conduct 168:11

169:8 170:2

conducted 41:22

175:11 181:9 222:9

conducting 42:13 confer 60:10 confidence 19:21

51:22 144:8

283:9 284:1

292:12,13,23 293:4

confident 175:14

203:23 253:1

confined 70:13

264:18

confirm 217:21 confirmed 120:5 conflicting 29:16

106:23 109:16

confounded 113:9

191:8

confounder 62:6

confounders

59:11,12

confounding 34:9

confused 89:18

95:23 164:17

246:19,24 287:1





confusing 32:13

159:12

confusion 164:25

165:1 289:22

congratulations

63:22

conjunction 34:8 connect 116:7 connection 59:15,17 connections 40:11
cons 78:8

consensus 12:20

136:6,12 215:6

consent 41:24 44:7 consequence 185:23 consider 27:13

34:11 35:14

60:3,17 80:10

83:19 88:13

102:2 103:3,22

104:5 107:16

109:14 132:8

133:21 136:7

146:15 163:13

283:12

considerable 80:9

280:15

considerably 290:21

consideration 5:5

14:1 49:12

152:12 158:13

169:12 259:9,11

considerations 36:6

94:2

considered 59:16

60:1 66:23

133:22 134:4

154:20 156:3,15

200:25 223:12

225:2 226:2 276:7

considering 156:6

187:20 288:24

considers 222:3

226:10 275:4

consist.... whether 131:19 consistency 10:21

11:8 22:10,20

33:17 34:20 36:1

103:10 107:5

123:15 125:14

126:15 133:8

137:11 138:1

271:24 272:22

consistent 13:15,18

20:20 21:16

23:14 28:14

38:14 46:13

47:25 67:12 106:12

107:5,19 130:7

132:25 145:4 187:3

239:18 241:4,10

253:24

Consortium 7:10

constant 78:14

145:3 246:3 249:24

255:1

constituents 196:8 constraints 44:6 construct 12:14

234:14

constructed 131:13

construction 238:18

283:19

contact 225:16 contain 296:23 contained 253:11 contains 229:23
contemplates 42:12 contention 246:4 context 19:17,24

34:15 57:10,14

92:2 149:4

159:13,24 193:18

195:24 196:13

continue 94:4 99:21

148:22 259:14

continued 33:12

continues 119:20

196:25

continuing 100:15

continuous 45:23

134:25

contour 79:18

contractors 5:13

15:8

contradict 30:3 contradiction 256:4 contrary 192:20 contrast 27:15
contribute 38:1 contribution 19:5

20:3 248:10,23

250:11

contributor 147:22 control 114:2 controlled 27:17

34:11 118:1 127:20

128:1 129:2,24

187:6 235:12

controlling 214:15 controls 214:12 convenient 166:24 conversation 143:19

171:23,24 223:14

conversations 88:22

283:2

conversion 77:12

266:21 268:20

convert 168:24 convey 221:21,22 convinced 252:14 convinces 146:9
convincing

190:16,19

copies 100:4,6,7

220:17

co-pollutant 24:1

114:25

copollutants

191:8,15

co-pollutants

20:9 23:24 59:10

87:20

133:14,15,21,22

co-presence

161:4,16

copy 11:7 247:15

Corporation

26:11,16





correct 26:3

45:13,22 108:21

116:4 136:13

139:16 165:5,19

210:1 244:10

corrected 245:23

250:14

correction 44:22

252:3

corrections 40:11

correctly 66:10

78:24

correlate 90:6

correlated 61:10

193:5

correlates 88:19

correlation 18:9

19:6,14,21,24

28:10 88:9

90:4,7,9 91:9,20

190:25 191:21

196:7

correlations

27:22,24

28:1,3,4,8,22,25

29:2,5 85:9

86:4,13,18,20,23

87:2 89:3,6,17

91:6,7 92:2,5

correspondence

55:10 129:23

corridors 64:19

Cote 4:12

9:16,19,23 10:2

14:25 15:12

24:23 44:21

45:7,22 48:14,18

52:13,17 53:3

100:11,14

101:12,22 102:12

122:10,14

181:17,24

182:4,12,17,23

185:16 186:8

count 91:1 229:18

292:21

counted 230:16

275:22 276:23

278:4

counter 136:17

counting 34:16,25

36:4 68:8 230:16

276:16 277:6

279:7,10

country 237:20

238:1,4 290:13

counts 229:14

230:11

county 222:10 226:7

264:3,5

276:19,20,25 277:1

281:24 282:3

284:20 285:23

287:19

couple 56:16

67:8,25 86:6 137:2

151:10,20 152:13

205:15 220:20

237:22 282:25

292:18 293:22

couple...well

123:23

course 76:22

77:10 92:21 150:22

184:6 186:21 187:6

214:1 216:25

226:10 231:21

232:16 233:8

244:15 276:8

court 46:11

court-ordered

219:14

co-variance 86:16

cover 20:16

234:5,6,9 285:6

coverage 48:6

covered 15:21,22

242:18 269:21

covering 47:20

295:25

covers 63:25

Cowling 62:18 131:8

184:20

CRA 233:19

crack 14:18

Crapo 7:11 101:23

102:18,19 110:23

112:2,5,9 130:23

135:25 136:2

139:12 145:20

146:4,14

197:3,13

199:13,18,22

200:2,23 201:1

205:12,15,24

206:14,23 209:12

292:4 293:17 294:3

Crapo's 112:14

127:25 203:1

Crawford-Brown

166:14,15,17,23

169:21 170:11

171:15,18,21

215:16,20

216:10,14,17

crazy 213:21

create 128:1 151:11

245:10

created 288:8 creating 284:9 credence 119:5

127:21

credibility

145:21 272:20

283:9 284:1

criteria 10:10,13

12:5,6 24:3

49:4,10 51:1

52:14,19,21

53:7,11,14,18 54:1

57:12 58:6 59:25

60:17 76:2

100:16,18 101:25

103:9,14 106:19

108:2 111:4

132:6,15 137:9

138:4 148:3

151:3,6,22 227:8

237:19,21,22

264:21

criterion 258:12





critical 4:24

42:3 43:18 86:1

119:9 120:1

174:4 206:2 241:14

261:15 294:23

criticisms 75:14

cross-referencing

86:12

cumulative 81:14

82:21 203:5

curiosity 272:24

curious 136:22

257:19 271:12

279:24 293:10

current 16:23,25

17:16 37:2,9

40:5 42:16 43:13

103:3 139:5,8

140:10 154:19

168:17 174:25

175:12,16 196:6

199:16 200:5 222:5

227:10,14,17

232:12 233:8

239:10 265:6

266:22 289:8,15

290:9,22

currently 169:4 curtain 283:5,7 curve 108:16 128:4 cut 100:25 163:21

164:1,2 166:9

171:22,23 253:18

D

daily 170:25 244:25

286:4

Dale 6:11 53:4

54:13,18 55:17

134:18 141:13

202:19 281:5

damage 128:17

dangerous 142:20

Danish 298:3

data 11:21 33:21

53:14 69:2,5

73:2 75:2 79:19

81:3 83:13

84:7,8 85:2,11

86:18,19 89:17

93:1 94:14 103:8

104:13,15 105:4

107:17,19

109:15,16 110:14

112:12,15,23

113:18 116:14

118:7 121:3

133:2 134:11

136:21 137:1,14

142:22 143:2,4

145:7,15,18

146:8

147:19,22,24,25

170:3,4

175:16,24 181:14

186:24

187:4,11,16,23

188:2,3

190:3,9,15,18

191:7 193:15

195:1,9,11

197:19,20,23

198:13,17 199:14

200:14

201:6,7,8,15,16

202:2 207:24

208:21 211:9

213:24

227:4,6,20

234:13 235:25

236:20 239:12,13

243:2,8 244:17

246:14 252:17,25

253:4,11,17

257:5 258:25

261:13,14

262:8,18,22 263:10

265:6

266:4,9,11,24

267:2 270:22

271:22 272:1 279:9

284:14,19,24

291:2,11

database 52:3

273:6,12 275:6

datas 84:7

data's 267:9

date 234:1 296:18

Davis 6:14

day 60:15 91:24

109:10 132:6

230:15 273:4,5

276:7 277:7

days 3:5 4:16 24:13

44:1,3 46:7,9

273:13,14 274:19

day's 298:18 daytime 277:8 dead 214:21 deadline 219:14 deadlines 41:7

219:24

deal 11:6 26:20

285:2

dealing 99:17 106:6

121:7 156:8 273:21

deals 63:25

dealt 201:6 debilitated 165:15 decades 49:7

decay 259:2 266:15

December 41:10

221:19

decent 181:4

decide 104:9

216:2 263:25

293:19

decided 41:9

77:18 225:6 248:5

deciding 60:3

decision 13:22 39:5

43:14 60:18 101:11

102:3 104:2 111:18

122:22 145:25

199:12 211:6

224:11

decisions 10:20

44:12 53:23 102:23

173:4 175:4 185:20

186:6

decisive 49:5 52:25





54:5

decline 268:10

declining 16:20

17:24

decrease 145:9

212:16

decreased 66:23

decreasing 18:1

212:16

decree 41:25 44:7 decrements 62:1,4 deep 130:6 156:9 deeper 59:15 128:16
defend 136:22 138:5 defended 197:17 defense 115:22

126:18 132:20

defenses 203:5 defer 110:12 define 164:22 defined 49:22

164:24 166:2 228:5

244:25

defines 286:6 defining 184:5 definitely 41:24

49:6 185:7 209:6

219:23 265:19

266:9 291:4 296:5

definition 121:20

165:2

degree 99:14 193:6

degrees 24:13 deja 48:10 delete 116:9 deleted 149:24

deliberation 4:24

deliberations 3:1

5:7

deliver 156:9 delivered 97:13 delivery 97:11,12

196:14

Delphino 30:2 delta 144:7 delve 193:25

demand 60:12 261:15

demonstrate 105:19 demonstrated 241:10 demonstrates 104:20

Dennis 7:2 densely 261:25 densities 255:20 department 22:16

37:10

depend 222:16 274:2 dependent 184:6 depending 127:2,9 depends 256:12

270:1 278:1

depiction 252:15

depth 55:13

Deputy 41:9 42:6

derive 253:4,6,13

297:7

derived 57:3

223:9,20

253:9,17 258:25

270:23 275:6

describe 78:6

94:18,19 95:2

132:17 199:7

described 43:9 55:7

112:16 230:5

describing 177:9

description 11:9

13:13 76:21,24

78:7 116:25

118:1 226:17 227:3

descriptor 12:23

13:19

descriptors 13:20

design 56:19,20

57:5 92:16 100:20

Designated 2:10 designed 224:21 designs 19:23 desire 217:9 detail 68:10
72:9

83:23 84:15

167:2 178:7,9

181:13 230:5 266:7

268:3 296:23

detailed 33:6 37:11

40:3,23 73:17

84:10

details 56:19,21

278:16

detect 109:2 detecting 109:3 determination 10:12

12:1 57:24 111:19

determinations

33:11 111:11 174:5

determine 12:15

31:16 32:4 80:12

109:20 112:18

187:24 230:9

248:19

determined 120:18

247:8

determining 248:12

develop 27:18 63:14

216:6 294:24

295:24

developed 40:13

43:12 159:10

developed... developed 252:10 developing 216:5

295:18

development 4:14

43:14 79:17 159:3

DFO 2:19 5:12 diabetes 154:10 diagnosed 119:13

Did...did 216:12 did...that 269:16 did...there 273:19 did...we 273:18
didn't...I 252:14 didn't...when

217:11

differ 189:4

difference 49:15

60:23 71:7 76:19

82:10 87:16 114:18

119:10 149:15





162:5,14,15 200:13

230:11 243:15

255:15 256:5

285:19,20

differences 85:12

87:21

125:19,20,23,25

126:5,8,10,12

149:21 150:5

164:23

different

19:17,23,25 24:8

26:23,25 30:10

31:12 34:17 37:9

38:19,20 50:22

53:18 57:14 66:9

71:10 76:5 80:14

81:15 82:19,20

87:13,19 91:14

98:14 102:8

107:7 120:21 126:1

137:10 138:2

145:15 159:21

160:12

161:21,24,25

168:3,5,20

189:24 198:9 211:5

223:12 224:9

229:24 230:20

231:10 232:1 233:2

237:17 248:16

249:22 250:23

253:10 265:8,25

275:20 282:20

286:10 294:8

different...if

98:13

differential 152:1 differently 287:16 differs 149:16 difficult 21:9

46:12 52:1 70:5

85:19 87:3

103:23 105:21

119:18 121:9

133:19

181:10,18,20 184:6

187:18 194:2

212:15 262:2 292:1

difficult...is

121:8

difficulties

99:12 181:13

difficulty 86:14

133:23 145:21

179:20 197:5,20

diffuse 196:22 dilutes 190:20 dilutional 154:18 dimension 144:15 dinner
218:4,6,22 dioxide 51:9 105:17 direct 226:18 directed 259:4 direction
78:14

85:13 140:23 206:1

directly 21:11

30:25 31:15 56:8

57:13 63:3,4 80:23

130:16 157:9,13

Director 2:20

Directors 7:10

disagree 112:22

113:10 127:25

190:5 194:24

disagreed 149:11

disappointed

58:17 176:13

disclose 249:7 disclosure 283:1 disconnect 116:2

118:22

discordance 30:14

discrepancy

245:11,17

discuss 21:5

31:11 60:23

100:3 110:2 118:11

148:12 180:10

181:12 206:15,16

222:3 265:15

296:24

discussant 63:17,20

148:18 215:3,9

242:23 295:17

discussants 298:8

discussed 29:11

34:7 78:2

93:3,4,6,12 106:25

151:1 153:7 154:21

158:8 173:18 184:9

187:15

discusses 102:15

discussing 70:22

89:23 149:2 206:10

discussion 14:15

17:7 19:9,14

47:4 52:23 56:18

57:2 61:7 63:11

78:4 86:4 97:18

102:18,21 103:24

106:20,21,22

107:1,22 109:8

111:8,21 112:17,20

113:1,25 114:7

115:5,19 120:12

124:15 126:17

127:15,17,18

129:22 130:1

133:13

136:3,13,17,19

138:16 139:3

151:16 158:13

159:14 163:16

166:21 174:14

182:25 183:10

185:14 193:12

194:17 201:12

202:21 206:18

223:23 228:9

273:19

discussions 25:20

50:20 68:15

99:22 114:9

125:9 148:22

154:12 216:20

222:7 293:24

disease 23:11 103:6

149:7 150:7 163:5

diseases 22:23

37:14 151:24





disentangle 21:10

24:3 113:11

dismiss 29:12 61:12

113:19

dismissed 113:15

201:9

dismissing 191:7 disparate 90:20 disparities 62:11 disparity 61:17

89:16

dispersed 245:9

284:8

dispersion 77:11

78:20,25 225:11

268:20 269:4

display 81:16

144:15

disregarding 27:2 dissect 155:11 distance 76:3

85:8 163:6

267:24 268:9

272:13

distances 253:12

259:2,7

distinct 120:20

231:3

distinction 50:1,11

89:17 91:20

92:25 93:5 175:3

176:4,6 181:1

distinguish

181:11 229:6

distinguished 4:4,6

distinguished...I'm

229:19

distracting 116:9

distributed

141:14,18

distribution

16:13,18,21

17:23 68:22 69:6

135:22 175:12,24

209:2 244:3

252:10,13,15,25

258:14 264:24,25

269:8 270:13

272:14,16 279:7

280:13 282:16,18

291:5,15

distributions

202:24 230:25

244:5 253:15

270:24 272:7

disturbing 79:25 diurnal 18:22 divide 82:23

248:21,22

divided 249:2

250:9,10

dividing 199:3 do...do 265:16 do...it's 293:19 docket 40:4,24 document
5:24

10:21,22

11:23,24 13:3,4

14:11,18 15:6,11

21:6 22:25 24:20

36:23 43:24,25

45:14

47:15,16,23

50:16 54:22

55:9,10 56:1,8

58:3,6,7,14,24

59:16 61:8,15

63:6,23,25

64:7,9,12,16

65:4,9,12,21 67:17

68:3,6 69:3

71:23 72:25 74:6,7

79:17,24 81:13

83:10,11,21

84:5,14 85:4,19

86:1 88:6 89:21,24

92:11 93:18

94:11,23,25

96:16 104:20

111:20 112:1,16

113:7 115:5 117:23

118:16 120:1

121:16 124:10,11

125:12 126:16

128:20 129:11,22

130:18 133:17

134:3 139:13

141:25 142:2,11,23

146:1,3,7,10,12,16

,17 148:3 149:6

150:12 154:25

155:22 159:11

160:15 164:16,18

166:6 171:3 173:22

174:16 178:19,24

179:14

180:2,7,21

188:9,15

197:5,8,15,17,25

203:18 205:19

207:2 219:15

220:1,5

221:12,16,24

222:11 231:19

235:9 240:2 241:18

242:19 250:17,22

251:9,13 254:12

270:2 274:25

279:15 290:1,3

295:25 296:1

document...not

270:14

documented 35:19

239:1

documenting 35:24

documents 5:20

11:8,25 12:7,12

13:8 14:13 15:7

44:12 46:8 47:14

48:6,8 56:12 68:25

79:23 168:10 185:1

221:2,21 241:7

dollar 131:2 133:11

domain 263:11,12

275:13

276:13,14,15

277:12 278:2 286:2

domains 285:22

done 10:24 14:10

44:23,24 45:3,4,14

53:10 55:13 58:3

59:20





73:9,11,17,23

81:13,24 82:15

86:3,9 96:17

98:24,25 103:18

105:3,12,20 107:23

109:6 113:17

126:24 132:13

134:24 137:17

144:13 145:10

153:8,15 160:12,13

162:3 170:13

186:24 187:5

200:17 224:19

239:19 244:8

246:2,16 251:25

260:10 261:13

262:7 264:12

265:22 266:17

268:18 269:1 272:5

280:1 281:21 288:7

293:3,21 295:15

297:6

Donna 7:9 67:14

69:8 251:23 254:19

265:20

don't...oops 281:4 don't...or 292:5 don't...they 116:7 dose 12:13,22

13:23,24,25

27:15 31:15

80:13

81:14,22,23

95:16,25

96:6,18,23

97:4,11,13,15,24

98:2,11,12

99:5,9 108:14

114:18,19

125:19,20 126:13

127:22

128:1,6,19,24

129:15 145:1,3

208:10 236:1

dosemetric 204:8

doses 30:8 32:9

82:21 83:1

114:22 128:15

236:14

dosimetry 80:3,11

82:4 95:11 97:17

126:8 154:12

155:23

dotted 30:12

31:8,19

double 34:16,24

36:3

Doug 166:20

171:13 172:4

215:10,11,15

216:12

Douglas 163:17

166:9,13

downtown 257:11

dozen 55:4

Dr 2:6,20,21

3:17,19

4:3,12,13,19

5:10,11,15,16

6:7,9,11,13,15,17,

19,21,23,25

7:1,2,3,4,6,8,9,11

,13,15,17,22,25

8:1,2,4,6,8,19,23,

25

9:2,3,5,6,8,10,13,

14,16,17,19,20,23

10:2 14:23,24,25

15:2,4,9,12,15,19

20:13,14 22:24

23:2,4,5 24:23

25:21,23,25

26:6,7,8,10,13

31:6,7,8,10,16,18,

20,23,25

32:7,11,12,14,16,1

8,20 33:2

37:19,20,22

39:2,4,10,11,12

40:7 44:17,21

45:7,22,25

46:2,4,5,15,18,20

47:8

48:14,16,18,21

52:10,13,15,17,24

53:2,3,4,5,6,7,8,9

,19

54:12,14,15,16,20

55:17,22

57:19,21 60:19

61:3,9 62:14 63:21

67:14,15

69:8,10,11,13

71:12,15,16,19,21

72:6,9,13,15,17,18

,20

73:3,5,8,13,20,21

74:2,9,10,12,14,24

,25 75:4,8

76:4,10,12

79:10,12

84:21,25

87:8,10,24 88:1

91:12

92:13,15,17,18,24

93:6,8,9,11,12,15

95:5,9,11,18

97:1,20,22

99:11,21

100:2,11,13,14

101:5,12,17,22,23

102:12,14,19

109:25

110:3,13,16,18,19,

22 111:6

112:2,4,5,7,9,11

114:12,13,14

115:8,11 116:16,18

118:9,13

120:8,10

122:10,12,14,19

123:20,24,25

124:1,2,3,20,21

127:12,16 130:23

131:6,11 132:17,19

133:6,12 134:18,19

135:25 136:1,2

138:14

139:2,12,14,17,21,

22

140:7,11,12,13,18,

20,22,24

141:2,11,12

142:9 143:15,16





144:17,20,22

145:11,20

146:2,4,11,14,18,2

0 147:2,3,4

148:10,23 150:9,11

153:24,25

154:3,5

156:19,24

157:1,3,6,7,22

158:2 159:4,6

160:2,4,21,23

161:18,19,23

162:19,25

163:12,19,24,25

164:2,4,5,7,8,12

165:4,7,8,10,11,14

,15,17,18,19,21,24

,25

166:1,7,13,15,16,1

7,19,23

169:18,19,20,21

170:11

171:13,15,16,18,20

,21,25 172:4

174:20 175:3,9

176:5,11,16,20,23,

25 177:20,21,22,25

178:1,3,4,6,7,9,10

,11,16,18

180:9,12,15

181:6,17,21,24

182:4,9,12,15,17,1

8,22,23

183:12,14,17

184:17

185:3,6,7,15,16

186:1,2,8,14

187:13

188:5,7,11,13,16,1

8,20,24

189:3,14,19,21,22

191:9,11,23,24,25

192:2,3,4,5,12,13,

18,20

194:4,7,11,13,14,1

6,21,22,24

195:14,16,17,18,19

,21 197:2,3,13

198:7

199:13,17,18,20,22

,23

200:2,7,23,24

201:1,13,25

202:6,14,15,16,18,

19,20 203:1

204:11,13,16,18,25

205:12,14,15,23,24

206:8,14,22,23,25

207:5,6,7,14,18

208:8

209:10,12,20,23,24

210:1,2,4,10,12,17

,19,21,22,23

211:12,13,15,17,19

,22,23

212:2,8,9,25

213:1,3,4,5,8,10

214:5,10,16,20

215:13,14,16,18,20

216:4,10,11,13,14,

16,17,18 217:21,23

218:5,12,14,16,18,

20,24,25

219:1,3,5,8,9,10,1

2,13,20

220:3,10,11,13,15,

19,22,23

222:22,24,25

223:1,4,5 224:16

228:8,12

229:9,12

233:13,16,18 234:4

239:21,23

240:3,5,21,24

241:11,13,17,19,20

,22,25

242:3,5,6,8,10,13,

15,16,17,24 244:21

245:21,22 246:1

247:2,14,19,20

249:1,3,4,5,14,16,

17,18,25

250:4,6,11,13,16,1

8 251:22,24

253:3,21,22,25

254:5,6,11,18,19,2

1,22,24

258:17,20,21,22

259:13,16,21,23

260:8,12,14,17,19,

20,25

261:4,6,7,17,18,20

262:4,6,10,12,14,1

5,16,17,19,20

263:2,8,9,17,19,22

,25 264:2,12,14

265:12,19 267:20

268:2,6,18,23,25

269:2,3,12,13,18,2

0 270:16,17

271:15,16

272:10,15,20

273:18 274:22

275:3,5,9,12,14,15

,17,19,21,23,24

276:11,16,18,20,21

,24

277:2,11,14,15,18

278:1,6,10,23

279:1,2,4,22,23

280:8,11,25

281:2,4,6,8,12,13,

14,18,19,22,25

282:3,7,10,14,15,2

4 284:6,18

285:15

286:19,22,25

287:8,14,15,17,18,

21,23 288:1,23

289:6,24 290:23,25

291:1,21 292:2,3,4

293:13,17,23

294:3,4,15,16

295:2,4,5,7,8,14

296:2,5

297:1,12,14,15,17,

21 298:7,15,17

draft 3:24

4:17,21 7:19 24:22

33:8,15 34:13

36:25 39:23 40:1,8

42:23,24,25

43:3,8,12,18

44:23,24





45:3,19,21,24

47:12 56:2

64:2,6,7,24

65:23 67:11

69:15 76:13,14

79:14 87:6 95:21

96:15 102:20 124:6

215:4 216:19

217:3,18 221:12

222:2 254:11,12

267:19

draft/Integrated

3:23

drafted 215:8 drafting 216:7 drafts 45:1

210:24 222:11

dramatically

85:16,17 107:20

291:22

draw 24:20 123:7

179:5 187:10

200:20 201:15

202:8 277:16

drawing 146:25

277:19

drawn 11:22 48:8 dreaming 214:14 drive 107:21

138:7 276:17

driven 41:24 85:10

driving 85:14

155:14 277:9

drop 259:3

drop-off 256:10 dropped 185:23 due 20:9,10 29:8

77:19 85:12 97:7

98:17 203:21 251:2

280:4

durations 31:14

during 17:9 19:7

24:13 35:4

66:11,17,20

99:22 125:10 222:7

276:6

dwell 234:10

dwindling 69:19

dynamics 70:16

E

earlier 33:7

38:12 111:9 116:20

117:21 154:12

156:5

167:8,18,24

171:3 174:22 184:4

189:23 190:23

197:18 224:18

230:21 247:6

earliest 137:21,24

early 210:24

221:14,15 278:8

easier 47:13 187:24

easily 86:13 113:15

280:17

eastern 280:24 eastward 280:16 easy 46:25 89:18

113:22 121:7 191:3

279:11

eat 138:23 298:10

EC 133:16,17

echo 147:21 178:22

Ed 6:9,17 22:22

23:20 92:14

110:2 113:2 153:24

156:19,20 159:4

160:24 180:11,13

188:5,10 192:25

193:9

195:14,16,17,20

211:13,23 278:24

282:24 292:14

edges 255:17

Ed's 110:21 212:10 education 156:20 effect 12:16,17

21:14,21

22:12,21 23:14

24:17 29:6 30:9,13

31:17,19,21,22

32:3,5 35:23

37:8 59:13 77:3

80:19,21 81:2

96:20,21

98:3,17,25

99:5,6 103:1

104:25 108:23

109:4 112:24

113:11 120:17

122:17 127:6 130:6

131:20

132:5,7,23

135:13,20 137:4,21

138:6 141:23

142:17 143:6 147:9

158:18 162:16

180:19 187:20

190:21 192:11

193:22 195:7,8

211:17,19,20

212:2,4,5 236:10

effective 64:2,3

99:18 179:2

effects

21:7,8,10,19

22:1 23:18 24:3

26:24 27:3,6

31:9 32:9 34:17,23

36:4,6,8,13,18

37:4,5 38:14,15,16

39:7 40:18

58:11,21 61:21

75:9,13,25 76:6,22

82:8 85:16 94:10

99:6,7 102:16

103:17,22 104:6

105:17 106:2

111:15,22 112:20

113:5,10

114:2,3,21,25

115:1,18,22 117:14

119:2,12,18,19,24

120:4,21,22

121:13,14,25

123:1,17,18 125:10

126:19 130:25

131:23 137:16,25

139:25 149:9

158:6,9,12

160:10 162:1,11





168:19,20 170:8

173:19,20 174:24

175:21,25 179:22

180:24 187:16,17

188:4 189:25

190:14 192:23

193:17 195:4 196:5

197:1,24

199:2,5,7,15

201:5,23,24

202:12,22 203:2,20

205:3 212:11,17

213:15,16,18,22,23

,25 214:1,18

228:20,25 236:17

238:25 281:11

291:18

efficiency 47:23

efficient 247:22

273:23

efficiently 45:2

effort 11:1,10

115:14 266:8 267:7

efforts 11:10

111:10 127:25

eight 107:15 eighty 119:25 either 11:7 51:18

56:10 66:15 71:5

78:24 138:11

143:22 148:9

155:20 167:24

168:9 203:13 204:1

205:2 267:25

either/or 183:20 elaborated 260:22 elaboration 227:2 elderly 151:18

159:23 164:21

Electric 7:13 electronic 47:19 elevated 18:25

285:16

elevation 285:17 eleven 99:24 eliminate 79:3,4

91:8

eliminates 88:17

eliminating 277:20

Ellen 163:1

Ellis 62:18,19 63:5

131:7 184:20

else 9:12 54:10

62:17 63:12

79:12 87:25 111:24

114:15 130:10

133:24 161:13

162:9 204:12

217:24 271:21,23

elsewhere 101:6

179:16 180:2

elucidated 197:17 email 8:10 295:3,5 emailing 216:7 emails 8:11

embrace 160:5

emergency 22:15

37:10

emission 226:4

255:20 283:24

284:13 290:15

emissions 226:3

244:13 245:3,5

255:10,11 280:18

emphasis 19:11

51:14 68:19 129:12

133:16 158:10

187:4 201:14

emphasize 116:20

119:8 206:5 279:5

emphasized 119:15 empirical 271:6 encompassed 245:1 encouraging 70:20
end....an 127:8 end...as 111:12 endogenous

155:1,8 157:19

endogenously 155:14

156:21

endpoint 35:21

98:7,16

endpoints 35:22

66:2 82:19,24

125:23 291:18

engineers 8:18

enhance 151:21

152:13

enjoy 294:22 enlightened 287:1 ensure 10:21 entire 85:22 86:1

89:13 91:3

107:21 175:11

243:20 275:12,14

entirely 281:17,19 entwined 214:3 envelope 82:25

266:19

environment

212:13,14

environmental 2:1

4:14 7:7,16 152:12

155:10

environments 223:12

226:11,13 287:20

envision 216:7

224:12

enzymes 157:11

eosinophils 155:21

EPA 2:16,24 4:8

5:22 6:23 7:1,2

9:15 13:18 30:3,18

39:3 41:6,9 49:6

60:21 67:18

71:13,14,23 95:6

102:23 109:12

147:18 164:19

165:5 169:2

194:1,5 200:17

207:3 217:12 235:3

237:21 241:23

258:18

EPA's 3:24 41:23

43:10 69:16

epi 19:25

36:14,24 60:24

71:8,18 72:5,16,21

73:9,15 74:22

77:24 85:6





110:14 114:16,21

116:1,14 118:4,7

121:10 124:23

125:5 126:2 186:20

187:2 202:2

208:4 209:19

213:14,18 224:13

epi-data 75:11 epidemiol 91:11 epidemiologic

21:2,16 22:7,20

27:11 30:7 36:17

76:6 153:1

175:5,10,23

181:8,11,14

187:15,25 197:23

198:16,24 199:4,10

200:18

201:7,8,15,16,22

202:11 206:10,11

208:22 209:3

epidemiological

19:17 27:5,16

28:13 30:16

31:21 32:2 40:15

71:4 123:13 134:21

168:7 179:21

199:14 202:22

203:20

204:4,7,15

207:20 223:22

224:1,9

epidemiologically

135:13 190:13

epidemiologist

191:13

epidemiologists

14:3,4,19 89:9

90:9

epidemiology

30:11 33:18

58:1,5,8,9 59:18

61:18 62:9,12

73:22 88:21

89:23 90:10

91:22

92:3,4,5,9,25

102:17 105:19

118:18,23 119:6

125:21 127:21

128:23 129:25

131:4 133:23

136:23 179:8 189:5

206:6

epi-studies 74:17

75:9

equal 124:14,22 equally 124:25 equate 288:12 equation 248:8

250:8 259:1

274:1,2 284:17

equations

84:9,10,12 270:8

271:4,5,8,9

equipment 105:13 equivalence 274:11 equivalent 99:4

248:5

error 57:2 282:5

errors 16:10

36:23 286:15

especially 60:2

70:2 91:23

114:17 128:21

149:3,7,18 151:1

153:16 181:7

187:16 202:13

280:2

essence 261:19

essentially 12:4,20

14:9 44:13 65:12

222:20 229:5

268:19 288:12,13

establish 11:16

36:1 97:14 119:3

established 174:16 establishing 97:10 estimate 22:18

23:12 24:12

82:19 98:18,19

225:7 226:24

269:13 284:20

288:11,16 293:1

estimated

226:5,13 227:15

228:1 230:6,12

267:21,23 292:11

estimates 22:9 24:9

51:21 85:18

144:2 168:24 200:9

223:19 226:5 235:3

240:11 256:15

estimating 99:4

232:19 248:11

288:24

estimation 252:9

estimations 297:5

et 27:21 28:20 29:6

30:2 35:6 38:16

57:3 64:18 94:20

142:4 154:12,14

155:24 283:20

Europe 73:11 75:1

298:6

evaluate 11:21

289:10

evaluated 33:17

35:22 125:24 225:1

229:8

evaluates 198:10

evaluating 13:24

34:14

evaluation 10:11

36:2 43:11 44:23

48:24 65:19 109:13

240:2

evaluations 45:3

57:23

evenly 245:9 event 149:1 eventual 224:11 eventually 63:15

167:8 214:20

everybody 54:7

62:17 79:12 130:10

131:2 141:2 161:13

162:9 164:3 180:10

194:20 215:22

217:2 265:14

everybody's 215:24

everyone 2:6,13





3:20 5:25 8:13

54:2 215:2

220:18 296:3,10

everything 25:22

200:10 226:17

246:23 284:3

291:10

evidence 11:20

14:21 21:13,21

22:20

25:1,3,6,9,13,15

26:24 27:5

28:14,19 30:22

36:17,20

37:5,6,9 43:11

48:24 50:6 51:3

53:22 58:4,10,24

60:14 102:25

104:6,21 106:17

111:17 115:17

122:21 123:10

137:10 144:11

145:18 149:10

169:23 170:7

174:2,18,24

175:6 179:10

180:18,23 181:7

182:7 188:1

190:12,21

198:13,14,16,24

200:8,17 201:17

202:9,11,22

204:5,7 206:10

235:13 241:5,6

evidence-based

224:2

evident 97:9 230:1 evolving 36:22 exacerbated 239:8 exacerbation 196:15
exact 129:23 272:18 exactly 49:20 60:22

145:8 169:15 170:5

185:4 193:6

205:11,19 207:8

228:10 249:18

270:9

examined 135:20

example 38:11

51:1 61:24 62:10

79:20 82:21

83:16 90:16

98:23 100:22 101:6

116:13 121:6 134:9

137:14 155:4,13

170:22 173:8 179:3

208:13 224:4

227:19 230:19

231:8,20 232:9

examples 50:13

101:20

exceed 118:19

exceedance 256:22

261:9 292:9

exceedances 85:18

228:1 230:15,17

279:8,20 288:9

291:3,11

exceeded 288:14 exceeding 199:16 exceeds 248:12 excellent 5:12 except
176:12

267:10

exclude 134:16

excluded 134:7

231:2

excursions 103:5,17

199:19 203:3,14,21

204:1,17,20

205:1 209:13

excuse 23:23 291:8

exercise 97:8

225:11 291:22

exercising 62:8

97:3

exhaled 155:5 exhaust 196:9 exist 123:1,3

181:16 182:8

274:21 289:1,2

existence 121:12,25

173:20

existing 82:3

170:17

exists 69:7 81:3

181:19,20 182:14

283:25

expand 19:8 61:7

222:14 260:16

277:12

expand...you 277:11 expansion 158:19 expect 107:9

130:4 141:16

251:4,5,18 255:16

experience 154:17

283:18

experienced 123:1

230:24

experiencing 231:5

experimental 14:5

30:23 56:22 57:1

60:8 121:4

197:19 201:6

experimentation

49:14

experiments

56:23,24 62:7

80:23 98:3,10

99:10 103:12 106:1

129:6 134:23

expert 15:7 expertise 198:21 explain 84:14

126:11 240:3 271:5

explained 247:21

287:9

explanation

83:11,20 237:1

explicit 10:23

87:15,22 134:16

172:14 183:2,3

185:23

explicitly 11:1

77:16 134:6 172:24

180:5 185:18

explore 51:25

explored 197:24

239:17

exponent 127:3,4,8





exponential

255:25 259:1

266:15

expose 129:6

exposed 18:25

21:4 104:5,11

158:24 159:20

161:8 187:7 195:25

205:7 208:24

231:23 232:5 233:9

237:4,16 240:15

277:5

exposure 3:25

4:22 12:18 15:3

16:3

19:3,6,7,13,15,16,

22,23 20:3,6,7,8,9

22:13 24:25 25:3,8

27:5,17 28:16

30:12 31:14

36:18 40:18

42:4,13,15 43:4,23

45:9,18 46:9

65:8 69:21 72:7

74:16,18,20 78:3

80:13,14,23

81:19 82:8,14

84:10 85:3,18

89:14 90:2,12,13

93:2 98:17 99:1

105:17 108:21

111:22 112:21

115:18 120:18

123:3 126:1 127:20

128:1,6 129:24

130:24

135:17,18,19

141:10 144:11

147:15 148:5

149:3,9 154:12

159:18 160:25

161:11 164:16

165:6,20 166:2

167:3,9,14,16,22

168:2,3,6,11,17,23

,24

169:7,8,16,24

170:2,14,21,25

171:2,10,12

173:2,3 177:8,11

178:24 179:5

182:10 183:6

185:22 186:12,19

190:1,13,17,18,22

195:12 197:19,20

198:18 202:13,15

205:4 207:22,24

208:1,4 210:7

211:6 221:7,12

222:9,14,19

223:11,19

224:18,25

225:5,10,11

226:10,14,22

228:21 229:19

230:4,19 231:4

232:20,24 235:12

237:13,17,19,25

238:14 239:10,16

240:9,11,13,19

241:2,3,9 242:19

244:3 247:17

248:8,11,17,19

261:2 262:11

271:7,19 272:1,7

276:9 281:10

287:21 288:25

297:5

exposure...the

237:19

exposure/response

144:12

exposures 27:8 62:8

70:4 81:20 88:4,18

89:12 103:2

114:19,20,24 118:2

123:1 125:21

128:14 129:2,10

143:5,6 152:2

155:9 173:9 177:17

187:6 189:8

198:4 199:16 204:9

221:22

230:10,17,23

232:23 233:1

236:6,14,15

237:3,16

238:12,14,22

239:4,13 249:23

265:3 288:24 289:2

297:8

exposure's 88:12

express 108:15

217:8 292:12

293:12

expressing 215:6 expression 292:23 extend 81:12 extended 17:7 147:6
extends 154:15 extension 44:2

46:16

extensive 48:6 extensively 40:14 extent 36:23

41:11 51:24 69:7

72:21 120:17 129:3

134:14 184:15

193:12 273:12

external 39:23

40:1,8,25 42:23

extraordinarily

155:10 156:7

extrap...between

271:25

extrapolate 80:13

extrapolated 82:5,9

83:2

extrapolating

70:3 80:24

extrapolation 80:12

81:4,10 82:16

83:12,21

98:1,11,21,22 99:9

269:23 271:23

272:1

extrapolations

84:13 99:16 270:22

272:5

extreme 239:10

265:3 288:20

291:14

extremely 12:20





118:17 292:2

extremes 264:23

279:7,9,13,21

291:4

F

face 150:16 151:2

faced 87:11 faces 238:21 facially 55:4 facilitate 26:4

facilitates 295:16

fact 27:13 34:6

43:7 45:13 48:10

50:18 51:20

53:21 54:8 55:1

64:15 69:19 76:2

86:2 114:17,25

115:1 116:8,10

118:21 119:3

120:13 128:6

147:11 149:15

152:1,15 154:21

155:5 173:5

196:6,12,23

203:4 204:25

207:20 209:13

210:6 212:16

214:18 238:24

239:2,9 254:25

266:24 274:15

282:2 288:16 290:1

factor 18:17

19:12,13 20:1,4

52:25 152:12

154:18 203:16

245:24 246:2 247:8

248:7,15,22,23,24

249:2,9,12 250:9

251:19 253:21

256:5 267:11

factor...I 248:2

factors 12:3

13:25 49:6 54:5

94:5 153:18

155:3 208:7

253:6,9,14 254:8

fair 58:13 101:6

185:12,15

fairly 8:20 15:17

69:18 74:7 86:9

111:23 152:17

256:14,16 257:11

fairly...a 256:14 fairly...I 257:23 fairness 283:1 faith 169:25

fall 128:21 193:16 false 36:25 familiar 11:3 70:7 fascinating 119:23
fashion 202:3,5

217:7

favor 136:25 feature 12:7 19:22 features 19:16

56:18

fed 116:25 117:1

Federal 2:10,25 feed 224:21 226:18 feedback 222:16 feel 11:18 48:2

58:2 64:13 87:2

113:1 132:4

169:3 184:16 185:2

211:4 215:10,24

217:24 252:3 253:1

feeling 62:20

66:5 80:25 238:16

Feldman 39:25

fell 173:16

felt 10:24 13:10

56:1 57:25 58:6

86:5,8 88:2 175:14

178:20 179:18

FEV1 145:1

field 269:25

273:4,5

fifteen 99:25

Fifth 37:3

fifty 104:3 132:1

133:11

figure 16:15,20

21:24 23:2,3

27:1,3,20,23

29:4 30:7,9

33:22 34:2 37:23

38:4,23 51:20

95:18,21,23

96:1,9,13,14,16,18

104:17 105:2

108:4,5,23

112:1,8,13

137:14 141:3

142:1,22

143:17,20,22 144:6

145:11 146:22

147:5 148:8 167:22

177:24 192:21

195:10 207:15

208:1 254:7

258:6 287:2 293:11

figure...I 105:2

figured 65:12

215:18,21

figures 18:19,24

19:19 28:20,21

29:17 97:16

117:4,6,20

123:10 183:1 292:7

filed 40:4

files 247:23,25

final 5:10

42:8,14 44:24,25

45:3,4,14,25

64:7 112:23 221:25

250:8

finalize 221:16 finalized 293:15 finalizing 221:24 finally 25:14 43:21

117:25 153:6 286:7

finding 61:10 106:4

207:1

findings 11:18 16:4

28:12 30:21

31:2,12 36:15

55:16 107:1

144:3 193:2 201:22

203:20 206:11





finds 58:11

fine 28:8 29:19

69:12 75:20 110:18

115:20 119:19

124:2 149:25 176:6

finger 180:23 fingertips 73:12 finished 115:10 first 3:24 4:21

10:8 12:14

16:6,9 19:18

23:3 24:22 33:14

40:1,22 42:24,25

43:3,8,11,18,25

44:1,9,23 45:19,21

47:10,12 48:5 56:2

62:21 64:2,23 66:6

67:19 68:3 69:15

73:8 76:17 78:9

79:14 80:5 85:24

87:6 93:18 95:12

96:15 106:8

140:14,15 141:12

155:2 164:13

172:12,24 174:11

178:22 217:11

220:20 222:2,21

223:8 226:17

228:17 234:18

235:20 238:23

243:12,15,23

254:11,12 258:23

261:12 265:19

271:14 281:20

288:2,6 295:25

fit 126:18,21

128:3,4

fits 71:20 285:9 fitting 111:18 five 12:23

13:14,17,20 18:2,4

20:18 41:7,20

49:20 53:25

56:24 94:22 101:16

111:13 121:23

122:5 123:6 136:20

138:21 193:20

237:25 261:11

262:10,15,24

263:4,5,6 280:1

five...so 108:20

flaws 125:1

flip 60:5 105:14

floor 5:10 flow 45:23 fluid 70:16

flunk 49:15,16

focus 10:16 37:15

47:1 69:20 92:21

128:24

201:1,9,11,13

211:14,24 231:12

234:18 284:10

291:23

focused 10:8

54:16,21 55:22

57:22 93:19 98:6

110:9 114:6

115:4 158:6 205:17

227:7

focuses 34:13

focusing 12:21

20:22 105:18 114:1

239:2 279:6

291:8,23

foggy 252:7

folder 100:9

Folinsbee 146:13

148:3

folk 9:15

folks 67:18 258:18

Follinsbee 106:10 following... following 228:18 follow-up 212:10
for...for 227:11

230:17,20 253:17

for...for...or

276:2

for...he 131:7 force 102:6 forget 114:9 forgot 7:22

form 12:16 70:23

81:17 103:3 122:23

156:2 157:8,9

formal 2:7

format 20:19 177:16 formatting 251:6 formed 156:21

former 143:4 forms 103:6 formula 65:14

formulas 86:7,8,10 formulate 121:17 formulation 251:2

266:22

forth 41:19 85:13

161:4 214:4

fortunately 290:14

forty 119:24 forum 3:2 forward 5:24

14:14 25:19

60:18 65:11

84:11,13 146:11

219:16 259:17

foundation 156:2

173:1

fourth 36:14 190:11

234:17 240:25

fraction 231:22

232:10 249:22,24

287:18

fractions 24:8 frame 224:17 framework

10:11,14,15,19

14:10,12,15,20

15:2 24:23

33:10,14 49:1,2

50:15 55:23 56:1

57:23 111:11,19

120:14 184:12

266:18

framing 172:23

184:10,15

Frank 37:21 53:6

58:24 62:2

74:13,24 131:9





176:10,15,24

178:16 191:23

211:12,22 241:12

Frank's 54:6 freedom 60:10 frequency 230:9 frequently 101:4

236:13

friends 118:12 from...well 253:10 front 64:20

146:10 148:7

207:25 232:25

242:2 278:19,22

283:7

frustrated 207:7 frustrating 286:13 fugitive 226:3

283:13

full 34:15 37:23

102:3

fuller 133:13

fully 65:21 88:6

209:11

function 62:1,4

85:8 95:25 113:6

115:20 131:24

135:1 147:14

190:9,14,22 195:13

203:7 263:16

272:12 273:3

functions 235:2

Fundamentally

109:18

Furthermore 238:8

fusion 244:18

246:14

future 44:12 47:1

160:19 274:18

fuzzy 254:9

gas/particle 125:17

gaseous 28:23 29:21

gases 58:15,21

59:14 61:14 125:16

161:6

gather 170:3

gauntlet 138:16

Gaussian

266:18,19 268:19

269:11

gee 193:20

gender 149:14,21

general 29:15 30:14

44:8 47:10 57:9

91:17 92:4 122:5

123:4 127:14

169:12 182:24

187:9 203:11 223:6

278:11

generally 27:25

28:6 29:13 30:7

33:15 66:7 90:11

91:4 110:4 123:5

General's 11:23

51:1 53:25

generate 272:14

generated 155:20

227:20 252:13

generating 155:14 generation 155:1 generic 42:5,11

45:6,11 46:7,14

47:16

genes 153:12,15 genetic 153:14 genetics 153:6 geographic 227:8
geographically

265:25

geometries 264:18

160:3,22 183:12,13

188:6 189:2,15

194:20 196:9 275:1

285:25

George's 52:3 156:5

191:4

Georgia 7:8

gestalt 131:20

132:4,18 133:7

get...if 261:23 get...may 220:6 gets 65:25 66:5

103:23 108:22

192:23 205:4

240:25 283:7 285:8

getting 87:3 89:1

92:7,8,21 107:8

108:9,22 120:25

125:2 138:16

162:20 173:16

213:12 215:4

237:24 265:2 270:9

276:8 285:9

291:12,13 295:12

girls 149:16,19

GIS 298:3

given 3:3 40:12

60:12 69:17 102:10

133:17 146:5

158:10 193:24

201:11 230:10,15

236:25 241:13

242:25 247:8

248:12,15,18,23

256:12 267:8 287:6

Given...given

239:23

gives 27:4 63:9

132:7 208:25

267:10 268:20



 		geometry 261:25

giving 145:21

 	G

gap 13:10 58:8

130:2

gaps 59:22 130:20

gas 230:22

gas/gas 125:17

George 6:19 57:20

62:15 87:25

92:16 94:7

114:22 117:16

125:15 127:13

131:6 136:22

188:25 245:14

glad 55:18 126:20

172:2 194:17 265:9

glean 122:20

global 102:22,24

109:22 190:5





globally 105:6

god 90:18

going...along

221:25

gone 74:8 81:9

83:22 115:6

120:1 274:25

gonna 90:8

Good...good 214:6 good...I 269:25 good...it's 219:2 goodies 155:19

156:14

goodness 164:4

Gordon 6:7

116:17,18 142:9

158:2 161:19

191:11

210:12,19,22

213:10 286:25

287:14,18

got...major 264:19

gotten 52:8 93:23

174:19 271:17

Gra...oop 220:11

gradient 26:11,16

103:12 256:14,18

280:15

gradients 244:18

256:7 280:21,23

Graham 4:19

220:11,12

222:22,25 223:4

224:16 228:12

229:12 241:25

242:5,8 245:22

247:2,14,20

249:3,14,17

250:4,11,16

253:3,22

254:5,11 258:22

260:12,20

261:4,7,18

262:10,14,16,19

263:9,19 267:20

268:6,23 269:2

270:16 272:10

275:5,12,15,19,23

276:11,18,21

277:11,15 278:1

280:8 281:22 282:3

286:19 287:8,17,21

288:23 293:13,23

297:1,17

granted 44:3

graph 23:1 39:9

232:23,25

grasp 196:1

gray 159:15 170:6

great 11:6 55:25

56:11 60:19 63:8

67:1 69:13 70:8

85:5 86:3 101:25

172:8 261:21

262:23 283:3

greater 11:17 65:19

69:24 75:12

230:5 231:24

233:10 245:1 259:7

greatest 255:21 greatly 5:1 37:24 green 178:2 greeted 26:6

grid 255:9 grill 219:2,3 gross 246:12 ground 285:18

grounded 235:16

grounding 234:24

235:7 236:25

group 26:18 52:5

135:2,21 205:6

219:6 220:9 233:21

grouped 228:6 grouping 32:2 groups 48:23

135:4 150:18,25

151:8,12 227:5

270:20

grows 276:6

growth 113:6 115:19

190:14 203:7

guess 5:4 32:8

38:10 40:21,22

49:3 60:16 64:6

67:17 68:24

74:4,25 92:12

98:25 106:14 109:6

111:8,21

112:16,19,25

115:11 116:19

120:17 124:23

126:22 147:21

151:9 158:2,21

178:12,21 183:8

190:7 227:2 230:21

247:22 249:5

252:1,11,22

254:7 257:3 264:15

265:9,13 269:21

277:16 283:15

284:7 286:25

288:19 289:5

guidance 158:17

179:4

guide 50:5

guidelines 10:19

12:8 13:16 158:16

guilty 62:20

guys 182:2,19

210:13

H

had...had 262:6 half-hour 143:5 hand 84:23 167:15

213:9 214:8

278:7,24 281:5

handed 128:3

handful 67:19

252:21 254:3

handle 155:23

189:16

handout 96:13

hands 218:21 294:5

hang 148:1

happen 90:21

117:8 217:16

289:17

happened 47:24

191:16 295:14





happens 245:17

289:15

happy 147:18

hard 5:18,20

60:22 100:18 105:3

113:11 120:12,24

123:7 126:11 133:3

166:11 181:16

191:14 192:8,18

247:15

270:3,7,11 284:4

hardly 8:3 122:12

164:14

Harvey 4:19 220:8

273:8

has...has 287:1 has...I 278:7 hat 148:2

hate 214:13

Hattis 6:11

54:14,16,20 134:19

141:13 202:20

204:13,18

209:10,23

210:1,4

249:5,16,18 263:25

281:6,12,14,19,25

282:10,15 287:15

have...really

252:23

haven't 7:20

15:19 53:9,20 81:9

140:1 184:9 268:23

having 8:10 45:20

48:25 50:14

52:21 82:12

83:16 87:22

89:22 96:3 102:3

135:14 138:5

166:11 167:19,21

182:2,20 188:4

194:17 197:4

201:12

hazard 12:12

head 4:18 179:19

269:14

headed 102:18

health 4:8 5:17

6:22 7:16 11:4

13:3,8 16:4 21:8

26:24 27:3,6

34:17,23 35:21

36:4,13,18 37:8

40:18 76:22 77:3

80:18,21 81:1

85:16 87:12,23

98:3 102:16

103:1 104:6 105:16

111:15,21 113:4,17

114:2,3,21,25

125:10 130:24

149:2

158:5,11,20

159:8,13,24

162:1 168:18,20

174:24 191:1 193:1

194:25 195:6,22,24

196:16 200:22

201:5 204:16,19,21

205:3,4 210:8

212:17,18,22,24

214:6 223:16

231:10,23 232:6,12

242:20 281:11

287:22 289:11,20

291:18 292:22

healthy 95:20,25

119:11 187:7

hear 8:3,15

9:4,5,15 15:18

33:3 69:11 94:4

110:16 122:12,13

124:1 139:15

140:19,20 144:19

163:20 164:8,11,14

171:7,24 175:7

176:19,23 181:3

199:13 259:21

265:13,14

297:12,14,15

heard 40:19 46:22

88:23 130:22

166:19 172:16

259:22

hearing 8:11 166:12

182:3,21 293:9

heart 61:6 150:21

192:24 194:18

293:15

heavily 11:22

201:15

height 117:10

held 41:20 84:3

Hello 259:20

help 8:14,18

11:19 16:3 55:20

69:3,4 83:25 87:23

109:17 121:22

134:1 183:1 194:11

208:7 214:24

270:25 278:19

292:25

helped 37:24 270:12 helpers 100:7 helpful 26:1

64:10 68:10,15

118:18 134:1,15

136:3 180:1

209:9,11 251:13

278:12 286:12

292:23 295:1,18

helping 112:18

helps 11:18

Henderson 2:21 3:18

4:3 5:10,15,16

6:23 7:3,17,22

8:1,6,23

9:2,5,8,14,17,20

22:24 23:4

25:21,25 31:6,23

32:18 37:19

39:2,10 40:7 44:17

45:25

46:4,15,18,20

52:10,15

53:2,4,6,8

54:12,15 55:17

57:19 60:19

62:14 67:14

69:8,11 71:12,16

72:13 74:9,12,24

76:10 79:10





84:21 87:8,24

92:13 93:6,9,12

95:5 97:20 99:21

100:2,13 102:14

109:25

110:13,18,22

114:13 115:8

116:16 118:9 120:8

122:12 123:20,25

124:2 127:12 131:6

132:17 133:6

134:18 136:1

138:14 139:2,14,21

140:7,12,24 141:11

143:15 144:20

146:18 147:3

148:10 150:9

153:24 154:3

156:19

157:1,6,22 159:4

160:2,21 161:18

162:19

163:12,19,25

164:4,7

166:7,13,16,19

169:18,19

171:13,16,20,25

174:20 176:5,23

178:16 180:9

182:18,22 183:12

184:17 185:6,15

186:1,14

188:5,20,24

189:14,21

191:9,23,25

192:2,4,12,18

194:4,16,22

195:14,17,19 197:2

201:25

202:14,16,19

204:11 205:14,23

206:25 207:6,14

210:10 211:12,22

212:8 213:1,4,8

214:5,20 215:14,18

216:4,11,16,18

217:23

218:12,16,20,24

219:1,5,9,12,20

220:3,15,22

228:8 229:9

233:13,16 239:21

241:11

242:10,13,17

251:22 254:19,22

258:17,21

259:13,21 260:8,17

262:4 263:2 264:12

265:12 269:18

274:22 278:6,23

279:2,22

281:4,8,13,18

282:24 286:22

287:23 290:25

292:3 294:4,16

295:4,7,14 296:5

297:14 298:7,17

here's 101:20

175:18,20,21 176:3

200:8,18 227:19

232:9

he's 39:14 60:16

181:4 215:11 221:4

hesitate 96:2

heterogeneity 35:12

Heuss 32:21

33:1,4,5

37:20,22 38:7

39:11

hey 220:11 253:7 hidden 206:17 high 12:21 34:22

51:18,21 66:12

77:4,14,15 79:22

82:8 103:4 113:7

114:19 152:17

154:14,15,16 161:8

162:7 190:25 193:5

196:7

203:4,15,21

204:1,9

209:13,24 227:9

260:5 261:2

265:2 273:14

274:19 291:6

higher 17:3 58:11

66:14 77:7

104:22 107:18

114:22 119:4

123:17 136:25

138:3 149:19

186:25 197:21

230:1 231:24

236:13 253:23

267:1 280:4

highest 43:19 236:1

247:9

highlight 64:10

208:4

highlighted 34:2

173:22 174:21

highly 28:2 37:17

85:8 239:14

highway 161:8

highways 161:1,15

Hill 12:4,6

49:4,9,11 52:14,18

53:7,11,20 57:11

59:24 100:16

101:25 103:9,14

106:19 108:2 111:4

129:14 131:14

137:8 138:4

hinder 35:3 hinges 121:2 histamine 142:4 history 227:6

hit 215:16 232:18 holes 151:14 holistic 36:2 holistically 38:17 home
28:6 275:7 hope 25:18 32:22

171:25 172:1 179:3

215:2 216:22

217:16,17 267:17

hoped 178:23 179:12

hopefully 44:9

121:5 259:20

hoping 179:17

216:18 268:25

Hopkins 6:21





hospital 22:16,21

23:19 37:10

90:17,19 208:13

hospitalization

152:13

host 87:2 132:20

hot 90:5,13

91:3,8,18 92:9

245:7,10,15

264:8,16 265:8

284:9 285:10,14

hour 17:2 30:11

78:14 108:12,23

119:24 137:18

179:10 227:11

247:25 287:15

hour...of 282:18

hourly 85:11

171:1 199:19

230:6,17 261:10

282:16,18 291:3,11

hours 47:17 81:19

137:18 179:10

house 101:13

Howard 39:24

How's 9:24

huge 85:12 293:5

human 21:17 27:17

30:8,15 31:12

81:3,16,18,23,25

82:6,17 83:3

98:4,22 99:5

105:16 106:6

110:25 118:1,19,22

119:6 124:24

125:5,21,24

126:7,9 132:11

135:17,18,19,23

187:5,11 197:19

201:6 205:17 206:2

213:24 223:17

228:14 235:12

humans 80:22

82:20 99:8,18

114:20 125:20

137:24 187:5

hundred 50:7 62:5

72:2 104:3,4

106:16,17

107:13,14,15

108:10 109:1

131:25 132:1

136:14,15

hundred...two

136:14

hurt 84:11 hydrogen 157:12 hydrophobic 157:9

hyperresponsiveness

141:8,13

228:16,17,23 229:7

hyper- responsiveness

96:11,17,24

hypothetical 290:4

I

I...and 261:20

I...but 293:18

I...I 213:12

215:18,22 226:16

252:3,11 254:6

256:23

259:16,19,24

263:20 264:15

265:9 268:23 281:6

286:19 287:8 289:4

297:9

I...I'm 280:8

I...my 259:25

I...okay 278:7

I...that 234:6

I...we 270:5

I...yeah 264:14

I...you 234:16

I.D 12:12

ICF 220:14

I'd 2:9,18

3:16,20 4:2

5:9,11,25 26:10

44:15,21 48:16

64:8 65:22 67:3,17

86:7,9 111:17

127:24 133:13,19

141:2 233:22

234:20 251:8

266:10 272:24

292:10

idea 27:21 29:12

70:8,24 74:6

86:9 91:25

122:25 133:6 154:9

156:1 160:5

171:8 184:19,20

216:5

ideas 27:18 31:2

103:10

identification

43:10 265:8

identified 40:10

44:12 45:18 137:25

150:21 223:19

229:3 235:11 236:8

262:11 279:25

identifies 205:3,6

identify 43:21 44:8

59:19 75:5 95:16

160:16 182:11,13

183:4 224:4

244:8,16 285:10

286:12

identifying 10:9

44:1 160:7 205:6

286:17

idiot 182:16

If...do 249:15

If...in 294:9 if...not 287:24 ignore 60:10 115:1 ignored 102:9 ignores
128:22

Ila 4:12 9:15 52:17

100:13 101:23

122:12,14

182:22,23 210:24

Ila's 49:19

I'll 20:13 27:18

39:18 47:12

60:20 66:8 87:9

90:16 110:12 115:6

131:10 140:14





143:14 151:10,20

154:5 156:18

164:12 169:11,17

172:7 178:22

186:16 188:5

202:19 218:7

235:20 249:6

266:12 279:4,17

282:22

I'll...I'll 229:20

illnesses...oh

115:16

illuminating 71:11 illustrate 146:13 illustrated

18:19,23 104:17

292:7

illustrates 27:21 illustrative 231:20 illustrative...an

231:20

I'm 6:15 10:16

14:14,25

15:21,24,25

20:15 23:6 24:19

26:3,15,17 31:25

32:6,12,16,22

39:19 40:20

43:20 45:5 50:10

52:16,19 53:11

61:1,19 62:2 65:25

71:3,20 72:2,23

75:2 78:22 80:2,24

82:17,22 83:5

98:11,15 99:19

105:18 106:14

107:11,20 108:15

109:8 115:13

122:10,13 123:24

126:20 129:18

136:13,21,25

137:20

138:1,3,8,10 139:6

140:7,23,24

141:1 146:7 147:24

155:2,3 161:13

163:25 164:7 165:6

166:25 167:1

170:20 171:15

172:2,15,24

173:5 176:8

178:3 179:15 181:5

184:24 186:2

187:22 189:24

191:12 192:4

194:16,19 195:17

197:4 198:1,20

200:12 201:10

206:1 207:2

209:3 211:7,10

213:5 215:13

217:14 218:14

220:12,19,25

221:4,8 222:25

224:14 234:9,10,18

236:22 240:16

241:25 245:22

246:1 247:3,14

253:16,22 256:8

262:7 265:9,23

266:1,13 267:8,9

268:7,15 269:25

271:14 274:5

275:3,17 277:16

279:3,24 280:9

281:6,10

286:1,24

289:5,20

293:9,10 294:12

I'm...but 262:24

I'm...if 267:4 imagination 126:20 immediately 278:16 immune 82:13 99:2

115:22 116:5,12

immunity 132:20

impact 77:23,25

158:22 212:17,23

impairment 203:5 impairments 203:7 imparts 151:25 implausibly 35:10

37:12

implemented 298:5

implicate 34:5

implication 35:18

77:13

implications 85:5

89:23 176:1 243:3

implicit 81:1 implicitly 249:8 implied 249:15 implies 239:3 implying
78:20 importance 85:21

146:24 153:14

important 5:5 10:15

14:13 20:2 34:10

49:21 55:4 66:1

81:7 83:9 84:3

85:3 88:13 93:21

94:10 98:5 102:8

113:6 119:5,7

121:19 123:18

129:14,25 130:22

131:5 132:8 133:16

134:12,20 135:24

145:14,24,25

146:16 152:19

153:4,12 159:24

160:17 162:1

169:14 171:2,9

178:14 180:25

183:19 196:17,25

203:16 205:2

239:24 251:11,12

258:12 267:16

283:20 285:24

292:18 296:22

impossible 156:25

impressed 9:22

147:24,25 266:6

288:3

impression 65:25

66:21 189:1

202:1 207:16,19

250:23

impressive 262:21 imprints 85:15 improve 40:2 47:1

64:8 201:20





202:1 225:12 267:5

293:2 297:5

improved 5:22 33:12

40:9 47:12 63:23

64:4 67:17 74:5

79:13 93:18 102:20

110:5 118:16 124:6

178:19,20

improvement 9:22

32:22 33:6 55:25

76:13 154:7

improvements 28:4

33:8 86:1

in...in 228:12

230:5 280:6 289:3

inadequate 13:2

25:4,13,16 243:18

inadequately 43:9

in-and-out 211:14

inappropriate 177:3

255:3

inaudible 218:25

Inc 32:23

incidence 149:16,19

190:9,17 203:14

204:9

include 4:10 6:3

13:6 80:8 90:20,23

91:13 189:8 222:15

266:20 269:16

277:13

included 13:7 29:20

33:21 34:2 40:10

73:1 90:25 97:17

118:17 134:17

153:3,22 154:11

224:23 238:20

273:14

includes 148:15

231:11

including 4:18

24:15 25:11 27:1

28:24 29:4 30:1

43:9 78:7 79:1

225:25 227:3

261:14 276:12

inclusive 79:14

incomplete 27:4

56:4 60:7

inconsistencies

239:11

inconsistent

37:14 106:12 130:3

135:16 144:24

149:22 190:19

239:14

incorporate 5:7

11:1 68:4

incorporated

68:13 69:3 101:2

129:22 292:16

incorporating 67:18

85:21

incorrect 44:22

244:10

increase 96:9

97:4,5,8 142:14

143:13 145:9

155:16

246:2,4,11 249:20

increased 24:12

95:24 115:23 116:5

147:14 150:24

151:2 156:16

increases 41:12

96:10 107:20

increasing 152:17

243:19

indeed 38:22

142:16,24 213:24

independent 4:7

21:14,21 24:3,17

107:25 131:12

152:22 193:22

195:7,8 269:10

index 179:23 193:5

indicate 36:19

37:25 47:19 75:9

108:7 131:16

133:20 238:3

indicated 4:23

36:15 38:5 42:22

75:25 96:20,21

173:7

indicates 75:11

113:7 149:6 202:22

indicating 47:21

indication 75:22

132:7 144:2 158:22

209:15

indicator 51:15 indictment 59:8 individual

18:10,11,15,16

67:3 88:9,10,11,17

89:7 135:20 147:18

196:5,15 296:21

individually 89:8

216:2

individuals

89:3,4 95:20,25

97:3,14 147:13

149:6

230:7,23,24

231:16,22

232:5,21,22

233:9 237:4 274:18

275:16 276:13

277:13

individual's

89:10,11

indoor 20:3,11 28:5

87:17 88:18 230:21

231:1,3,6,11

232:4,13,14

233:3,4,10,11

246:25 248:11,23

249:2,8,11,20,23

250:1,9,14

indoor/outdoor

19:10

indoors 165:13 induce 157:16 industry 39:22 inequality 248:22 inert
156:8 infants 152:6,23 infection 115:23

116:5,13 152:7

infections 150:24

152:10 203:6





infectivity 82:12

infer 12:24,25

13:6,11

25:1,4,7,10,13,16

49:25 123:16

190:3,4

inference 86:24

203:19

inferences 200:20

infiltration

19:12 20:1,4

infirm 159:23

inflammation 137:16

152:2 155:4,6

inflammatory

95:19,24 96:4,8,14

108:7,8,25

137:14,21 157:5

influence 83:16 influenced 73:15 influenza 152:10 inform 31:3 33:25

38:9 39:5 121:14

122:21 134:4,7,8

169:21 170:9

199:12 211:1,5

222:1 224:10

information 15:10

16:13 17:22

45:1,23,24 56:11

59:21 61:16

67:24 68:4 71:3

72:3,11,24 73:11

87:13 99:14

110:6 121:11

144:5,14

160:8,18 167:23

168:15,17,22,24

170:18 171:3

179:1,13 180:1,6

182:11,13 185:9,24

186:20,22 187:1

189:10 198:10

209:7,8,18,21

210:21,23,24

211:2,3 218:9

226:21 236:2

237:16 238:2

241:3,5,20

242:21 251:10,14

253:18 261:15

270:13 271:18

273:9,16 274:18

279:14 283:14,16

284:22 293:21

294:11 297:19

298:4

informative 144:14 informing 43:3 ingredient 286:3 inhalation 228:18
inhaled 81:14,22

95:25 155:18

inhaling 157:20

initial 64:23 124:5

244:9

initially 46:14 initiate 136:12 inner 276:22

input 217:3 225:8

226:9 260:15

263:10,23 265:13

inputs 261:14 insight 171:4 instance 56:18,22

57:11 78:24

86:17 99:15 135:14

214:23 251:3

instantly 14:3

instead 200:5

247:23

Institute 7:14,16

11:5 39:14,20

105:12

institution

107:23 109:6 111:3

137:3

insufficient

36:1,16,19 40:16

integral 70:12

integrate 103:24

116:7 121:12

201:18,21 207:12

integrated 31:3

42:8,10 50:20

111:23 116:15

169:22 186:18

188:8,14

integrating 86:3

87:12 115:17 174:1

integration 20:22

26:23 94:14 102:16

114:15 116:23

118:1 125:5 184:13

213:11 246:7

integrative

130:12 172:19,20

173:14,24,25

178:13

integrity 41:14 intend 146:5 207:9 intended

129:13,14 168:13

199:12

intensity 69:22

intent 11:15 169:22

183:11 199:8

intention 169:6

290:6

interaction 4:15,20

58:15 125:16,17

129:19

interactions 61:5,8

interest 69:20

120:19,23 121:13

122:1 267:12

282:22

interested 5:2

88:14 115:9

186:2 297:10

interesting

73:14,25 74:3

124:10 258:7

259:3,18

interestingly 28:9

interference

66:5,13 77:14,23

interferences

17:5,8,9 65:24

66:1,11,24 94:19

internal 80:13





internally 10:20

13:18

International

220:14 233:19

Internet 59:4

interpret 105:21

158:17

interpretation

27:10 34:9 35:3

36:24 60:4 72:4

73:1,15 85:2,6

143:18 158:9 174:6

193:9 234:25

interpretations

71:5 85:15

interpreted 19:24

193:18 205:16,21

234:15

interpreting

72:16 87:12,23

133:23 181:14

187:19 192:24

193:2

interrelate 196:12 inter-species 99:15 interval 19:21

144:8 292:12,13,23

intervals 293:5

294:1

intervention 119:23 intractable 70:17 introduce 2:19

6:1,24 26:10 220:7

introduced 7:20 introducing 8:21 introduction 7:18

31:1 33:9 62:25

invent 11:12

inventory 244:10

283:12,22,24,25

284:13

investigate 293:14

investigated

151:7 152:4

investigation

153:14

investigator 58:23

investigators

145:16 193:1

investment

113:16,17

invoked 255:6 involvement 4:24 involves 3:22

36:5 46:7

IOM 11:23,24 50:4

53:9,17

IRM 13:4,8 irrelevant 82:7 irrespective

12:17 90:21

is...at 224:12

is...is 232:6

233:10 261:18

263:10

is...is...I 270:18

IS...the 221:25 is...they 273:9 is...this 115:14 is...you 258:13

ISA 3:24 4:17

5:21 7:19 26:21

29:12,18

33:19,22 34:7,12

35:1,13,24

36:10,14 38:8

39:4,23

40:2,5,14

45:10,19,20 46:8

55:10 64:24

65:16 67:5,12 68:7

70:1,14,22 94:14

118:8 122:21

139:18 148:15

167:3 168:12

169:3,4 173:2

174:23 184:25

185:5 197:15

201:2,3,11

205:16,20,25

207:20 208:1

210:13 221:25

228:12,13,19,25

229:5 235:11,21,22

258:10 270:20,24

271:8 295:11

ISA's 40:17

isn't 57:6 138:24

168:21 180:23

182:6 184:11

188:7,8,13,14

197:11 226:18

241:20 271:22

issuance 42:7,14

issue 16:6,8

17:13 19:2,5

20:6,24 49:17

50:14 51:6,23

57:16 65:16,23

67:6,8 74:18 76:18

77:22 84:2

88:3,7 94:4,13

97:19,24 98:1

109:9 115:12

120:14 125:2,19

126:4,22 133:14

154:15 155:1,21

156:7 159:8

161:20,24 168:12

169:14 179:20

180:16 183:19

193:11 195:2,22

196:1,22 209:12

223:23 234:19,20

252:4,6 263:10

284:24

285:2,5,24 286:1

289:6

issues 15:25 16:3

17:8 20:8 34:8

35:2 36:20 40:24

41:21 44:7

49:1,3 64:17

70:5 81:4,7

93:21,23 95:11

99:19 103:16 125:8

126:10 159:12

164:14 167:4 168:8

170:16 183:21

234:9 252:1 257:17

262:1 285:3,13

295:17





issues...it's

121:16

it...because 256:4 it...I 262:20 it...it 267:13

289:24

it...it's 260:3 it...they're 107:25 item 68:17

it'll 80:20

Ito 35:6

it's 2:7 5:5 9:20

11:24 12:14

14:13 21:9 23:2

32:21 34:10

35:19 42:24

43:16 45:16

46:11,12,24,25

47:19 49:18,19

50:20 51:15

53:10,20,22

58:19,22 59:8 60:5

64:14 65:20,24

66:4 67:25 69:18

75:1 77:8 79:16,25

80:20 81:6 82:7

83:18 85:3,19 87:3

89:18 90:6,11

91:5,17 92:10

93:3,24 94:10

96:22 97:6,21 98:5

100:2,18 103:1,7

105:20,21 107:5,24

108:21 109:5

112:15 113:3,11

114:3,7 115:16

116:1 117:15,16,17

118:20 119:7

120:12,24 123:7,21

124:6,18 125:12,16

126:6 128:11,20

129:13 130:3

131:25 133:3

134:12,19

135:21,24 136:16

140:3 142:20

145:24 146:16,20

147:10 149:24

152:3 154:25

155:21 157:25

159:11,15 163:15

165:10,21,22

166:21 167:25

168:9,21 170:13,19

171:19 172:1,18,19

173:12 175:17,19

176:6,8,13

178:14 179:1,13

181:10,16,18

183:7,19,23,24

184:19 185:12

187:2,5,6,24 188:3

189:14,16,17

191:14,20,22

192:7,17 197:24

198:11 199:6

200:8,16 203:12,25

204:1,2 205:15

207:7 209:11,13

211:3 212:13,14

213:22 214:19

221:24 223:6

225:19 231:3

233:16 234:1

237:13 238:18

240:17 247:14

249:19 250:11

252:8 253:6

256:4,24

257:5,13,24 258:14

261:20 263:13,20

264:18 265:4,22

266:2,3,18

267:11 268:10

269:10

272:10,11,12

273:3,4,16

274:1,12,20,23

275:12,17

276:5,18,20 277:25

279:10 281:2 284:4

285:18 286:12

287:1,8,10,13,18

288:22 292:14,18

293:13,15 298:3

it's...for 285:18

it's...I 288:5 it's...it 258:5 it's...it's 214:3

254:22 256:23

265:25 286:3 291:4

it's...it's... it's 280:17 it's...the 105:20

245:19

it's...this 223:5

it's...what 282:12

I've 12:7 54:21

67:2 86:11 88:23

130:22 161:4

164:21 166:17

171:18 193:11

205:20 215:23

216:3,11 234:4,6

235:19,24 242:25

249:16 256:8,9

268:10

J

James 7:11

95:9,10 99:11

102:18 109:25

110:23 111:7

112:14 120:11,25

122:7 123:22

127:25 138:15

139:6 140:13

147:22 148:6,7

154:16 174:19

181:4 197:2

202:1 205:14 207:1

210:19 213:12

292:3

James's 138:24

Jenkins 4:19

220:10,19,23

222:24 223:1,5

260:14,19 289:6

294:15

Jewish 7:12

Jim 7:4 31:23 79:11

84:21 85:3





207:16 269:19

274:22 292:3

job 2:17 53:10

57:25 58:3 85:20

86:3 87:5 129:21

149:2

167:6,10,16 168:14

262:23 298:18

Joe...Joe 266:24

John 7:25 8:1

9:11 32:21 33:5

June 217:5

just...a 213:13

just...because

212:13

just...I 214:10

252:1 258:4 261:21

295:11

just...it 133:3 just...it's 265:4 just...it's...it

289:25

know...because

264:22

knowledge 37:1

59:22 60:9,11 91:5

152:3

known 34:21 59:13

81:20 198:11

Koran 163:1

Kotchmar 7:2

14:23 15:4 25:23

kudos 266:9

110:13,15 111:5

just...let 283:6	 	

115:8 116:16

124:12,19,23

126:17,20 143:15

147:2,3,6

148:11,17 150:9,12

158:1 159:7,18

174:20

188:10,13,17

194:21,22,23

just...we 276:25

justification

235:23 236:19

justified 56:4

176:7 236:20

justifies 88:20

justifying 145:21

197:20

L

L.A 266:2

lab 117:15

labeled 178:15

260:4

labs 117:16

lack 125:7,14

169:15 244:12

 		lags 45:9,10

195:20 213:3,4,8

Johns 6:21

John's 146:21

212:10

join 39:15 218:9

joining 194:20

Jon 6:21 47:6 57:15

63:13 100:15,24

111:8 120:9

122:10,15 125:8

150:1 172:3

192:1,2,19

194:24 196:2

278:6,7,23 283:8

Jon's 150:8 journey 275:4 judged 28:14 judging 12:3 judgment 53:15,22

93:22 94:15 95:3

190:2,6 194:14

224:10

judgments 192:10

judicious 193:8

July 219:14

jump 10:4

 	K

Karolinska 139:10

140:9

kazoo 176:7

Ken 127:1

Kenski 7:9 67:15

251:24 253:21,25

254:6,18

Kent 6:13 118:12

120:9 127:19

279:22

key 16:1,8

19:4,11 24:20,21

93:21 105:16

109:10 170:19

173:16 237:24

279:19 286:3 292:6

kids 152:16,23

kinds 54:7 168:15

170:15 255:14

Kinney 8:25

Kleeberger 7:15

150:11 159:7

km 264:6

knock 129:1

laid 102:4 121:22

130:12,17 177:10

252:8

Lake 7:9 lamp 160:16 land 285:12

landscape 177:16

language

11:6,7,12,13,16

20:19 24:22

laptop 295:21

298:16

large 18:6,20 20:23

28:15 54:24

77:2,18 92:20

157:10 161:5 246:6

257:18,25 273:12

280:14,16 283:19

293:8

larger 17:9,10

160:1 161:9 245:17

246:12 249:22

255:7

large-scale 280:21

Larson 8:4 9:13

69:9,10,13





71:15,21

72:9,17,20

73:3,5,20 74:2

254:20,21,24

258:20 259:16,23

260:25 261:6,17,20

263:8 264:14

280:11 281:2

297:3,12,15,21

last 10:18 11:22

19:8 20:6 30:17

37:7 39:12,24

57:22 58:16

70:19 80:7

120:13 127:14

154:22,24 158:4

159:2 161:19,20

173:15 191:12

194:8 201:2 223:14

250:20 295:6

296:18 297:15

lastly 27:14 30:6

31:1 184:7

late 149:18

later 5:24 51:5

56:19 63:7 66:18

75:14 87:11,23

94:22 110:12

112:17 172:1 175:1

177:6 185:1 234:16

249:6 254:10

257:22 260:9

273:11 282:23

latter 169:3 launch 102:21 laundry 181:17 lavage 157:16 law 2:24

lawyer 217:12

lay 133:4 170:4,7

272:9

laying 138:15

140:24 169:23

lays 10:12 70:8

160:19

lead 34:16 44:11

63:17,20 107:22

148:18 163:1

166:20 214:15

215:2,8 220:16

239:4 242:22

293:19 295:17

298:8

leading 24:16

leads 42:19

Leanne 144:21

146:18 147:6

160:3,21 186:1

192:25

leap 278:16

learn 129:4

learned 56:5 101:20

172:21

least 13:17 41:24

52:4 64:5,8

71:21 74:1 75:10

81:21 82:5 83:20

86:10 96:23 108:25

125:4 133:14

141:14 143:10

144:7,14

151:1,4,25

160:16,18 173:15

176:9 178:20

183:10 195:4

223:24 224:1

232:6,24 236:19

239:1 251:17

252:22 255:9

263:14 267:17

leave 219:10 276:13

278:7 298:13,16

leaving 218:19

278:3

led 47:6 111:9 legal 217:6 218:2 legend 177:23

287:9,24

legends 117:4 length 256:2 less 31:1 33:18

42:19 50:6,10

55:24 86:21 97:3,9

124:16 174:24

202:23 209:17

231:5 234:16

238:10

lesson 8:24

let's 15:8 47:4

50:7,8 56:16 63:19

89:11 101:7 102:14

104:14 105:23

108:12,19 128:11

131:23 132:20

140:22,23 176:9

183:25 265:13

278:8 281:20

letter 43:8 63:15

215:5,6

216:20,25

217:3,4,7 219:18

296:10,22

letting

168:15,16,18

level 9:24 16:9

31:9 32:3 50:1

61:21 88:11 89:7

96:3 98:17

103:22 105:1 106:1

107:5,10 108:18

109:15 111:2,18

122:23 126:1

133:9,10 135:16

136:10,11,13,25

137:2,6 138:12

152:4 158:23

161:22 168:22

169:10 175:12

176:2 197:23

202:23 203:4 204:1

208:25 234:20

236:23 237:5

240:15,18 256:6

264:10 284:20

285:18

289:11,15,20 291:3

292:15

level...I 136:5 leveling 244:22 levels 27:15

28:10 30:9,13

58:10,11





66:12,15 79:21

81:20 82:14

88:9,10 91:21 99:6

103:5 104:2,7,10

105:5 106:2,4,16

109:4 110:24

114:24 119:2,24

128:8 129:9 130:24

132:22 136:9

139:23,25 140:6,10

161:8 175:16,20

177:11 178:8 181:9

186:25 187:18

199:2,9,24,25

200:19,24

203:2,14,15 207:22

208:4 222:3,4,12

223:9 224:7,8

229:1 230:10

231:10,25 233:2

235:3,4,12

237:17 238:24,25

266:25 273:10,21

280:4,16 290:21

liable 60:9

Lianne 6:15 55:21

57:19 84:23 87:8

88:2,25 89:1

92:7,13 212:8,25

239:22 275:2 279:2

288:19 290:25

library 55:6 59:2

117:17

lies 50:11

light 68:19 79:6

86:2 152:15 204:6

likelihood 41:12

likely 12:25 13:7

25:1 37:16 50:1

51:19 65:6 66:3

77:14 113:9

125:9 150:18,20,25

151:11

153:16,17,21 160:6

202:11

Likewise 64:21

limit 36:21

limitation 57:1

284:14,19,24

limitations 35:15

40:16 65:6

78:10,15 168:17,19

limited 29:16 67:23

112:15 141:20

225:15 262:24

limiting 151:16

266:4 269:17

limits 97:14 174:18

line 2:9 16:16

31:8,19 32:21,25

41:19 92:14

105:7 107:4 131:23

133:1 148:15 172:8

173:23 174:7

204:17 205:1

206:12,13

221:2,9,10

233:23 255:3

274:25 277:16

line...for 233:23

line...you're

277:19

linear 180:25

263:16 274:1

linearity 273:24

lines 11:20 26:23

27:1 30:12,22 68:9

134:1

line-up 54:5 link 275:20 linkage 281:14

link-based 226:4

linked...roadway

226:1

linking 118:7 links 226:1,8 list 38:8 54:2

63:3,4 64:9 139:24

154:8 162:21

181:18 218:8

260:11 263:24

listed 18:4 81:23

147:10 196:6

199:11 234:7 248:8

294:8

listen 292:9

listened 215:22

294:18

listing 159:22

160:5 177:5

literally 104:2

literature 10:10,13

24:6 34:10

35:21,25 36:3

43:10 51:9 52:8

54:17,21

55:1,13,24

57:1,6 70:17,21

82:3 118:17

151:14,24 153:19

156:1 223:17,22

224:1,13 228:14,15

253:4,11,24

254:2 255:5

256:9 258:25 297:4

literature... there's 255:4 little 12:13

43:22 45:8,20

50:12 52:22,25

55:24 56:6 67:23

68:9,10 76:18 78:9

79:5 80:2 83:1

93:16,23 94:22

95:18,22 99:16

103:23 106:20

107:20 109:15

117:1 120:12,24

121:1 122:18

132:14 133:13

155:15 158:5

159:2,11,15

178:7,9 180:18

181:7,12 182:7

191:6 203:19

213:25 217:15

221:1 222:6 226:20

227:2 253:6

254:9 258:14

267:12 270:12

286:14,21





little...I 255:24

live 154:17

170:12 257:5,16

258:15 260:5

261:24 275:10

276:17,24 277:6

living 75:20

91:14

161:1,7,14,20

258:1,11 277:13

load 75:23

local 226:24 245:20

284:8

located 287:15

location 155:18

222:10 225:19

228:2,5 246:11

247:7

locations 29:24

208:16 225:25

226:6,22 227:7

228:4 229:25 240:8

246:13 257:4

locked 298:15 loctations 225:14 log 141:14,18 logic 102:1

138:13 161:3

197:16

logical 271:1 logistics 5:14 long 25:8,15

26:7,11,13,15

31:7,10,18

32:7,12,16 71:6

73:3,4,13,21,23

87:6 111:15 112:21

114:7,11

115:7,18 116:12

125:10 154:19

190:17

203:4,5,13

204:2,10,18,20,21

205:3 209:15

214:20

longer 64:7 110:9

162:21 163:16

174:14

263:14,20,21

longitudinal 226:13

232:20

long-term 69:21

70:3 173:9

189:25 190:18,22

282:12

look...fly 257:10

loop 112:12

Los 257:17 259:17

261:1 262:14

263:13 265:22

267:16 280:14

283:17

285:4,6,23,25

286:6 293:4

lose 142:24

loss 253:16

Lost 172:8

lot 11:10 27:25

46:22 56:3,4 61:12

64:14 68:13 74:8

80:21 83:23 86:3,5

92:20 94:2,17

101:4 106:25 110:6

118:21 128:6

132:24 142:22

157:23 169:25

178:25 179:12,22

180:1 195:18

201:14 236:12

240:14 245:4 255:5

257:13,14

277:3,5,20 281:7

285:6 286:14 287:5

288:3 290:14

293:21 296:23

298:3

lots 137:10

158:24 279:15

lot's 283:4 loud 8:7 louder 182:19

love 14:11 86:7

157:21 181:3

292:10

low 79:21 82:14

89:19 90:9 99:4

103:22 104:7

105:1,5

106:1,2,3 107:10

109:4,15 110:24

111:2 117:19 128:8

136:21 137:2 140:6

152:6,23 187:18

228:20 229:1 256:3

280:15

lower 12:21

109:23 114:24

117:15 119:2

138:7,12

139:5,8,20

140:10 197:23

208:16,17 230:1

232:7,17 235:10

256:21 288:12,14

290:16,17,20

lowest 32:10

137:4 181:9 186:19

231:23 235:11

236:1,7

Luben 6:25 14:18

15:9 20:13,14,15

23:2,5 39:4

luck 194:10

lunch 138:17,23

139:1

lung 50:9 59:15

62:1,4 98:18 113:6

115:19 128:7,16,17

130:6 131:24 135:1

147:14 155:20

156:10,13 161:7

190:8,14,22 195:13

203:7 273:3

M

ma'am 233:18

magnitude 96:8

145:5 187:19

286:15

main 33:21 36:16

57:18 85:23





208:1 228:13 243:6

246:17 251:5

270:14,23,24

mainly 110:22

224:20 234:9

maintaining 41:14

major 47:21 64:11

71:25 77:25

78:15 84:19 112:19

117:9 125:22

163:13 168:12

178:12 226:8

228:22 244:24

245:8,18 250:20

254:15 257:9,16,24

258:15 264:6,7

285:1

majority 22:17

23:11 145:7

makers 31:4

management 47:2

289:14

manner 41:14

73:19 74:1

Manufacturers 32:24

Marcus 41:10 margins 293:7 marked 140:16 markedly 37:9

102:20

marker 155:5

Mary 4:13 46:6

73:22 75:1

173:12 174:22

175:7 185:1

Massachusetts 26:17 massive 50:2 248:1 matched 273:11 material 48:11

99:16 127:10

materials 36:12

102:2 234:2 298:16

mathematically

247:1 248:5

matter 51:14 85:7

145:15 193:6 194:2

196:7,23 204:2

209:14 212:21

265:1,3

mattered 209:25 matters 92:16 204:2 max 231:5

maximum 17:2

30:11 66:11,16

208:19 230:13,23

277:7

may 2:5 27:13

41:3 59:11,12

60:13 62:17 68:9

71:8,9 74:10

77:3,18 81:7 87:17

92:10,19 96:15

102:5 118:10

119:18 123:3

124:16,25 126:23

127:3,6,7 133:18

134:8,9 135:7

143:25 149:8

151:21 154:17

155:16 165:12

166:3 167:19

177:1,2 178:15

180:12,16 184:9,25

185:9 195:23

196:12 204:13

216:21,22 239:14

241:3 267:13,14

269:25 270:1

273:5,13 274:23

280:25 285:13

291:7 293:6 296:18

298:1

maybe 51:13 53:10

60:5 67:9 72:13

73:6 81:16

82:23,25

83:18,22 86:11

89:22 93:4 99:21

110:1 111:24 115:6

119:8 121:19

124:18 138:16

162:14 163:9 174:7

176:13,16 179:13

181:5 183:2

203:10,12 207:1

214:24 220:6

241:7,23 251:16

252:7 266:10

268:16 271:4 287:9

292:5

maybe...you 293:6

McKreaner 30:2 me...there 257:24 me...to 256:16 mean 16:16

18:12,13,16

49:13,21 59:4

71:13,16

72:10,20 75:20

77:8 79:21 82:5,17

83:14 86:13 88:7

90:16 92:18 93:3

96:25 100:17

101:1,11,14,18

103:13 115:9 116:1

121:6,15 122:16

128:2 131:7 132:18

133:2,7 137:23

139:10,15 145:23

148:7 151:15

157:17 164:1 168:4

172:10,18 173:6

174:25 177:9 182:9

184:18 188:20

189:17 192:9 194:6

195:15,22

199:18,24

200:2,3 201:14

202:2,9 205:21

207:10,25

208:14,15 209:17

213:21 215:20

228:1 234:25 235:6

237:2 240:4

242:1 243:12

245:10 247:23,25

248:3,18

249:1,10,19

252:3,13 253:18

254:1 255:4

256:9,13

257:13,17,19,23

261:3 262:3,20





264:14 265:14

266:16,19 267:25

270:8 272:16

273:25 275:25

277:15 278:10

280:12 282:15,21

288:24 291:19,24

293:3 295:9

296:6,17 297:5

298:1

meaning 84:17 163:6 meaningful 104:25 means 3:1 39:6 54:2

78:13 135:3 141:23

148:8 165:22

169:16 173:11

180:22 209:15

211:7 227:25 230:1

264:24

meant

124:17,18,19 166:8

176:4 181:6 188:16

202:10 207:9 270:9

measurable 108:9 measure 108:24 measured 16:21

81:18 93:22

94:15 137:21

156:20,23 243:16

268:4

measurement 16:7,10

57:2 64:1 76:20,25

77:19 94:19 156:25

204:23 265:7

measurements 76:1

94:17 243:14

244:2,19 254:1

255:3 286:11

291:12

measuring 134:25

280:3

mechanism 101:7

127:9 132:12 187:3

mechanisms 121:3,11

123:16,18 126:18

157:8

mechanistic 186:25

Med 58:22

Medical 7:12

Medicine

6:8,10,20 11:5

55:6

meet 41:7 111:4

138:4 217:6,22

219:22 237:21

238:9 290:18 298:9

meeting 2:8,14

3:3,4,16,21,22

5:14,17 39:15

43:5,6 58:16 76:15

78:2 103:10

108:1 154:22,24

218:1 219:17

221:13 222:5,13

227:14,16 232:12

233:8 234:1,2

288:7 289:8,12

291:8 296:24

meetings 94:3

meets 103:13 106:18

132:6 138:22

154:19 289:17

member 62:19

members

3:12,13,15 4:4,6

5:2 39:21 217:22

218:7 233:23

memo 47:21 75:15

84:16

memorandum 70:25

71:17,22 72:6

73:17 229:22,24

254:17 256:20

259:12

memorized 61:2

memory 53:11

Meng 7:1 14:24

15:2,9,19,20

74:10,14

mentally 157:23

mention 11:3 22:3

51:19 62:23

74:15 228:3 230:21

mentioned 12:8

40:8,11 51:24

56:3,25 65:3 73:22

74:23 77:1 84:6

85:3,4 114:15

134:3 142:10,11

143:1 170:22

225:10 247:6

279:24 298:2

mentioning 155:12

Merely 30:18 merit 158:13 message 140:2 messes 79:19

met 10:19 290:5,11

Meta 106:9

meta-analysis

142:16 143:8

145:10,22 147:20

148:4

meta-regression

144:11

meteorology 256:12

meter 256:24

meters 67:5,7

72:2 75:17 161:7

254:15 256:2,11,21

257:9,16 258:24

259:8 267:14

method 16:11

83:12,22 97:12

255:25 269:23

271:23

methodology

270:7,10

methods 16:7

129:9 272:21

metric 28:16 metrics 30:12 metropolitan 16:22

Mexican 66:8

Mexico 66:8 113:4

191:1

mic 15:16,17 26:3

31:24 52:16 54:19

Michigan 7:9

micro 223:12

226:11,12





microphone 8:21

39:18 140:23

176:18

mics 8:14

middle 66:17,20

205:19 235:10

might...I 267:17 might...of 270:24 might...that 289:7 migration
211:14,24 mike 154:4 180:14

182:20 220:12

281:18

million 56:24

104:14 108:11,22

136:15

137:17,19,23 138:1

225:23 242:9

million...I'm

225:22

mind 17:4 80:25

132:23 186:15

228:8 232:11

269:22 288:22

295:12

minimize 196:22

minimum 82:18

mini-tables 131:14

minor 79:16,25

117:3,10 146:20

minority 135:15

162:12

minus 105:3 253:20

minuses 131:16

132:25

minute 23:6 99:25

108:19 222:7

minutes 3:4 27:18

29:11 99:24

108:10,13,17,18,21

109:1 112:3,6

117:25 128:11

132:1 137:19 229:2

mischaracterizes

36:14

mischaracterizing

36:23

misinterpreted

124:13

mislabeled 148:24 misleading 27:4 miss 81:7

missed 86:11 110:20 missed...I 110:20 missing 56:21 58:13

121:24

122:6,10,16,18,24

179:15 244:11

251:1 255:11

277:3,5,25

mistake 11:12 mitigating 274:19 mix 114:4 195:5 mixed 105:21 mixing 89:2
mixture 20:25

21:5,12 115:2

187:21,25

mo...the 269:21 mobile 64:14 226:4 mode 78:3,6,7,10,12

79:1,5

model 24:8,15

29:15,17 30:1

34:14,19,22 35:2

44:4 68:24 70:16

78:8,13,20,21 79:8

127:5 172:13

193:21 226:10

230:24 231:2,9

232:3

243:3,7,17,25

244:6,11,17,23,24

245:20 246:7 248:4

251:2,4

252:9,12,18,24

254:8 255:2,8,25

257:21,22 274:13

275:4 276:11

277:12 278:2,17

282:8 285:8 298:4

model...a 265:7 model...the 244:22 modeled 225:24

226:25 245:5,25

276:25 285:16

modelers 283:2

modeling 80:12

81:10 223:10

225:11 229:20

237:19,25 239:16

240:9,13 241:2,9

243:19 244:2

245:12,13

246:3,11,16 247:18

253:3 261:15 262:2

263:12 266:18,20

269:4 275:12 279:8

280:6 285:11

286:25 287:7

288:25 291:22

292:6

models 21:20

23:24 24:2

29:14,18 30:4

34:21 51:4,16

64:25 65:3,5

69:4 70:1,12

79:1 126:6,8

192:24 196:25

212:13 251:6 254:2

255:5,8,15

273:25 274:9

278:13

modest 134:24

modifications

56:5 57:12

modified 12:4,19

41:11 60:9

modify 218:1 296:21

moment 74:5

229:20 234:18

money 103:7

monitor 65:23

67:6 70:3 74:16,19

85:1 88:20 89:8

90:5 91:3,21

92:9,20

243:15,23

245:12,16,18,19

247:9 253:23





256:17,18 260:5

282:11,17,19

monitor...come

240:11

monitored 83:13

240:21 241:4

monitoring

17:16,23,25 18:2

69:17 85:21

monitors 16:23

18:2,5 65:19

67:4,6 71:1,7,9,24

72:11,18,22

73:14,23 75:5

76:7,8 88:21

90:7,14

91:8,9,18,23

223:10 224:23

225:2,6,15

226:7,23 228:3

240:7,8,17

243:8,9,11,12,22

246:5 252:21

254:4,13,14

256:6,20 259:11

281:16,21,23

282:1,5,7

month 90:25

months

42:7,10,12,16

morbidity 24:25

25:2,9,11 190:1

morning 2:6 3:10,20

5:4 15:20 39:13,19

66:18 100:5 114:15

196:11 270:18

298:10

mortality

move 15:8 22:14

24:10 25:8 26:2

47:4 54:13 60:18

63:19 76:23

77:10,17

84:11,13 96:12

102:15 107:18

140:23 160:14,15

219:16 221:17

240:12 243:21

255:12 270:19

moved 96:9

moving 69:20

96:16 257:2

much-improved 69:15

multi 20:7

multi-hour 62:8

multiple 181:2

193:12 195:25

210:7 232:22 233:1

multiply 248:15

292:21

multi-pollutant

21:20 24:1 29:15

30:1,4 46:23 51:11

94:1 189:12 192:24

194:18 195:2

multi-pollutants

94:5 186:21

189:5,9 214:22

multi-variable

193:13

mute 8:17 my...I 261:3 my...my 246:17

252:4

myself 83:19 170:20

264:17 293:18

nagging 79:17

NAQS 154:19

nasty 204:14

National 4:8,14

7:12,16 16:25

152:18

natural 36:12 56:24

nature 13:22

82:13 93:19

239:5 286:14

293:12

NCEA 6:25 7:1 15:21

46:6 52:11 73:4

74:15 100:5

118:11,12 139:15

162:23

nearby 218:6

near-road

18:23,24 65:18

68:23 69:7 280:23

near-roads 71:24

necessarily 40:12

42:18 76:7 97:17

112:22 117:12

126:13,14 138:9

163:8 165:16 190:5

209:22 224:20

226:18 284:9 297:6

necessary 54:10

204:22

negate 196:22

negative 28:1 33:23

35:11,13 61:11

96:22 106:13,23

107:7 134:10

141:5,6 147:11,17

negatives 107:3

NEI 284:19

23:9,10,15,21

 		neither 236:9

25:5,15 29:8

most...more 243:2 most-likely 60:3 mostly 110:9 132:21

251:24

motivated 288:19 motivation 291:20 mouse 126:7

 	N

N02 16:9

NAAQS 36:22

41:3,4,8 42:5,11

45:6,11 121:22

122:22 174:25

175:5,13 191:20

196:6 239:10

network 17:17,23

69:18,22 73:24

neutrophils 155:21 nevertheless 112:12 next...next 232:15 nice 65:15

68:5,13 75:16

76:23 78:7 79:7





82:2 96:23 118:2

140:3 209:18 251:3

258:5 298:2

nice...I 298:1

nicely 76:15 77:6

121:6

nine 49:10 ninety 3:5 nitrate 155:6

157:8,10,13

nitrated 157:17 nitrates 77:13 nitrating 157:14 nitration 155:14 nitric
17:6 66:13

77:12 157:4

201:5 280:5

nitrogen 2:4,12

4:1,9,10,11

39:23 41:3,18,22

51:9 52:6 105:17

130:24 155:2

173:10

nitrogrn 85:8

NNMAPS 35:6

NO2 4:1

16:14,19,21,22

17:11,14,16,18,19,

23,24,25

18:2,3,5,6,9,13,14

,20,22,25 19:10

20:8,24,25

21:3,4,9,11,14,22

22:13 23:14

24:9,15,17

27:2,5,8,11,13,16,

22,24

28:10,19,22,25

29:3,7,12,13

30:4,8,11 31:15

34:6,7,24 35:20,25

36:9,11 37:6,16

38:15,24 39:1

51:14,20 59:9 61:7

66:4,15 74:21

76:6,19,25

77:4,7,12,15,19

78:16,23 83:7 88:9

99:2 111:22 112:21

113:5,8,11,20

114:10,24

115:18,22 116:6

119:18 120:3

128:7,9 129:7

130:5 133:24

141:10 142:18

143:13 147:13

149:3,9 151:5,22

152:4

153:2,9,16,20

154:17

155:7,8,14,15,18,1

9 156:9,12,13,21

157:9,12,13,19

158:6,25 161:17

162:2 170:9 175:13

179:21 180:24

183:18,24 188:4

190:1,13

192:11,12,23

193:21 195:7

196:25 201:23

202:12,15,22

211:14,15,21,24,25

212:6,11,16,23

213:15,17,19,20,23

214:6,12,18

222:4,12 224:23

225:1 235:13 239:4

244:12 266:22

267:2 273:5,10

280:14,22 290:15

NO2s 191:15

NOAEL 32:15

nobody 75:20 91:5

noise 88:17

non-asthmatic

137:24

non-cancer 13:19

none 29:22,23

199:15 286:14

non-Hispanic

162:9,16

non-Hispanics

162:10

non-linear 274:9

non-MSA 228:7

non-response 128:14 nonspecific 137:24 non-specific 36:8

141:15 142:7,18

228:24 229:7

non-specific... inhalation 228:24 non-supportive

30:23

nor 145:6

normal 126:7 normally 141:14,18 normals 104:21 not...based 277:21
not...I 249:19 not...why 253:25 note 54:23 67:3

108:3 134:22

135:19 141:6

229:21 233:22

noted 11:21 33:7

36:20 43:11 89:4

95:22 141:13

notes 33:19,23,24

67:4 215:23 229:13

nothing 93:25

98:2 116:10

177:8 235:21

274:13

notice 3:2 221:18

236:5 281:15

noticed 229:22

not-so-careful

188:25

November 43:8

221:16

NOx 2:12 5:17 14:17

26:21,23 40:5,18

41:19 43:9 61:20

80:7 83:7 87:19

98:17 280:4

NOx)PRIMARY 2:4

NOz 67:22





NRC 125:11

nuance 56:7

nuanced 13:13

Nugent 2:6,9,16

5:11 8:8 9:10

15:15 26:6 32:20

33:2 39:12 95:18

192:5,13 216:13

218:5,14,18,25

219:3 233:18

242:15 296:2

298:15

null 33:24 131:23

numerous 298:5

NYU 6:7,19

occurred 267:24

287:10

occurrence 230:9,14

235:13

occurrences

229:14,17 259:6

287:12

occurring 131:25

187:17 236:13

268:9

occurs 34:7 268:13

o'clock 138:17

October 39:24

oddly 68:5

of...because 247:23

46:15,19,21

52:12 53:2,4

54:12,15 59:19

63:13,19 66:5

69:11

74:12,13,14 75:8

76:10 84:24

92:13 95:5 100:2

101:22 102:12,14

103:20

110:13,17,19

112:12 116:16

118:13 123:25

124:3,21 127:16

130:19 135:6



 		of...I 281:15

138:25 139:2,21

 	O

OAQPS 45:1 170:1

198:17 200:3,10

203:11,17 205:21

206:24 211:2,11

220:9,12

ob...obvious 265:23 obesity 151:23,25 objective 40:4 obs...that 224:8
obscure 135:3 observable 135:13 observational

35:8 36:3 57:5

60:8 121:4 135:23

observationally

123:19

observations 123:13

134:21 135:8

obtain 253:10 obtained 221:23 obvious 80:18 209:5 obviously 15:6

38:10 44:24

184:2 216:1 221:20

occasional 203:3

204:20

occur 21:1 103:5

130:25 198:4

212:12,14 226:12

228:1 259:7 297:9

of...I'm 98:25

of...of 228:18

254:13 255:4 260:1

of...there 286:8 of...there's 217:12 of...you 94:6

offer 289:22 offering 26:8 office 2:2,17,20

4:13,17 15:5 220:6

294:7

Officer 2:11

off-road 271:18

283:19

offset 256:1

oh 14:25 31:25

32:11 53:6 59:9

61:11 67:3 124:2

128:3 163:20,25

164:4 166:15

171:25 172:9

177:25 194:22

195:6,17 219:12

222:24 240:7 247:3

268:25 269:2

275:24 281:25

okay 6:23 7:24 8:23

9:8,14 10:2 13:5

23:4 25:25

32:11,15,18

39:11 45:25

140:22 141:11

143:15 144:20,21

146:18 147:3 150:9

157:23 159:4 160:2

161:18 162:19

163:12 166:9

171:25 173:12

176:5,25 178:12,18

182:18 183:12

184:14 186:1 189:1

191:23 197:13

200:23

202:14,17,24

203:16 204:11

206:22,25 210:10

211:12,22 215:14

216:10 219:6,20,23

220:4,16,22,25

221:19 222:18

223:4 224:14

233:16 241:11,22

242:13,17,24

244:20 245:21

249:1,4,17 250:4

251:22 254:6,20,22

258:21 259:13

260:8 262:4

263:2 264:12

268:2,8

269:18,19 274:22

275:24 277:18

279:2 281:4,13





282:10 285:21

286:22 287:14,24

292:3 294:3,4

295:4 296:25 298:7

old 59:2 117:17

146:12

older 22:22 149:8

150:7 165:12

on...I 115:9 256:9 on...is 115:13 on...on 212:10

224:11 286:7

one...one 255:23 one...the 113:23 ones 72:11,12

75:6 79:1 82:11

105:5,11 121:7

130:4 160:8 181:8

ongoing 269:23

on-road 64:14 65:18

68:20,23 69:6

225:7 227:15 230:2

252:9,10,16,17,22

253:12,19 254:8

256:6,15 271:18

272:4

on-roadway 266:14

onto 206:7

on-vehicle...on- road 283:14 onward 270:22

Oops 219:9

open 62:16 63:11

84:22 131:9 180:10

217:14 274:23

operate 3:2 operates 2:24 operation 90:25 operational 298:4

opinion 40:12 77:24

263:7

opportunities 3:8

opportunity 4:2 5:6

39:22 43:22

44:14 59:21

opposed 111:12

180:24 194:1

240:22

optimal 161:6 options 70:9 or...or 283:11 oral 3:6,8,9 orally 172:6
oranges 272:11

order 234:23 239:17

256:10

ordered 46:11

Oreck 117:13 142:10

143:9

organic 77:13 organics 29:23 organized 180:3

185:9

orientation 240:13

original 52:20

107:3 118:16

229:23

originally 264:5

OS 80:7

others 33:24 34:3

40:7 42:2 47:11

48:2 52:5 58:2

71:25 76:9 131:9

135:10,12 148:11

160:2 206:24 215:4

216:8 251:12 259:5

265:16 268:8

273:14

otherwise 42:3

76:15 149:25 206:6

237:3 288:15

ought 38:22 75:22

132:8 146:9

151:1 172:14

238:17,19 242:4

our...are 95:6 out...well 264:3 outcome 113:21

135:14 184:3,5

outcomes 25:12

204:16,19,21,24

210:9

outdoor 87:17

246:21,25

249:9,12,21

outer 255:17

outlined 97:25

123:22

output 231:8

outset 284:2

outside 22:5 200:12

253:17 276:13,15

277:7,23 287:5

overall 18:13 22:11

26:21 28:2 81:22

83:1 85:25

102:3,22 111:21

178:19 190:2

overarching 20:23 overcoming 79:22 overkill 79:2 overlap 226:20 overly
37:4 overnight 295:15,23

298:14

overprediction

243:22

overseas 215:11

oversimplification

77:8

overstatement 239:5 overt 108:24 overview 10:3

220:17 221:3

222:18 278:12

overwhelm 280:23 oxidant 155:17 oxidation 155:6 oxide 52:6 157:4

173:10

oxides 2:4,12

3:25 4:9,10

39:23 41:2,18,21

85:7 130:24 201:5

Oz 283:3 286:25

ozone 29:21 34:18

35:4 38:14,25

51:14 61:24,25

79:24 80:7 99:15





113:25 114:8,23

120:4 151:25 152:2

153:9 162:3,5,6

269:17 273:4,10,14

274:19

P

p.m 298:19 package 196:23 packets 3:12

page 19:19 22:25

23:5,6 75:15

104:17 105:16

115:16 124:8

126:17 130:11

143:3 172:24

173:23 180:17

192:21 205:19

208:14 286:10

287:3

pages 106:24 116:11

132:16

PAH 29:3

PAHs 29:22 pairs 18:11 panel

2:4,12,14,22,23

3:1,14 4:6 5:12,18

28:25 29:2 31:7

43:7 125:3

150:18 157:24

214:1 260:15

panelists 216:21

panels 132:14

PANs 17:6 66:13

paper 36:19

52:18,20 53:20

57:3 58:18,20

59:2,25 60:6

117:13 125:18

127:1 129:8 152:24

papers 70:18

90:15 117:22 152:7

193:9 253:17

297:23

paragraph 53:13

63:15 68:6 78:18

115:20 205:19

215:4,25 292:16

294:24 295:18

paragraphs

115:21,24 116:3

179:25 206:18

216:5 217:18

parameter 109:1,3

parameters 19:11,25

131:13 135:1

274:2,6,7

paraphrasing 139:10

Pardon 112:4 114:12

parentheses 18:14

Parenthetically

204:25

park 272:19

parking 287:5 participants 295:10 participate

218:8,21

participated 41:4

particle 29:18

119:19 129:19

particle-bound

29:23

particles 28:9

29:19 58:15,20

59:13 61:14,22

128:15 129:7

156:8,11,14

161:5,9,10,16

particular 10:17

30:21,22 80:17

82:24 83:8 84:17

98:7 102:6

134:22 135:4,16

150:13,23 161:1

203:14 204:9

207:22 227:7 228:2

239:3 244:24 258:6

280:20 281:1

287:22

particularly

33:9,17 34:10

65:10 85:11

94:16 110:7

173:1 186:3 190:15

193:14 230:1 235:1

239:6 267:15

particulate 28:24

51:14 85:7

particulates 153:10

156:6

partly 245:10 pass 247:3 passed 108:4

112:2 137:15

247:15

passing 70:14 71:23

100:7

past 16:19 200:18

295:15

pasted 235:19

Pat 8:25 9:2,11 patiently 93:14 patients 201:10 pattern 37:12 patterns
35:7 81:20

273:6

paucity 147:24,25

pause 138:17

pay 82:14

Peacock 41:10

peak 65:18 66:3

70:4 119:24

259:6 267:3,22

268:5,9,12

287:12,15

peaks 287:10 pediatrics 152:25 peer 48:1

Penn 7:4

people 2:8

7:21,22 8:14,15,21

9:24 14:2,17 18:24

21:3 22:24 62:8

64:15 89:24

90:24 91:19,22

93:2 96:3,19 102:9

104:5 117:21

123:21 130:22

135:4 141:9 142:14

150:6 154:4,16





156:20 157:24

158:24

161:1,7,12,14,20

165:23 166:8

168:15,16,22

176:23 184:9,16

187:7 189:7 196:10

203:11,17 205:5

220:5 231:4 233:22

237:5,16 240:12,14

257:5,15,25

258:11,15 260:5

261:24 273:2

274:23 275:7,10

276:4,12,17

277:3,5,6,20,22,24

278:3,4 286:22

292:8,19,21

294:6,20 295:3

people...some 135:9

people's 165:1

176:10

per 104:3,4,14

105:24 106:17

107:13,14,15,16

108:11,22 119:25

133:18 136:15

137:17,18,23,25

179:21 197:9 199:7

208:18,19 282:11

283:21

percent 50:7 66:6

144:25 231:12

233:9 243:13 245:2

255:10 257:15

258:11 265:2 276:6

277:22

percentile 16:18

238:12

percentiles

227:25 230:2

perfect 61:23 75:21

77:9 100:22

perform 293:25

performance

243:7,9,17

performed 227:22

230:8

perhaps 47:23 81:14

97:6 150:15

151:17,22 154:20

183:9 207:3 279:25

280:4

period 26:5 73:23

125:10 182:8

periodically 171:21

periods 187:8

Peroxidase 157:11 peroxidases 155:7 peroxide 157:12 person 110:2

216:1 220:16

226:15 287:4

personal

19:3,5,7,10,15

20:3,6,7,8,9

27:8,22,24 28:7,10

74:21

88:3,11,12,18

91:21 224:25 225:5

personally 26:7

131:4 184:18

personnel 5:13

persons 225:17

287:19

person's 88:17

89:14 98:8

perspective 151:4

212:18,22 260:3

261:22 289:10

perspectives 102:8

pertinent 78:12

110:24 268:22

perturbed 78:9

petroleum

39:14,20,21

phase 28:24 290:7

Philadelphia 222:10

225:18 226:7 242:9

243:5

257:5,7,10,19

258:13 260:2

261:3,5 262:23

263:12 264:3

265:22 276:2,22,25

277:1,4 281:23

282:3 283:21

285:23 287:19

293:4

Philadelphia's

277:20,21

philosophers 54:8

Phoenix 266:1

phone 7:21,24

8:15 9:10,12,25

22:25 52:16 69:9

96:12 110:15,20

122:13 127:13

163:17 166:8,10,14

171:22 215:12

217:24 218:2

254:20

phones 8:16,17

photochemistry

255:13

phrases 206:17 physical 94:9 physician 119:13 pick 109:9 126:16

193:7 197:9

picked 104:23 197:9 picking 167:7,12 picture 27:4 38:2

132:16 133:5

278:15,19 283:10

289:10

pictures 50:4 pie 287:2 piece 56:25

121:24

122:6,11,16,18

178:14 240:9

pieces 122:24 144:5

158:7,23 172:21

Pierson 18:9

Piloto 119:22

Pinkerton 6:13

118:12,13 127:19

279:23

pinning 193:3

placed 30:21





places 56:17 180:21

plagiarize 11:7

plague 34:9

plan 42:10,13

169:25

planning 221:15 plans 42:9 251:15 plausibility

21:18 35:15

36:5,10,11 49:14

94:8 103:13

119:4 137:12

155:24 173:19

193:17

plausible 128:11

138:5 153:13 187:3

212:15

player 197:1 playing 64:25 please 8:11

15:15,17,22,23

16:5 17:12 19:1

21:23 23:22

24:18 26:19

27:19 28:17

29:10 30:5,17

32:25 39:15

45:12 49:5,7,8

140:17 157:20

166:14 180:13

189:21 220:24

221:19 227:18

232:2

pleased 118:3

194:20

plot 39:9 79:18

128:2 258:6,7

plotted 79:20 plowing 47:17 plus 96:20 105:3

132:25 205:18

253:20 276:4

plus/minus 143:18

144:15

pluses 131:16

132:21

PM 24:7,8,11,15

28:10 34:18

38:16,25 60:24

61:7 120:4

125:10 133:17

183:18,23,24

191:16,17,21 195:8

211:20,21

212:1,5,7

213:14,20,25

PM10 29:20

PM's 213:21,23

point 10:4 38:17

43:15,17 44:25

45:12,20 53:3 54:6

57:6 63:2 74:7

78:5 79:16,17,25

92:17 96:1,7

98:8,23 101:5

102:21 108:20

112:14 113:24

114:6 116:8

117:9,10 130:1

136:15,16,17,24

137:1 144:1 148:10

150:8 154:23

161:12 180:23

182:6 189:6,22

191:4 192:18 194:8

199:3,4 207:4

208:8,23 213:6

214:7 221:8,24

224:12 234:12

235:20 236:23

237:2,9,24

239:24 240:25

244:13,22 245:18

259:4 264:7 266:12

270:25 272:16,25

274:14,16

279:18,19,23 293:6

pointed 35:4 94:7

110:25 129:20

196:2 266:10

pointer 13:5 pointing 167:16,17 points 16:8 19:4

54:5 64:3,5

81:17 95:14

105:7 107:4,8

113:13 116:18

118:14 124:9

126:1,3,10 129:6,9

167:12 171:5 190:6

199:11 209:1

238:24 239:8

250:24

policy 2:24 31:3

41:21 62:24 63:6

121:21 173:4

185:11,19,20 186:6

199:12 211:5,8

234:15 237:6 238:5

pollutant 20:7

21:5,12

29:14,17,21

33:21 36:6 114:9

150:4 193:3,5

194:3 214:15

pollutants 21:20

24:4 27:14 28:8

29:9

34:5,8,16,17

35:8 37:13 51:12

105:21

113:10,12,20

114:1,6 115:2,4

151:3,7,22

152:11 153:20

190:25 195:25

214:24

pollution 38:25

57:6,10 90:22

114:4 195:5,13

276:9 277:4

pollutions...the

277:4

poor 27:22

162:11,17,18

poorly 27:9

population 13:24

16:3 74:16,20

75:16,23

88:3,8,13,15

89:13,14,15

90:3,12,13

91:3,14,16 92:19





104:10,23,24

109:21 134:24

152:16 158:23

159:20 161:15

162:21 197:11

199:15 201:5 205:4

208:24 212:17

225:22 231:22

232:10,11 238:21

241:15 242:7 265:2

276:1,3,6

population...the

275:25

population-based

90:6

populations 75:19

119:9 138:19

148:13,19 150:21

153:17,21 154:9,21

158:7 159:21,25

160:6 162:12

164:20 168:25

205:7,9 231:14

242:2 274:20

port 275:8

284:13,15

portion 35:12

198:17 226:2

239:15

portions 69:5 portray 288:20,21 posed 80:18 position 198:6 positive
17:5,8,9

22:12,17

23:12,14

35:10,20 79:5

106:18,21

107:1,6 140:18

141:4,6

206:18,19 212:23

possibilities 136:7

possibility 51:25

244:12,15

possible 18:11

43:19 47:19

71:10 80:22 98:6

135:22 137:4

151:21 160:6

167:25 168:9

200:4,16 201:19

207:12 219:17

295:13 298:9

possibly 81:12

144:14 208:3

270:19

post 160:16

posted 217:4

Postlethwait 6:17

153:25 154:5

156:24 157:3,7

180:12,15 182:9,15

195:18 279:1

postpone 60:11 potent 201:3 potential 36:13

37:8 61:8 94:19

154:8 158:5,20

160:19 176:2

192:22 200:24

205:3 222:13

223:16 224:3,5,7

230:10 231:9,23

232:5,11

potentially 51:8

52:7 85:5 150:25

152:18 153:4,12

156:12 222:14

power 7:13 126:25

141:21

powerful 164:7

283:3

ppb 16:24 17:1

200:6 227:10

229:17 236:6,15,19

261:9

ppm 17:1

108:10,13,18,21

109:1 117:14,24

128:11 131:25

132:1 137:19

138:25 143:8,11

228:20 229:1,4,6

287:4,6

practice 47:25

90:11 91:17 92:4

precedent 191:14 precision 144:2 predicted 282:19 predicting 279:20
prediction 210:8 predictions 255:2 predominantly 55:23 pre-existing

149:7 150:6 151:23

152:5 163:5

preference 101:1 preliminary 203:18 premature 152:16,23 prematurity
152:16 pre-meeting

296:9,13,20

pre-natal 25:11 preparations 3:16 prepared 280:8 preparing 125:10
presence 25:4,13,16

28:5 157:11

present 15:9 39:7

41:17 61:22

81:15 93:21 95:2

123:2 198:3 201:17

220:6 270:13

presentation

4:15,20 9:21 25:24

26:12 42:20

49:19 109:16 213:6

220:1 233:14,24

242:16

presentations 33:22

196:21

presented 32:1 79:8

92:2 100:5

110:6,11 124:15

231:19 236:20

288:17

presenting 5:3

26:17 79:8

146:22 185:24

233:20 290:20





pressed 46:10 60:22

presumably 85:10

173:2

pretty 10:25 58:4

67:22 70:20

118:2 161:13

172:12 177:23

181:4 188:4 190:12

243:10,11 246:8

270:4

prevalence 115:25

190:9,16

prevalent 36:9 prevention 114:3 previous 36:24 37:1

68:12 76:13,14

78:2 93:17 94:2

124:10 167:11,13

236:2

primarily 63:25

106:25 201:1

245:20

primary 2:12

41:18 51:14

principal 128:19

129:2

principally 26:20 print 79:18,23 prior 35:9 36:15

172:22

prioritizing 263:23 privately 270:6 prob...is 215:11 probably 11:24

48:6,11 49:12,15

66:19,22 67:1

73:16,18 77:24

101:9,16 104:17,22

105:2 115:6

122:8 128:7

153:3 158:13 159:3

165:12 174:4

177:15,16 192:25

230:5 242:3

255:3,18 256:19

259:2 263:14 265:1

267:6 269:25

272:11 274:13

276:10 277:20

279:19 293:19

294:22

probably...Ted

280:12

problem 8:5

46:22,24,25

48:19 58:2

79:23,24 80:18

85:1 87:4 91:1,4,5

92:11 111:24 114:7

121:9 126:23 151:5

153:8 167:20

186:22 187:7

189:6,13

194:9,18 206:14

239:5 249:6

250:6 261:23

269:23 276:10

278:5 279:11 281:1

problematic 51:8

113:22 147:8 195:9

problems 8:12 56:23

78:23 94:16 111:24

182:20 186:23

214:25 216:25

280:3

procedure 265:8 procedures 48:23 proceed 212:21

220:22 243:18

244:2 260:23

proceeded 222:20

223:7

process 4:25

12:11 14:7

41:2,12,17,18

42:5,9,11,15

43:9,12 45:16

46:7,14 63:2

101:18 114:5 115:4

120:16 186:10

210:13 216:24

221:17 222:1

247:22

processes 42:4 63:9

produce 135:13

144:25 156:14

169:6

produced 114:20

155:9

producing 15:7 product 5:22 products 214:23 profile 232:20 profiles
226:13,14 progress 22:14 23:7

42:22 80:7,15 87:5

232:24 290:14

progressing 23:19

pro-inflammatory

154:10

proof 128:18

129:2 182:16

proper 27:10

40:12 48:9 81:10

properly 216:3

238:20

proportion 238:21

241:15,24

proportional 248:9 proportions 248:16 proposal 137:8 propose 136:13

138:14

proposed 54:7 137:6

140:8 221:18

pros 78:8 prospective 113:4 protect 214:5 protecting 200:21 protection
2:1

289:11,20 290:4,10

protein 155:14 proteins 157:13,17 protocols 48:23

144:24

prove 101:10,15

138:11 258:13

provide 3:8

21:18,21 26:22

28:18 39:15 43:1





47:20 50:12

93:22 113:18

115:24 179:4

197:25 211:2

235:18 237:14

provided 3:4,7,11

16:13 17:22

18:19 29:17 33:6

39:25 40:1 58:18

64:3 70:11 97:16

167:23 235:13

241:21 253:15

provides 42:6

48:5 186:25 227:21

providing 4:7 177:4

196:13 207:21

235:23 252:5

261:21

provisions 2:25 proximate 78:16 proximity 75:10

154:14 205:5

264:6,21

PSD 270:15

Pub 58:22

public 2:14

3:2,5,7,11,15,20

4:23 5:2,3,7

6:22 26:2,4,7

32:20 43:23

44:1,10 46:21 88:5

117:1 125:4

131:1 149:2

158:5,11,20

159:8,12,24 195:24

196:16 200:21

203:11 205:4

212:18,22,24 214:5

217:7,8

233:17,19

289:11,20

publication 27:20

28:21 35:2 51:9,25

57:17 58:1 146:13

publications

33:20 113:3 196:20

published 24:6

29:25 59:3 82:16

145:22 152:25

pull 15:16 52:18

59:24 63:14

140:2 146:11

172:20 185:8 266:9

pulled 56:13

123:6 140:1 254:14

pulling 208:22

296:7

pulmonary

213:15,16,18,22,25

punching 215:19 pure 61:20,25 purely 271:6

purpose 86:24 169:3

172:16,25 174:8

221:2,20 224:24

225:12,16 289:7

purposes 43:15

237:6 238:6

pursue 147:19

push 82:25 103:24

262:25

pushed 49:13 170:24 pushing 219:22 putting 123:13

138:15 161:4 186:4

219:15 272:3

Q

Qingyu 7:1 15:20

74:14

qualitative 71:7

73:18 74:1

134:14

198:15,16,24 202:4

211:7 224:10

qualitatively 74:17

131:15 134:8

200:17 224:1

quality 4:9 17:1

43:19 47:3 87:16

111:16 112:18,25

113:7,8,19 173:8

175:12 221:6

222:21 223:8

224:22 226:20,23

227:3,13,23 230:12

238:8 239:12,15

240:16,19,22

242:19,21

243:2,3 247:5,16

248:14 251:25

260:2 261:8 271:10

272:5 274:3,24

281:9 283:2 286:11

288:4 290:17

294:11 297:2

quality/modeling

267:5

quantification

121:14

quantified 30:20

quantify 21:9

134:11 198:19

200:9,11,14,15

211:4,18 212:3

235:3

quantifying 198:17

quantitative 39:8

68:4,7 84:12

120:17 122:3

123:2,8 131:22

132:10,22 144:10

145:13,17 187:19

198:20 202:3

251:16

quantitatively

27:15 168:4

quantities 86:14 ques...my 295:11 question 10:8,16

15:22 16:1 20:16

30:25 35:23 46:1

47:5,9 48:3

49:9,20 52:2 55:23

57:22 62:18

63:4,12,20,24

71:14,19 73:14

84:2,23 90:1 98:12

100:3,12 102:15

104:8 105:15

106:15 107:11





120:13,15 121:23

122:1,4,5,15

131:3,5,8 133:11

136:5,8

138:19,21 139:4,19

144:9 145:13

148:12,14,18,21,24

150:16,17 157:18

158:1,3 163:15

166:21 167:14,20

173:18,19 174:15

185:24 186:15

189:18,24 203:1

204:6 210:13,19

14:14 18:6 37:23

55:1 73:23 89:19

90:9 92:16

119:22 134:19

171:6,17 175:10

187:18 202:9,10

235:21 252:12

255:15 256:21

266:13 278:22

283:5 286:18

290:19 296:16

quotation 48:9

quote 36:15 60:19

134:4 235:9

209:14 229:3,8

235:11 236:18

261:2 288:10,16

ranges 20:4,5 104:4

199:11 200:19

rapidly 45:2 259:3

rare 135:14

rat 126:7 211:10

rate 12:22 98:20

152:17 268:20

rates 238:14

rather 11:8 34:14

36:20 54:17

80:11,23 82:23

235:5,16 237:10,11	 	

99:9 132:16 136:15

241:23 258:23

260:19,24

262:3,7 268:13

276:1 283:11,21

285:7 288:18 289:5

292:4

questioning 201:10

questions 10:5,6

17:14 27:10 31:6

32:19 36:9 37:19

39:11 44:15,18

62:25 63:8

118:10 120:21

121:19,20 123:4

130:10,11,14,20

140:4 168:1 170:20

171:9 172:23

173:6,17 176:9

178:21 181:2

184:10,15,22

185:10,18,25

186:10,12 189:10

201:2 215:7 217:19

222:8 239:22

241:12 242:11

257:3 258:18

260:24 266:11

283:1,6,8 287:6

292:10 294:8,10,18

quick 97:22 210:12

quickly 38:4 183:15

261:13

quite 9:21 10:24

R

racial 162:4,14

Radiation 4:18

radon 121:6

Raiders 172:8

rail

284:16,18,20,21,23

rail...for 284:20 rail...the 284:18 railroads 64:18 raise 103:2 136:9

168:8 171:5,9

285:3

raised 154:24 247:7

283:8 285:9

raises 35:23 285:25

raising 27:9

range 18:9,14 21:25

98:23 101:25

102:3,11 104:9

106:16

107:12,16,18

109:13 119:25

120:18,19,23

121:13,25

137:10,22

138:1,4,24

173:20 175:22

177:17 197:9,10,21

198:1,4 199:1,8

202:23

208:15,18,25

150:14 164:24

191:6 196:15 204:9

208:6 214:23

234:12 248:2

264:21 268:15

ratio 243:16

rationale 34:1

126:19 153:13

ratios 269:5,9

271:18

rats 82:20 98:25

99:1

raw 87:1

REA 64:16,22,24

65:21 67:4,10

117:12 118:4,8

128:22 210:14

235:10,17,22

236:24 237:12

238:2,16

239:3,18 241:7

reaching 59:14

128:7 237:22

reacting 155:19

156:13

reactions 20:11

reactive 155:2

156:11

reactivity 152:3

155:16

reacts 157:8

readable 79:15





reader 70:6 77:17

83:20 181:5 206:20

251:9 278:20

readers 148:9

188:25 209:5

278:13,21

reader's 207:18 readership 270:2 reading 53:11 54:22

64:12,22 68:18

89:24 117:18 146:7

266:16

reads 176:22 180:18 ready 47:7 60:14 real 61:21

62:7,10 64:13

81:11 90:5

114:23 115:2

123:21 129:3 136:4

189:7,8,17

197:5,20 210:12

235:5 252:17,25

254:1 256:18,24

276:3 292:13

real...the 252:25

realistic 129:9

288:16

reality 105:4

128:15 155:12

183:24,25 195:24

212:12

realize 82:6

114:4 196:5 214:19

realized 215:3 realizing 209:3 really 5:18

14:11,13,14

20:18,21 21:9 28:4

38:22

43:3,18,19,24

45:3,18 48:17

49:13 54:25

57:4,7,8,21

58:1,5,17 59:6

65:2 66:16 69:25

71:8 74:6

75:9,11,15,22 80:5

83:10,14,24

84:16 85:3 90:20

93:1 96:2,22 97:23

103:24

104:7,12,13,15

105:5,10,25

106:2,3,5,18,22

107:2,10,24

108:1 109:4

110:3 111:16

112:11,13 113:8

114:1,7 117:17

118:15,18,19

119:9,22 120:11

122:19 127:18

128:19 129:13,25

131:20 132:10

133:24 134:11

135:16

136:3,8,16

138:18,20 139:4

143:20 144:4,6

147:9 148:14

149:11 153:15

157:25 159:1

160:11 164:25

166:25 167:13

173:6,11,21,25

174:1 176:21

177:11 179:19

180:19 181:5,23

182:20 184:8

190:19 191:24

194:11 197:25

198:2 201:4 203:16

206:1,10 212:14,20

214:11 217:14

218:1 224:22 235:6

237:5,14

239:7,16 240:18

241:8,23 252:16

253:2

262:12,17,24,25

266:3 267:4,22

268:13 269:24

270:3 271:24 274:3

278:12,15 279:14

284:10,12 285:19

288:5,19 289:19

292:24 293:2,20

294:23 295:16,22

298:1,2

really...in 116:10

really...they

130:15

realm 65:7 198:21 real-world 252:24 reason 21:14

52:14 54:4 73:21

95:8 98:5 147:23

167:1 192:21

226:16 245:6

reasonable 72:2

252:12

reasonably 55:14

110:11 132:6

190:21

reasons 13:9

96:25 97:6

125:6,11,13

129:20,23 181:18

279:16 292:19

recalculated 19:20 recall 53:10 60:24 receive 96:13

232:22 287:21

received 96:19 receiving 5:1 recent 11:5 27:20

28:18 34:18 55:1

70:11 222:3

227:6 297:22,23

recently 29:25

198:12 293:24

receptor 94:10

246:10

receptors 225:24

247:24 263:13

264:9 269:7 272:13

recess 100:1 220:2 recitation 31:2 recognize 14:3 43:2

73:8 89:9 90:22

156:17 266:7

recognizing 44:5





225:4

recommend 66:23

109:19

recommendation

77:21 79:3 186:7

246:9

recommendations

30:18 40:2 109:11

recommended 102:23 record 42:20 234:3 records 3:3

red 22:5 178:1 redid 62:7 254:15 redo 246:22

redone 143:20 reduce 214:18 reduced 42:15 reducing 214:6,17
reduction...NOx

290:15

reductions

290:6,9,15

refer 190:8 297:10

reference 47:9

54:23 71:23

references 54:24

55:11 67:25

70:12 80:9

referred 58:25 66:7

96:14

referring 72:7

181:7 195:20

246:14 247:17

refers 141:7 refined 226:22 reflect 19:16

40:5 45:11 67:11

143:4

regard 35:14 48:5

59:24 75:15

99:15 111:17

112:12,20

113:8,16,25

114:2,8,17,18

124:17 141:13,15

147:8,14 148:4

149:12 150:3 174:5

181:14 190:16

195:21 228:17

285:7,22

regarding 4:8,16,21

27:10 28:15

39:25 40:17 164:14

228:14,15,22

250:21

regardless 254:13

Regine 163:18

189:20 205:16

region 244:11

276:23

regions 157:10 regression 193:13 regulate 214:23 regulated 133:18
regulation 212:23 regulatory 26:18

44:22 115:3 233:21

reinvent 48:24 reinventing 11:9 reiterate 118:15

154:1,6

reject 102:5

relate 92:5 93:3

121:9 126:5 172:24

284:2 291:17

related 16:1 21:4

39:21 44:6 47:5

49:17 67:24

74:20 75:10 78:1

85:24 122:2 124:14

148:19 158:1

163:15 164:23

235:14

relates 105:23

127:5

relating 71:16

283:17 291:17

relation 97:2

164:15

relationship

12:25

13:1,2,7,12,23

19:9

25:2,5,7,11,14,17

50:8 95:16 96:23

97:11,15 144:12

153:1,2 190:4

204:23 225:5

297:8,20

relationships 101:8

103:11 123:3,8

173:9 192:22

193:14,24

relative 38:25 84:8

86:17 87:19

155:9,19 162:7,8

208:7

relatively 112:15

114:19 119:21

134:24 147:25

release 42:12 released 43:24,25 relegated 34:3 relevance 59:4

105:4 158:12

relevant 62:25 77:3

80:8 88:8 89:7

102:5 105:11

167:7,13,17

168:6,18 172:22

182:7 196:17 210:8

237:6,7 238:5

280:7

relied 273:24 relies 40:14 rely 100:23 relying 235:2 remained 29:14
remains 27:9

28:15 30:15

remarks 2:19 3:17

100:8 278:11

remember 60:20

117:17 125:9,18

142:12 183:20

269:14 271:14

281:5

remembers 62:3

remind 139:17

219:13

reminder 15:15





remote 17:10

removing 259:11

rename 49:5

renaming 49:6 52:14 reordered 235:25 reorganization

56:20

reorganize 19:14 repeat 273:23 repeatedly 54:7 repeating 40:20 rephrase
32:8 replace 100:9 112:3 replacement 112:6,7 replicate 56:24

62:9

reply 258:19

report 11:4,5

50:4 53:10,25

97:25 125:11 153:3

208:14

reported 27:11 30:8

35:7 55:15

147:12 253:24

254:12,16 258:24

259:11 268:11

reporting 55:15

reports 11:23

51:2 106:12

represent

30:10,13 91:3

235:10 286:12

289:2

representa... representativeness

260:1

representation

237:15 244:4

representative

168:2,3

238:4,18,19

240:8 257:13

288:21

representativeness

239:24 240:1

represented 4:18

92:10 227:25 228:4

241:16,18,24

253:14 280:17

representing 242:4

represents 15:5

39:20

reproduced 253:8 reproducing 244:1 request 131:8 requested 68:2
requesting 297:18 requests 3:6,9 required 200:9

261:14

requirement 217:6

218:2

requirements 162:22

requisite 200:21 re-reading 124:4 research 4:13

7:7,12,14 151:14

160:9,13,19

274:16,20

reservation

218:5,16

residence 276:3

residential

257:8,14

residents 276:4 resistant 141:17 resolved 43:16

Resource 32:23 33:6

resources 32:23

44:7 263:15

respect 79:6

151:6,25 228:23

257:18

respective 94:12

respiratory 21:19

22:1,2,6,8,12,23

23:8,10,15,18,19,2

1 24:24 25:9 33:18

37:4,5,8,11,14

38:9,20 112:20

115:16,18 116:1,11

149:7 150:6 187:16

188:4

190:1,10,17 201:23

202:12 203:6

208:13

respond 71:13 92:13

135:15 171:24

183:14 204:13

216:25 258:23

260:12 261:4 280:9

293:11

responders 147:17

response 12:13,22

13:23,24,25

16:12 17:6,21

18:18 48:4 60:21

95:16

96:4,6,8,23

97:5,11,15,24

98:12 99:9

108:8,25 114:18

118:24 119:10

126:13 127:22

128:1,4,19,24

129:15 130:5

131:15 141:14,19

148:21 235:2

242:12 249:7

responses 39:2

95:19,24 96:14

108:7 111:1

116:5,12 122:3

140:10 142:8 206:3

215:8

responsibility

195:3

responsible

179:22 215:3

responsive

135:10,11 143:12

147:13 152:2

responsive...I

132:2

responsiveness

95:24 97:3,7,8

104:18 112:13

117:11 132:24

142:14,18 143:13





144:8 145:9

162:6 228:23

235:14 236:17

responsivity 151:21

152:1,9

rest 43:5,6 99:23

272:9

restate 71:19 restaurant 218:6,11 restricted 86:19,20 result 126:14

243:20

results 21:15

29:15,17 30:1

33:21,24

34:11,14,15,19,23

35:10 37:12,13

39:8 55:2,7

74:22 76:22 77:5

90:20 106:12,13

138:2 149:21

170:25 171:1

221:22 227:23

230:20,23

231:17,18 232:9

244:3

246:7,11,16,20

251:18 259:17

retrieved 55:5

review 2:4,12

3:22 5:6 15:25

19:8 35:4,9

36:15,25

37:1,2,7,9 40:25

41:2,5,17,20,21,25

42:5,7,10,11,16,17

43:9 46:8,9

47:13 48:1,5

55:12,19,24 57:9

64:23 68:12

78:25 80:7,17

168:8 169:13 185:4

203:18 242:18

289:9 297:4

reviewed 34:17

117:22 260:22

reviewers 154:1

reviewing 5:21

135:19

reviews 34:18

41:4,5,8,13,15

43:14 51:3 68:3

revised 15:10 41:12

100:7 108:5 122:23

214:21 217:3

revises 108:5 revision 154:7 revisions 219:17

Rewrite 79:5

rich 8:2,8 9:12

123:23,25 127:12

129:16,18,19

163:23,25 164:10

165:4

Richmond 4:19 220:8

233:15 273:8,9

274:15 290:2

291:16,25

rightly 209:12

rigid 53:13

rigor 11:17 286:21

rigorous 33:16,19

36:2 50:24 147:19

risk 3:25 4:22

13:16 34:20

42:4,13,14 43:3,23

45:9,17 46:9

65:8 68:6,19,25

69:4,16 70:2

72:7 78:3 80:19

81:2,6

83:9,11,21 84:3,13

85:4,19 98:6

104:1,24 109:13

115:23 116:5 123:3

128:23 137:7 148:5

150:24 151:2

152:6,9,13,24

158:14 159:10

162:8 164:15 166:2

167:9,14,17

169:17,24 170:2

173:2,3,21 174:5

178:24 179:4 183:6

185:22 186:12

187:11 197:8,15

198:13,18 199:6

200:4,9,15 201:4

203:12 205:18

206:7 208:7

211:6 221:5,12

223:13 234:23

235:1,6,7

237:2,7,14 238:4

240:19 242:20

274:24 290:6,8

292:8

risk-related 167:4

risks 109:21

162:7 221:22

223:15

Riverside 280:16

road 17:19 18:24

76:3 83:7

84:7,8,12

85:9,13 136:4

138:22 252:22

254:3 256:15,21

257:9,16 260:6

261:24 264:15,22

267:22,24,25

268:1,5,7,14,16

278:24 279:3 284:7

285:1

roads 71:1,2,25

72:12,22,23

75:7,10,12,19

92:22 161:20 163:7

244:24,25 255:11

257:24 258:1,11,15

285:3,6

roadside 83:13,17

roadway 65:13 71:17

91:11 226:4,8

253:13 254:15

256:3 259:8 266:23

267:1,4,14

268:17,21 272:13

297:9

roadways 154:14

205:5 245:8,18

264:6 269:6,7





277:9 285:17

Robert 60:13

robust 29:14 30:4

120:3

Rockfish 219:3

Rogene 2:21 4:3

5:10,15 9:3 46:2

52:9 87:10

100:11 135:25

219:8 296:2 297:1

role 2:10 30:23

64:25 186:10

roll 246:19

247:20 248:6 252:2

273:20

rolled 248:25 rolling 94:24 rollup 288:9 roll-up 239:9

Ron 7:13 87:9,24

131:10 133:11

134:18 178:17

180:9 183:19

286:23 287:25

Ron's 182:25

room 8:13 26:8 33:3

298:14

Rosenbaum 220:13

244:21 249:25

202:6,15,18

206:8,22 207:5,7

208:8 209:20,24

210:2,17,21,23

219:8,10,13

rough 82:16 253:18

roughly 81:20

82:5 108:20

round 44:9 67:19

68:3

route 192:6 214:11

routed 17:11

RSV 152:7,10

RTP 4:15

rubber 136:3 138:22

rulemaking

45:4,13

221:17,18 222:1

run 131:22 231:2

232:3 248:3 257:21

278:2 280:19

running 132:14

170:25 257:22

Russell 7:8

63:20,21 214:10

262:6,12,15,17,20

263:17 265:13,19

268:2,18,25 269:12

295:8

satisfied 86:6

288:15

save 40:19 262:23

saved 42:1

saw 48:3 83:8 135:8

162:9,11,15

187:4 207:25

282:10

saying...so 257:17

scale 17:18,20

18:3,7,20,22

204:23 246:23

255:6 280:14,17

291:19

scaled 250:1,2

scales 69:24 256:10

SCAMP 78:19,20,23

79:1,4

scant 123:10

SCAQMD 79:2 scarce 69:2 scenario 165:20

288:13

scenarios 227:13,21

schedule 41:23

43:2,17 44:1,4

45:6,8,11

46:3,10,11,13

215:15

250:6 263:22 264:2	 	

216:11,12,14

269:3,13 271:15

272:15 276:24

280:25 282:7,14

284:6,18 285:15

Ross 4:13 38:11

46:2,5,6 61:3

71:19 72:6,15,18

73:8 75:4 76:4

139:17,22 162:25

169:18,20

175:3,9

177:20,22

178:1,4,7,10 181:6

185:3,7 187:13

188:11,16,18 198:7

199:17,20,23

200:7,24 201:13

S

safer 203:12

Samet 6:21 47:6,8

48:16,21 52:24

53:19 97:1,22

101:5,17 111:9

120:9,10 122:19

125:9 143:16

145:11 172:3,4

181:21 192:3,20

194:7,13 196:2

278:10 283:8

Samet's 150:1

194:25

sampling 154:19

Sarnett 27:21

sat 187:14

217:17 219:21

scheduled

90:17,20,23,24

217:5 221:14

scheme 13:5,14,17

71:2 101:3,13

Schlesinger 8:2

9:3,12 123:23,24

124:1,3,21

129:17,19 163:24

164:2,5,8,12

165:7,10,14,17,19,

24 166:1

School 6:8,10,20,22

science 2:2,17 3:23

12:6 40:6 41:21

42:3,8,17 169:22





185:11,20 186:19

188:8,14 235:17

237:1

sciences 7:16

16:4 152:18

scientific

2:3,11,15,23

3:21 4:5,25

12:19 15:5 41:14

42:20 57:23 60:7,8

65:7 122:21 146:23

156:1 167:3 175:18

180:23 206:7

scientifically

33:19

scientist 26:16

scope 221:2

Scott 4:19 220:10

225:10

screen 238:10

screening 255:25

258:8

scrubbed 128:10,12

se 133:18 179:21

283:21

seafood 219:2,3 seamless 178:25 search 10:9,13

54:17,21 55:1,13

season 86:19,20,21

seasonal 18:21

seat 26:8 39:16

second 3:23 4:16,21

13:21 16:9 17:13

33:8,15 34:13

39:23 40:8 42:23

44:25 45:24

47:11 49:21

69:14 78:1 95:17

113:24 120:22

123:2 134:2

143:3 151:10

169:22 221:11

223:10 225:9,16

234:17 237:9

240:25 243:4,11,12

247:21 254:24

255:22 265:20,23

288:18 290:7,24

second...there's

115:15

secondary 40:24

116:6

secondly 27:7 30:22

180:5

section 20:21

37:3 54:23 57:9

60:21 67:21

68:20 69:1 70:1

76:23

77:16,17,21 78:9

87:15 110:8

115:13,15 116:21

117:2 124:7 247:21

271:3

sections 20:21 51:2

68:14 80:4 94:12

116:24

see...get 267:12 see...it 108:12 seeing 49:19

64:24 85:10

119:5 132:5

187:2 245:11

seek 181:22

seem 58:2 73:5

115:3 124:24

126:18,21 170:8

172:11

seemed 13:12

55:11 72:2

149:13 201:8 256:3

257:19

seemingly 71:10

seems 29:12 59:9

61:17 69:20 71:5

75:24 76:13 80:8

94:15 116:21 118:6

120:14 124:22

130:6 159:11

164:19,23 169:14

175:18 191:12,13

203:23 209:9 249:6

256:4,13,18

257:1,12 277:10

282:15 283:4,18,25

289:25 292:5

294:11

seems...it 256:13

seen 10:3 53:9 56:7

61:11 74:15,19

75:13 82:8,14 98:3

99:7 105:23 119:13

139:25 140:6 147:9

161:25 175:21,25

179:23 228:20

229:1 254:4

256:9 268:10,19,23

seen...most 256:8

Seigneur 7:6

76:12 242:22,24

245:21 246:1

247:19 249:1,4

250:13,18 295:2,5

select 226:21

selected 31:2 33:20

34:22 55:3

225:18 227:7 235:4

237:11 241:24

261:12 264:5,9

selected...it 264:2

selection

10:10,14 35:2

122:21 234:21

237:21 279:25

280:2

selective 237:20

238:10

semantic 180:16

semi-disagree

117:20

semi-quantitative

251:17

semi-quantitatively

198:25

send 295:3 296:6,14

297:24

sending 8:10,11

215:25 294:22

senior 26:15 147:4

sense 13:10 48:9





49:23 61:13,23

71:8 79:13 85:24

99:10,12 156:22

163:4 196:13,17

198:15,16,24

213:14 238:4 247:1

250:19 257:10

259:10 267:20

288:8,10 291:3

297:6

sensitive 98:8

104:23,24 109:1

274:8,20

sensitive... excuse 256:16 sensitivity 97:13

98:9,16 125:23

293:24

sensitized 109:5

sent 70:25 71:22

124:5 216:21 234:5

sentence

173:22,24 174:21

180:17,18

sentences 149:24

separate 11:19

89:22 119:18

120:12 131:18,21

132:9 191:15,19

196:21 210:14

212:11,20 213:20

227:5

separated 89:5

191:18 227:5

September 221:14,15

sequential 229:17

230:17

series 29:7 35:16

51:12 52:5 57:4

92:19 132:14

serious 35:23 serve 224:24 serves 225:21 service 62:21 serving 29:8

SES 163:6

SESSION 298:19

set-out 46:14

48:23,25 50:23

sets 3:11 36:25

41:19 51:12 76:5

173:1

setting 101:18

113:9,19 120:23

156:2 191:14

settling 197:6

several 26:25 29:25

40:2 41:4 49:1

106:10 115:21

134:2 143:20

178:21 193:6

196:10 242:25

269:20 278:14

284:21

severity 23:18

152:6

shape 13:24 shaping 145:24 share

210:21,23,24

268:24

sharp 158:17 sharply 121:18 shed 204:6

Sheppard 6:15 55:22

84:25 92:15,18

146:20 160:4 186:2

212:9 214:16

239:23 240:5,21

279:4 290:23

291:1,21 292:2

She's 219:22 shipping 284:11 shocked 75:1 118:4 short 24:25

25:2,5 27:5

36:17 42:24

50:17 63:15 65:4

71:6 74:7

103:2,5,16

109:19 111:14,22

115:20 132:24

136:6 173:9,16

179:9 187:8

202:13,16 205:2

shorter 42:17,19

69:24 70:4

shorthand 181:10

shortly 152:8

198:22

short-term 69:21 shout 176:24 showed 28:7 32:1

38:14,15,16 62:4

88:5 113:5

152:25 162:4

164:21 243:8

showing 22:12,21

119:23 143:12,22

144:1 153:2 208:17

232:23 238:14

shown 16:15 19:18

30:6 103:12 152:14

154:7 162:2 200:19

shows 5:19 16:21

17:21 18:2,8,12,14

22:9,20 23:14,17

27:23 28:25 29:2

75:11 77:6 98:16

108:6 137:4

141:9 235:18 243:7

267:2

shut 207:4

sides 248:21

SIDS 153:2 sight 142:24 sigmoid 128:4 sign 96:20,22 significance

141:7 146:23 149:3

158:6,21 159:8

196:16 236:12

significant 10:25

22:11,18 23:13

24:16 39:7 40:16

51:18 96:5,20,21

108:8 119:23

125:25 138:6

141:22 142:9,13,17

143:6 144:1,6

162:8 193:21 205:7





211:20 212:5 236:9

243:23 244:9

261:10 266:8

significantly

117:15

similar 35:7

37:7,13 130:11

160:9 231:18 266:2

similarity 11:6,13

Similarly 42:11

91:2

simple 98:20

225:6 255:18

266:18 270:4

simple...simplified

255:12

simplifications

49:11

simplified 227:14

simplistic 12:13,16

195:23

simply 43:16

45:16,19 79:3

118:15 156:9 244:4

246:22

simulate 225:16

239:10 289:4

simulated 226:15

230:7,24 232:21

233:7 272:6 275:15

287:19

simulating 289:8

291:6

simulation 227:24

simulations

227:22 271:19,20

289:16

single 33:21

114:1,6 115:4

214:24 222:10

247:25

single-family

257:14

single-pollutant

34:14,18,21,22

site 18:11,13,15,16

88:20,21 89:3

90:1,3 91:2 92:9

234:20 280:2

sites 17:10,11

18:12,15 90:4

siting 85:1,21

sitting 100:14

267:14

situation 156:17

157:5 290:4,16

situations 100:20

101:14 189:8

288:25

six 53:25

108:12,17,18 233:1

294:18

sixty 46:7 131:2

137:19

size 17:25 24:8

98:18 159:20,25

247:23,25 263:11

291:14

sizes 134:24 skills 170:1 skirting 122:7 skirts 121:16 sleeping 287:4
slide 10:3 11:22

12:2,10 14:8 21:23

23:22 24:7,18

26:19 27:19

28:17 29:10

30:5,17 32:1

220:24

221:7,9,19 223:2,3

224:14,18 227:1,18

230:3,18 231:7

232:8,15,18

233:4,6,11

234:22 235:18,24

236:2 237:8

247:2,15 249:15,16

250:8

slide...and 234:8

slides 26:12

100:4,6 220:20

221:1 247:13

slight 243:21,22

slightly 229:24

slippery 191:13 slope 191:13 199:6 small 17:20 18:22

135:15 141:20

143:24 147:7

206:17 246:6

257:1,11

smaller 86:19

255:11 285:3

smattering 54:25

Smith 233:19

234:4 240:3,7,24

241:17,20 242:16

Smith's 100:8

smoking 11:4,23

13:3,8 50:9 121:6

smooth 240:11

So...and 261:23

295:22

So...and...and

252:20

so...I 123:7 so...oh 275:14 so...so 283:17

SO2 29:22 135:8

141:15 142:4

so-called 49:4

147:16

socio 162:13

socioeconomic

161:22,25 162:15

solely 117:13

solid 16:15 207:10 solve 46:25 194:9 some...put 270:21 some...some 285:8
some...there 273:13 somebody 87:25

108:22 111:24

114:14 128:8

155:13 157:20

163:22 271:21,23

276:2

somehow 178:23

179:1,14 180:2,8





191:19 196:1,18

208:3 270:5

274:8 292:15

someone 64:11 90:18

117:19 119:11

127:1 170:13 184:4

213:1 267:14

someone's 91:10

something...the

258:5

somewhat 43:25

65:20 66:22

68:24 96:7

192:20 267:3

269:10

somewhere 64:9

94:11 251:1

255:6 267:24 268:9

277:16 278:19

sophisticated 11:25

98:12 280:19

sorry 7:23 14:25

23:6 26:3 31:25

32:12,17

98:15,25 115:13,14

122:10 163:25

171:15 175:9 178:3

182:4 192:4 195:17

222:25 226:1

241:25 247:3,14,15

271:14 286:24

297:16

sorry...a 275:7

sorry...and...and

245:23

sorry...distinct

229:19

sorry...the 225:22

sorry...there

230:14

sort 14:4 32:10

47:20 49:3,23

50:1,4,9,19

65:11,15,16

68:11,14,21

69:17 70:16 71:2,6

72:3 88:16 89:2

91:11 92:6

93:19,21 94:6

100:23 101:13,18

110:8 112:19,24

118:4 120:12

121:16,24

122:1,7 127:25

130:10 131:12,14

132:3,4,14,15

133:3,19,24

134:6

144:7,10,14 147:21

150:5 151:16

152:22 153:11,21

158:18,24

169:11,15

170:12,17

173:19,22,24

174:14,15 175:23

176:12 177:3,5

179:14 180:5

193:14 196:15,18

199:6 203:8 208:25

241:14 252:24

256:2,9 257:11

266:17 267:10

278:12,16 283:3

286:13,14,16

289:21 292:7 297:7

sorts 38:21 73:24

sound 9:24 sounded 124:13 sounding 55:5

sounds 52:25 162:20

185:15 293:8

source 56:11

77:11 157:19

160:25 195:12

196:9,14 205:21

226:4 244:13

248:23 250:9 251:5

264:7 268:10

283:13 289:21

source...if 249:20

sources 17:11 20:10

28:6 55:6 57:13

64:9,11 68:5 78:16

137:11 155:10

226:3 230:22

231:2,3,6,11

232:4,13,14

233:3,4,10,11

244:23

245:14,19,24

246:21,25 248:11

249:2,8,10,11,12,2

3 250:1,14,23

251:10 253:11

277:13 283:15,19

285:16,18

286:10,17

Southern 6:10 space 255:1,18 spaces 75:23 spare 10:2 sparse 69:18
sparsity 69:17 spatial

17:13,15,17,19

18:3,6,20 69:22,23

74:21 225:2,13

240:12,13

244:1,4,18

264:11 271:2 272:6

spatially 255:15

284:23

speak 8:14 9:9

15:17 31:23 33:2

44:14 54:18 140:17

169:20 175:8

177:4,20 180:13

182:19 183:16

188:10 192:25

193:10 198:8,22

290:23

speaker 7:20

8:5,16,17 10:1

15:18 32:20

54:18 114:11

140:17,19,21

144:19

163:18,20,21

164:9,10

166:5,11 175:7

176:15,18 180:13





182:2,5,6,19

183:16 188:22,23

192:1 197:12

205:8,10,11 218:23

247:12 263:1

298:13

speakers 93:17

speaking 52:19

148:1 184:18

198:20 200:12

216:1 220:12

287:13

species 4:11 24:8

28:24 67:22,24

80:15 82:20 155:2

specific 27:23 28:1

30:24 32:14

37:14,25 38:5

57:10 63:9 104:2

115:12 116:13

134:11,13

141:16,17

142:3,4 143:16

164:20 169:12

179:25 185:8 193:3

199:10 228:4,5

246:10 297:19

specifically 29:4

38:12 40:14

49:18 69:6

121:17 122:8 125:3

specification

123:12

specificity 11:17

103:11

spectacularly

280:22

Speizer 37:21,22

53:5,7,9 58:24

74:25 75:8

131:10,11 132:19

176:10,11,16,20,25

177:21,25

178:3,6,9,11

191:24 194:11,14

211:13,17,23 212:2

217:21

241:13,19,22

242:3,6

Speizer's 91:12

spend 80:2

165:12,13 267:6

287:19

spending 47:17

287:5

spent 118:7

223:12,14 226:11

287:11

spikes 154:16,17

spoken 15:19

spoon 116:25 117:1

spot 63:9 87:9

90:5,14

91:3,8,18 92:9

265:8

spots 245:7,10,15

264:9,16 284:9

285:10,14

sputum 157:16

St 52:3

staff 2:2,17 5:23

6:24 15:5,7

33:13 36:19

37:24 102:23 148:2

165:5 187:14 198:3

stage 63:7 120:16 stand 15:12 142:1 stand...or 291:8 standard 17:1

67:1 103:3,4,15,21

108:11 109:20

112:18,25 113:9,19

120:24 136:7

137:18

139:6,9,20

140:10 156:2

175:13,16,17 176:2

199:16,19,20

200:5,21 222:5

224:11 227:14,17

232:13 233:8

288:7,14,15

289:4,8,12,16,18

290:5,9,11,18,22

291:9

standardized 86:14

standards 4:8,9

35:17 191:19

214:14 222:14

224:3,5,8 290:8,10

standing 9:18 standpoint 132:3 stands 234:12 star-6-pound-6

215:17

start 2:7 6:5

22:1 42:7

64:15,18,24 66:2

67:10 96:3

100:10 103:20

104:1 106:5

107:9 108:9 120:11

137:15 139:3

140:14 146:6

184:11 214:14

220:4,16 238:16

242:21,24 298:10

started 54:22 65:2

starting 9:15 23:18

138:1

starts 174:1

start-to-finish

278:19

state 7:4 37:1 40:5

66:10 120:2 154:10

stated 56:14 116:10

119:16 146:21

239:1

statement 2:10

44:13 70:13 139:16

174:8 176:8

189:4 190:8,24

283:11

statements 68:13

235:8

state-of-the-art

43:13

states 16:15

17:24 18:21

73:10 75:6 76:9

168:2 169:1 200:19





224:24 225:21

static 259:22 stating 175:14 stationary 226:3

283:15

statistical 16:22

61:10 141:7,21

146:23

statistically

22:11,18 23:13

39:7 51:18

96:4,19,21

141:22 142:9,13,17

143:6,25 144:5

236:9

statistics...that

258:6

status 161:25 177:4

stay 196:25

staying 172:2 steady-state 78:13 steep 256:7,14

Steichen

39:13,17,19

44:20 45:5,15

46:16,19

step 12:14 14:7

82:2 103:20

169:8 197:7 219:14

251:9

Stephen 4:19

220:11,12 221:4

224:15

Stephen's 220:20

steps 219:11 222:21

223:8

Steve 7:15 35:6

150:10 153:24

159:7,21 160:5

221:3

Steve's 154:8 stick 217:17 sticky 155:25 stimulus 228:25 stop 115:6
143:14

172:9 185:4

stopping 200:5

story 123:14 239:18

stoves 230:22

straight 48:8 136:4

235:9

Strand 105:11 106:9

strategy 10:9,13

105:14

stratified 28:5 street 297:4,8,19 strength 34:19

103:10 107:8,19

108:2 137:12

189:7,12,15

strengthened 111:10

strengths 168:16,19

171:8 206:4

stress 155:17 stretch 149:13,14 strike 266:17 striking 119:10 strong
28:22,25

29:2,6 37:6

58:5,10 103:9

107:25 113:14

175:2 188:1,4

189:14 190:12,21

stronger 28:7 45:20

211:21 212:6

strongly 10:24

170:8 195:11

202:10

struck 131:11 structural 93:16 structure 101:8

180:17 234:13

structured 10:5

100:24

struggle

278:20,21,22

struggled 278:15

struggling 71:20

254:7

strung...strung

278:18

stuck 114:5 115:3

studied 195:5

studies 10:11

19:17,25

21:2,16,17,18

22:3,4,5,6,7,10,17

,21 23:8,9,12

24:14 27:16,17

28:3,7,18

29:8,20,23

30:11,14,24

31:13,21 32:2,5

33:23,25 34:1,5,12

35:4,9,16 36:24

37:15

38:13,18,22

40:15

51:11,12,17 52:5

56:22 57:5,24

59:20 61:4,25 62:3

71:8 72:5,16,21

73:9,11,15,22

75:5,25

76:5,6,22

77:4,24 80:20

81:8,15,18,22,23,2

5

82:1,4,11,16,22

85:6

87:13,14,17,18,23

88:8 90:23

92:19,21 95:15

97:18 99:3,18

102:16,17 104:16

105:10,18,19

106:6,11,23

107:3,6,21,23

111:1 113:2,6,7,14

114:16 116:1,6

118:19,20,23

119:1,6,17,21

120:2 123:15

124:15,16,22,25

127:20 129:24

131:17 132:11

133:24 134:3,15

135:17,19,23

137:2,22 138:8

139:9 140:4,5,8

141:4,5,9,20,25





142:2,5,11,15,16,2

1 143:24 144:23,24

145:4,6,7 147:7

149:20 151:3,5

152:14 153:8,10,11

158:11 162:1 168:7

175:10,11,15,19,20

,23,24

177:10,17,19

178:10

179:6,7,8,21

181:8,11,21

183:5 186:20

187:2,5,25 188:1

199:2,10,24 200:20

205:17 206:3,12,15

207:10,21

208:4,5,9,15,17,22

209:3,19 210:16

211:14,24 214:2

224:9 235:12

236:2,3,7

253:4,6,7

259:4,5 268:3,4,7

studying 135:5

152:21

stuff 158:23 style 20:20 sub 13:21

subcategory 283:23

subcommittee

2:15,22

subgroup 152:19

153:4 162:17

subject 27:23

28:1 109:5 147:18

subjects 28:5

109:13 143:12

145:8

submit 229:21 submitted 129:8 subpopulations

149:5 150:3 151:11

168:4,25

sub-routines 274:7

subsequent 40:2

114:9 222:11

subsequently 53:12

substantial 35:12

42:22 48:4 102:25

substantially 63:23 substituted 95:21 substrate 157:12 suburbs 276:17,21
suddenly 116:12

128:13

suffer 40:15

sufficient 12:24,25

13:1,6,11

25:1,6,10 26:22

30:19 49:25 151:18

190:3,4 235:13

sufficiently

25:18 27:13

sug...yeah 297:17

suggest 67:9 100:25

136:2 151:9 195:11

212:11 241:7

suggested 19:8,13

63:1 67:25 70:23

140:9 184:20 297:3

suggesting 72:15

82:22 103:4

268:7 287:11

suggestion 69:25

73:6 81:11 82:18

139:8 151:13

182:15

suggestions 53:14

suggestive

13:1,2,11

25:6,10 112:23

190:3

suggests 54:25

102:25 133:9 187:1

236:11 270:17

282:11

suite 34:15

sum 215:1

summaries 68:12

111:14,16 297:23

summarization 124:9

291:2

summarize 57:13

70:1 110:23 180:8

summarized 55:8

72:21 105:15 133:2

179:1

summarizes 106:8

summarizing 58:4

106:8 167:12

291:10 294:17

summary 20:20 21:24

58:23 68:13

70:23 73:1 94:23

110:7 111:7 115:17

116:4 124:7 130:13

144:10 145:13

159:2 166:25 167:5

172:19 176:20,21

177:2 178:13

179:10 227:19

229:16 230:8,22

summer 17:10

66:11,20 77:8

supplemental

84:16 298:4

support 2:18 5:13

15:6 21:14 26:25

37:7 103:15 115:25

136:21 147:5

169:16,24 174:3

197:16,25 198:6

206:11 241:7

267:12

supported 68:22

199:15

supporting 168:10

236:8 274:16

supportive 30:22

41:8 134:12,13

202:11

supports 103:8

supposed 116:22

179:14 187:10

194:9 240:9

supposition 213:21

sure 6:2,3 47:8

50:10 52:16,19

53:6 61:1 62:2

74:6 75:3 78:22





80:24 99:19

102:7 109:8 122:13

150:4 155:3

172:15,24 173:5

176:8 179:16

184:24 191:17

205:14 207:11

211:8 229:11

239:17 240:17

246:1 253:22 256:1

258:20 261:6 262:7

263:3 264:10 267:9

270:16 275:17

286:1,18 289:21

290:25 291:25

296:24

surface 155:20

156:13 161:5

Surgeon 11:22

51:1 53:25

surprise...it 260:3

surprised 61:19

75:2 117:18

118:8 210:15 256:8

260:3

surprising 35:20 surprisingly 51:21 surrogacy 34:9 surrogate 16:2

27:14 28:20

29:9,12 36:9

74:16,20 224:25

surrounding 255:17 surveys 274:17 susceptibility

13:25 149:15

151:19 153:17

159:9 162:5 163:4

susceptible 119:9

138:19 148:13,19

149:4,9,17

150:3,19,25

151:8,12

153:5,21

154:9,21 158:7

159:14 160:6,7

161:12,21 162:21

164:17,22,24

165:3,9,16

197:11 205:8

susceptible/ vulnerable 231:13 suspect 37:17

Swiss 66:9 switchboard 296:4 symbols 30:10 sympathize 69:16 symptom
116:1

190:18

symptoms

22:2,6,8,13

23:19 89:11 116:11

190:10

synonymous 146:23 synopsis 65:5 synthase 157:4 synthesis 31:3 51:3

145:17

system 82:13 99:2

298:3

systematic 35:5,11 systemic 43:14 systems 54:8

T

table 3:12 6:1

17:21 18:2,5 37:23

57:11 68:8 82:18

89:5,22 91:6 100:5

105:16 106:7 109:7

124:8 138:10

143:17 151:11

154:10 177:7,22

178:5 199:9 229:13

235:18,20,25

236:4,8,14 243:7,9

281:15 282:4

286:8,9 287:7

tables 34:3

146:11,12

177:13,14 183:1

292:19

tabular 81:16

tails 291:15,23

Taiwan 24:11

take...have... take 273:1

taking 140:2 215:23

226:9

talk 10:5,6 12:22

14:19,24 15:1,2

20:16 40:22 41:1

45:17 50:12

52:15 61:24

84:25 94:6,17

136:4,6 194:22

220:19,25 221:5

222:6 270:5 286:5

talked 10:18

23:25 94:2 130:9

158:4 172:25

196:10 197:4 235:4

283:13

talking 8:20

15:25 24:24 43:4,5

50:5 82:24 88:2

91:12 105:5

106:3 109:2 124:13

129:18 138:20

139:18 143:23

148:16 159:25

189:7,23 194:19

219:11 223:15

234:19 250:7

264:23,24 265:1

271:2 274:11 281:9

285:22 286:10

289:13 291:14

talks 10:14 57:11

58:20 87:15 115:15

tallies 291:11 target 56:1 task 65:1

team 4:17 14:17

167:16,17

teams 50:22

tease 126:11 196:4 tech 7:8 269:24 technical 168:10

229:22,24 241:6

254:17 256:19





259:12

technically 116:4

technique

76:20,25 77:19

105:13

technology 107:24

Ted 7:8 39:13,19

63:20 67:14,16

68:2 70:23 76:17

77:1 94:18 214:8

262:4 265:13,18

269:19 295:7

297:24

Ted's 69:25

teleconference

32:21 217:5 219:23

telephone 2:8 6:3

7:23 154:4

telephone...a 217:5

temperature 24:13

273:11

temporal 17:13,15

103:11 225:2

temporality 54:10

ten 104:22 108:19

132:16 136:20

282:5,7 293:2

tend 77:6 128:23 tended 208:12 tendency 128:3

134:25 135:2,5

292:20

tends 135:3

tenor 205:25 207:19

210:22

tentatively 221:14

tenth 16:17

term 24:25

25:3,5,8,15 27:5

36:18 42:24 70:4

71:6 103:2,5,16

109:20 110:9

111:14,15,22

112:21 115:18

132:24 136:6 179:9

190:17 202:13,16

203:5,13

204:2,10,19,20,21

205:3 209:15

termed 227:5

terms 13:19 45:17

64:19 65:1

66:24,25

75:10,18,24 76:1

81:14 82:6 87:18

94:24 99:8 107:4

108:1,7,15,24

109:12 118:23

151:18 152:2,7,9

153:19 155:23

157:18 158:18

159:15,18,22

161:14 162:1

164:19,23 179:5

193:4,15 205:2

211:4 235:6

273:2 283:12

286:8,10

terrain 285:16,18

Terry 6:5,7

116:17 118:9

126:22 161:18

191:9 192:6 210:10

213:9 286:24

terse 93:19

test 49:16 101:10

104:10 106:5

197:10 198:5 260:2

270:1

testing 107:13

text 34:2 57:15 than...that 272:24 thank 3:13,19

4:3,12 5:11,16

7:17 8:9 9:23

13:16 20:14 23:6

25:20,21 26:1,9

31:5 33:4 37:18

39:10,17

44:19,20

46:15,17,18,20

52:10 55:17,19

57:19 60:20

62:14 63:21

67:13,14,17

69:8,13

74:8,9,24 79:10

84:21 87:8,10,24

97:20 102:13

109:25 116:16

118:9,13 120:8

127:16 131:6

134:18 150:9

156:19 160:21

171:13 174:20

178:16 180:9 191:9

206:25 213:8

220:15 233:13

234:4 239:20,21

242:15 251:22

252:4 254:24

259:13 274:22

278:23 284:25

297:1 298:17

thanks 9:2 33:2

67:15 101:22

127:12 131:7

148:10 153:24

157:23 166:24

251:24 260:7

that...are 104:6 that...does 289:22 that...for 228:19

229:7

that...I 112:25

268:2

that...in 228:19 that...it 283:25 that...levels

222:12

that...of 106:12 that...on 235:3 that...that

212:21,22 243:17

254:9 276:7 288:17

that...that's

287:13 291:19

that...what 234:9

that's 8:23 9:4

12:13 14:9,17 15:5

20:1 32:5 35:21





42:1 43:16,19,24

44:13 45:19,22

46:13,23 47:6,16

49:12 51:6,7

52:9 53:1 54:3

55:19 56:20

58:12 60:19

63:14,17

66:10,14 70:8

71:10,11

74:2,4,7

75:20,21 76:2,18

77:15 79:8 80:3

81:5 85:23

86:22,24

91:19,21 92:6

93:23 95:19

96:12,14 97:13

98:5 103:18

107:9 109:1,3,23

110:10 112:9

113:23 116:13

117:11 118:8,17

120:6 121:4

122:1 123:9 124:17

126:4 127:11 128:2

130:3,25 131:2,3

133:1 135:13

136:18 137:5,21

139:12 145:23

146:15 149:10

153:22 156:24

157:1 162:2,25

163:7 165:19,24

166:3 167:23

168:22 169:25

170:5 171:1 173:25

174:9 175:1

176:3,4,6 177:23

178:6,11

180:6,25 182:8

184:17 186:4,22

188:16 189:1,5,6,9

191:2 194:1,16

195:5 196:5,14

199:8,12 202:16

203:16

206:21,23,24 207:1

208:8,20,21 212:19

214:16

217:13,15,16,21

218:18 223:23

224:12 229:9,18

232:13,23,25

233:15 234:2

235:15 237:5

238:10 239:13

240:10,18 241:2

245:10,23 246:13

247:4,21 248:8

249:14,24 250:4

251:21 253:13

254:12,16,18,19

255:1,23

256:14,19,25

258:12,16 260:6,21

261:3 262:12

263:7,8 264:9,20

266:4

267:6,22,23 268:13

273:16 275:6

279:18,19 286:1

287:11 288:21

289:4,11

291:9,19 292:20

296:15 298:5

that's...I 263:2

that's...that's

246:3 257:12

the....so 122:20

the...and 95:22

108:3 277:8

the...at 223:24

269:5

the...based 221:6 the...but 292:14 the...does 273:1 the...from 284:4
the...going 270:11 the...I 106:22 the...if 274:6 the...i'm 271:13
the...I'm 230:14 the...in 241:1,18

273:20

the...of 269:8

the...on 228:16 the...or 269:22 the...so 245:10 the...that's

98:24 270:13

the...the

215:8,15 218:19

221:9 222:5

227:3,17 228:16

229:3 230:25

231:12,17,24

232:25 243:11

245:8 253:15

254:25 255:7,20

258:18,23 263:11

269:5 273:19 276:8

284:19 294:21

the...the...the

250:8

the...these 294:1 the...this 234:2 the...those 280:21 the...to 264:22
the...unless 256:25 the...we 223:7

273:11

the...whatever

286:4

the...when 104:12

285:15

the...your 283:22

296:9

the..the 291:2 them...break 236:3 theme 26:21,25 themselves 6:24

8:22 37:15 135:7

217:8 226:24 270:8

then...then 264:25 theoretically 271:6 there...and 256:1 there...if
222:7 there...what 120:16 there..is 104:6 therefore 34:10





36:10 41:8 89:8

104:5 107:24

193:21 195:8

279:17

there's 9:10

12:11 24:10

32:18 35:9,12

38:7,19 45:10,23

49:1,3 51:13 53:12

56:18,19,20

59:10 60:21

62:10 64:5

67:5,8 70:5,18

71:6,21,24,25 72:3

74:7 80:9,21

81:1,17 86:6,21

87:6 89:2

94:7,22 98:1

101:15 103:1,12

104:21

105:7,10,19,20

106:25 108:6,8

111:3,25 112:14

115:20 116:9,10,12

117:22 120:15

121:23 122:6

123:10

125:11,18,19

126:14 129:12

130:2,3,5 131:19

132:7,23 133:8

134:24 135:1,4

136:5,12

137:10,11,12,13

138:6 141:23

144:12 157:18

158:7,23

160:8,12,17,18

172:7 177:8

178:7,21 180:17,18

182:7 186:14

187:20 190:11,24

193:21 194:8

197:24 198:8,12,14

199:3,4 200:13

204:22

213:15,18,19,24

218:9 231:25

235:21 236:18

237:10 245:16

255:4,9,15

257:10 259:19

265:16 267:2 271:1

274:13 277:24

279:14,15 280:15

290:5 292:20

298:1,2

there's...there's

245:4 266:11

these...an 261:9 these...at 104:1 these...these

253:13

they...if 272:21 they...that... from 268:6

they...they 280:23 they'll 184:24 they're 6:4 8:16

24:4,5 54:5

61:10 65:1,3 66:18

86:15,16 90:21

91:10,11 100:9

113:6 130:16

136:20 165:13

191:7 194:9

198:19,21

203:21,25 206:17

210:25 211:8 215:3

217:25 238:11

239:1,18

241:8,10 271:2,6

276:22 277:9

279:16 289:14

296:22

they're...they're

276:9

they've 58:3

151:7 193:8

thickheadedness

83:19

thing...one 115:12 think...I 272:16 third 19:2 35:1

39:12 234:17 242:6

243:10,11,12

265:24

thirty 46:9 this...for 229:4 this...have 216:8 this...on 116:11

286:23

this...or 278:20 this...that 115:24 this...that's

259:19

this...the 123:4 this...these 228:6 this...this 258:2

287:9

this...we'll 216:23

THOMAS 73:3,13,21

THOR 298:3

thorough 55:14,19

153:13

those...the 255:10

those...those

281:22 297:11

thoughtful 172:5

thousand 107:16

131:2 133:11

thousands 273:12

291:12

threshold 96:3

106:4 135:22

180:19,24

181:7,18,19,22

182:10 184:2,3,5

250:2

thresholds

181:11,15 182:8

threw 112:11 202:2

throughout 12:6

21:6 33:15 34:12

50:15 111:20 131:1

169:13 298:6

throw 150:6

186:16 293:10

throwing 75:18

190:12

Thurston 6:19 57:21





61:9 88:1 92:17,24

93:8,11 125:15

127:14,16 160:23

161:23 183:14,17

189:3

275:3,9,14,17,21,2

4 276:16,20

277:2,14,18

Thurston's 114:23 thus 274:10 tidbits 241:6

tie 110:8 121:5 tied 180:2 tightening 20:19 tightly 214:3

til 72:13

Tim 8:4,9 9:13 69:9

74:9 85:4 92:8

254:20 258:17

259:14 262:7

264:12 265:20

280:11 297:3,14

time-line 42:19

timely 41:13

title 172:19 177:1

titles 55:5 titling 176:13 to...going 220:21 to...I 296:17 to...if
244:16 to...I'm 115:14 to...this 291:17

to...to 222:3 227:7

239:9 250:6 251:13

253:16 266:8

289:17

to...unless 124:12 to...want 293:1 to...we 284:22 tobacco 121:6

today 3:7

4:16,18,20

5:4,21 40:8,19

44:14 56:6

189:23 222:3,17

223:24 234:1

298:18

today's 3:22

129:8,9

toes 11:14

token 100:19

Tolbert 30:2

Tom 6:25 14:18,22

15:1 20:15 73:4,12

tomorrow 103:25

107:23 136:19

137:7 223:24

269:15 279:5,18

281:7,9,10 298:10

tomorrow...this

103:25

tonight 60:13

218:4,6,17

269:15 295:6,12,19

298:9

tool 24:2,5 258:8

tools 193:13

top 50:1 155:15

208:14 269:14

topic 3:22 102:6 topics 20:23 234:6 topology 285:12 torn 166:25 167:1

total 108:14 231:16

241:24 249:22

total...a 249:21 totally 147:8 289:7 touch 215:15

touched 76:17

271:13,17

tough 121:7,17

123:20 189:17

194:8

towards 22:15

149:14 167:17

232:25 280:16

town 255:20

tox 14:17 60:23

102:17 114:16

116:14 126:2

186:24 190:11

213:24

toxic 29:9 101:13

toxicologic 116:6

toxicological 21:17

61:4 121:10 134:23

188:3 201:22

206:12 207:24

toxicology

30:9,13,16 31:12

58:9,11 59:18

61:18,20,25

62:12 118:20,22

119:2 124:24 125:5

129:1,12 132:11

134:21

track 49:2 109:18 tracked 152:20 tract...I'm 275:7 tradition 198:12
traffic 20:10 64:14

91:8

245:1,2,9,20

255:20 257:18

264:17 280:18

284:8 285:25 286:4

traffic-related

190:25

transformations

94:8

transition 56:13

178:24 179:15

transitional 180:6 transitions 56:10 translate 108:10,14 translated 71:4
transport 255:7 transportation

64:19

trap 128:21 trash 207:9 travel 261:14

traveling 276:22 treated 159:2 treatment 250:21 tremendous 113:16 trend
211:20 212:6 trends 16:14





trials 207:23

137:20 138:10

typo 256:25

tried 11:6 20:18

145:2 149:13,14	 	

38:3 48:14 96:18

101:19 119:8

131:13,16 133:4

139:24 143:3

144:23

196:2,3,21 244:24

trigger 205:2 triggered 49:19 triple 34:16,25

36:4

trouble 8:10 81:9

182:2

troubled 143:18

true 60:14 65:10

178:11 235:1 257:6

273:25

truly 122:5 trust 241:8 truth 284:4

try 24:2 38:23

44:15 47:1 52:7

61:7 80:12,13

82:16,18 91:7

101:11 113:18

129:10 139:22,23

140:22 147:19

155:11 160:14,15

163:8

185:8,10,13

196:4 201:18,19,21

207:11 209:7,10

210:24 213:20

219:17,22 236:22

240:9 244:23

289:9,10 290:6

293:25 295:23

298:7

trying 15:24

43:20 47:18

52:18,20 80:1 81:9

86:23 95:15 100:20

106:14 107:12

108:15 110:8

123:9,14 128:24

130:23 133:24

155:25 157:22

163:22 170:7

180:15 181:25

183:13 188:11

202:6,8 207:8

213:7 234:10

240:10 245:13

268:22 273:22

277:17 279:12

284:10 285:10

288:20 289:1

291:17 292:6

trying...the 107:11

TSD 253:15

tuck-away 47:12

Tuesday 90:24

turf 200:12

turn 3:16 7:18 8:16

20:13 26:3

173:13 221:4

224:15 237:9

254:23

turned 133:15

162:14

Turning 48:3

turns 80:19 83:8

264:3,4

twenty 105:24

112:3,6

twice 263:14 twitchy 155:17 two...two 252:1 two-fold 225:12 two-step
12:11 type 78:21 81:16

135:22 146:10

types 19:25 30:10

82:19 87:13

92:5,25 102:1

209:1 237:15

289:16

typical 27:15 297:9

typically 44:5

66:10,14,19 77:4,5

279:11

U

U.S 2:1 22:4,5

27:24 29:24 66:9

238:17,21 239:25

242:6 252:21 275:6

ultimate 111:18

127:6 197:6

ultimately 156:1

221:23 237:6

253:14

Ultman 7:4

31:8,16,20,25

32:11,14 79:12

95:11 207:18

269:19,20 270:17

271:16 272:20

273:18

ultra 119:19

ultra-fine

161:5,8,16

unambiguous 281:20

uncertain 53:23

123:12

uncertainties

30:19,20 65:6

76:21,24 77:2,19

251:2

uncertainty 28:15

35:3 76:19

123:11 192:15

250:21,24

251:4,5,10,16

265:4 286:9,11

unclear 96:7 122:18

288:22

uncomfortable 68:24

186:9 266:13

uncovered 81:8

undercounting

279:17

underestimating

255:19

underestimation

243:24 246:5

underlying 35:23





97:13 123:18

236:25

underneath 53:13

underprediction

243:21 244:9

246:12

underrepresenting

258:2

understand 16:4

20:2 39:14 47:18

52:7 61:13 62:13

69:1 84:17 101:3

117:7 118:18 146:4

148:2 168:23

188:21 193:4

234:14 240:14

246:19 250:15

270:10 275:18

283:5 292:25

understanding 58:14

61:17 91:17 121:11

122:2 151:19

161:14 197:5 270:3

understood 27:9

52:11 88:7

under-the-lamp-post

153:11

unedited 296:12

unfortunately 23:20

41:6 95:20 96:6

204:4

uniform 11:16

United 16:14

17:24 18:21

73:10 75:6 76:9

168:2,25 224:24

225:21

University

6:9,12,14,16,18

7:5

unless 156:8 262:24

unlike 66:4 85:7

181:4

unquote 36:18 unrealistic 289:25 unscheduled 90:18 unstable...or

256:16

untangle 70:6 up...upwards 233:1 up/roll 252:2 upcoming 61:8

up-front 174:8

upon 60:10 65:18

76:17 131:21

185:12

upper 16:17

235:10 288:8,11

up-regulate 157:4

upset 60:9 upwards 259:8 urban 17:18

18:3,6,13 70:13

255:8 261:25

use...we 223:25

useful 12:14 43:1

44:10 47:22 48:2,7

50:3,5,11 51:4

52:4,7 73:7

87:14 94:24

101:2,4,9,10,15,16

,21 110:6 112:17

130:13,18 132:3,13

134:5,6 177:5,18

189:10 234:14,24

240:17 251:20

253:2 295:3

USEPA 27:12,14

usual 47:25 150:6 usually 39:5 209:19 utility 26:18

233:21

utilize 183:5

V

validate 271:23

272:8

validating 107:25 validations 108:1 validity 28:16

273:20

valuable 124:16

126:23 293:20

value 101:25 150:17

245:19 246:3,12

260:21 293:7

values 70:4 71:17

147:8 207:24 208:1

228:10,11 243:19

246:5 256:17,22

260:5 271:25 291:6

van 218:10,19

Vanessa 2:20 5:16

26:9 217:10

vari...variable

225:4

variability 35:18

86:17,21 92:22

135:3,7 225:3

231:25 250:21,24

261:10

variable 135:1

209:16,17 226:11

273:16

variables 193:20,24

270:8

variance 264:11

variation

17:14,16,18,19

18:3,6 69:23 74:21

88:18 244:1

variations 18:22,23

271:3

varies 18:16 184:3

variety 103:6 157:8

168:7 227:12 267:2

various 11:25

15:7 36:12 81:17

82:22 131:15 168:3

211:9 216:15

225:24 226:11

227:22 259:2

vary 27:23 28:2

85:8

vastly 79:13 178:19

vector 59:14

156:9 161:17

vectors 161:6

vehicle 196:13

283:14

vehicles 285:5





ventilation 28:6

98:20

venues 193:12 verbally 62:23 verify 163:3 versa 214:13 version 47:24

101:19 227:15

260:10

versus 27:16

49:10 57:9 71:1

72:22 75:17 88:3

90:1 98:4 157:20

209:17 224:13

255:17 272:1 293:3

very...it's 257:7 vetted 65:16,21 vice 214:13

view 38:1 115:6

161:12 173:15

237:2 270:2

views 150:17 215:7 viral 152:9 virtually 157:5

vis 130:19,20 visibility 283:9 visits 22:16,22

23:20 37:10 208:13

vu 2:20 3:17,19

26:8 48:10

vulnerability

159:9,18 160:25

vulnerable 149:4

150:19 159:14

161:2,15 163:6

164:18,22,24

165:2,6,9,14,17,18

,20 168:4

 	W

wait 72:13 249:6

265:11

waiting 92:14 93:13

walk 206:12 warped 213:13 warrants 155:12

was...and 252:6

was...that 284:14 was...the 212:5 was...this 272:23 was...was 212:5

225:18 231:5

251:25 252:15

was...why 246:14

Washington 6:16

wasn't 32:14

56:25 57:3,8 68:17

80:5 86:6 114:13

125:22 146:5 153:3

166:1 191:24

201:13 210:2,5

211:17 212:2

225:20 228:5

230:16 250:17

252:7 256:1 259:23

263:3 276:7 286:17

wasn't...and 289:20 wasn't...it 250:16 wasn't...on 255:24 way...can't
271:22 way...to 236:14

ways 128:5 185:17

196:11 198:9 210:7

211:5,9

we...and 295:15 we...I 247:22 we...I'd 109:14 we...so 250:3 we...we
222:20

223:18 224:5 229:8

250:3

we...what 130:19 we...why 222:6 weak 28:3 83:22

85:9 107:8

weakness 206:16

weaknesses 170:17

171:9 206:4,16,19

web 217:4 219:16

276:5

website 233:24

we'd 42:25 43:21

44:4 260:14

week 33:7 216:19 weigh-in 66:15 weight 30:20

40:12 98:19

152:6,23

weighted 89:15

90:12 124:25

weighting 124:14,22

weird 191:20

welcome 2:13 3:20

5:17 8:1 26:8

welcomed 93:19

welcoming 3:17

we'll 5:23 6:2 7:18

8:12 9:8 26:3 43:4

48:15,19 61:7

76:12 99:12 103:25

112:16 116:17

123:22 127:14

148:21 160:3

163:8,12 185:13

188:18,19

201:19,21 207:11

215:14 216:25

219:22 242:17

259:9 265:11 275:1

281:10 286:23

well-applied 50:15 well-below 62:5 welll 231:16

we're 5:20 6:4

10:2,4

12:5,20,23

14:16,23

21:1,2,3

45:17,19 46:17

54:12 78:5 82:24

85:10 86:23

87:2,11 100:20

104:23 105:5

109:2,18 110:25

114:4 115:9 118:10

119:5 121:7

122:7 136:10,11

138:20 139:18,19

140:2 145:17,25





146:22 148:16

159:13,25 166:11

169:23 170:7

171:24 172:10

174:13,23 176:3

181:15 182:20

187:22 189:7

192:9,13,14 193:25

194:17,19 197:22

200:7,8 201:12

202:7 203:18

207:11 215:1

216:18,23

217:13,14 218:1

219:6,16 220:4

221:11 222:2,11

232:19 234:19

237:24 242:20

245:6 262:24

264:15,23 268:11

272:18 278:24

279:12,17 288:24

289:1,19

290:8,20 291:16

292:6 294:1

were...there 228:13 were...we 284:10 were...were 248:1

282:8

we're...we're

221:15 276:12

we're...we've

104:23

we're...when 218:10 were...you 271:12 weren't...in 245:12 we've 3:9
10:5,25

11:22 14:9,10

21:24 23:25

46:10 85:20 94:2

96:17,18 101:12

110:10 127:18

138:18 139:24

141:20 158:4,6

161:25 162:3

172:25 219:5

220:17 221:21

222:19 227:24

228:9 230:25 235:4

274:25 290:13,14

293:15,23 294:11

295:15 296:25

297:6

we've...because

130:15

we've...we've

221:23

what....and 109:11 what....I 122:18 what...we 264:5 what...what 235:6
whatever 39:1

229:15 236:24

242:7 258:10 262:9

267:11 296:10

whereas 240:15

where's 137:11

WHEREUPON 100:1

139:1 220:2 298:19

whether 38:5 49:9

50:20 52:3,5 53:17

60:7,24 71:3 72:25

77:23 96:7,19

107:12 121:2,4

122:22 123:17

124:16 125:16

126:6 131:18,25

132:4,12

133:7,20,21 135:20

139:19 140:5

142:12 144:9,25

145:2,14 147:9

150:5 151:13 158:4

173:18 174:23

181:4,19

182:7,13 200:21

201:11 227:9

232:22 244:9,10

246:5 248:12

250:14 256:23

263:3,4 267:23

285:11 293:2

Whew 274:25

which...is 289:22

while...while 291:7

white 16:16 79:19

162:8,10

whites 162:16

whittling 183:7

Whoever 181:3

whole 38:8 85:1

87:1 88:3 94:23

117:2 132:13

133:14 155:1

156:14 171:19

186:15 206:17

207:20 208:24

213:12 233:14

237:15,20 238:17

252:2,8 262:1

273:12 276:18,20

282:8,9

whole...oh 286:24

who's 7:24 159:14

294:17

why...I 246:19

252:12

why...that's 214:16

wicket 155:25

wide 37:13 51:22

59:7

widely 12:6 27:23 wide-range 35:10,14 willing 171:7

wind 78:14 85:12 winter 17:10 77:7 wintertime 77:6 wise 101:24 102:10
wish 128:8 207:4 with...with 252:25 woman 164:7

wonder 184:11 186:4

287:7

wondered 288:10

wonderful 53:20

252:13 297:11,20

wondering 71:3

72:2,23 271:11

275:3

worded 167:20





work 3:14 5:18,20

8:12 58:6 60:7,8

80:6,10,21 98:24

100:18 109:15,24

110:10 117:23

133:15 139:2

151:16 183:25

206:8 210:18 216:6

217:17 280:1

288:3,4 298:18

worked 14:17 15:6 workers 275:20,21 workinformation

275:4

working 21:25 95:13

97:15 162:17,18

295:21

works 95:18 100:20 workshop 41:20 world 60:13 61:22

62:8 94:9 114:24

115:2 129:3 253:3

worried 65:25 179:9

266:1 267:8

worry 81:4 131:24

worst 237:12,23

238:1,6,11 239:2

241:14 257:21

260:4

worth 49:12 60:5

68:1 70:22 93:24

186:4 191:22 223:6

would...I 255:22

264:17 265:9

would...this 229:18 would...what 136:24 would...would 297:9
wouldn't...it

285:20

wow 206:20

write 145:16 188:22 writers 164:19 writing 50:22 172:5

215:25 243:1

295:9,24

written 3:8,11 12:1

26:24 33:6 35:19

36:22 37:12

40:3,23 54:3 61:23

62:22 68:17

70:11 83:14 88:6

117:2 120:10

128:20 129:11

130:2,9 148:24

149:1 150:13,14,15

164:15 166:5,6

169:5 184:8 190:20

201:11 218:13

234:5 278:11

296:13 297:22

wrong 65:25 66:21

95:21 98:15 111:25

112:9 128:5 138:11

141:1 165:6 178:15

186:17 211:11

214:15 215:16,21

244:4 258:13

280:22

wrote 58:23 181:3

185:17 186:4

  HYPERLINK http://www.epa.gov/casac  www.epa.gov/casac 

233:25

Wyzga 7:13 87:10

133:12 178:18

288:1 289:24

Y

Yang 24:10

yards 284:20,21,23

yell 154:5

yet 15:19 107:1,5

167:1 175:1

York 162:4 257:15

you...because

120:20

you...how 293:11

you...I 280:12

288:4

you...if 249:20 you...that's 249:12 you...well 265:10 you...what 214:4
you...where 135:21 you...you 108:17

252:13 260:18

you...your 249:7

296:20

you...you've 218:12

246:2

you'll 43:5 94:3

136:22 216:22

233:25 236:5

277:24

young 58:23 younger 149:17 your...in 257:22 your...your

252:24 298:8

you're...that

257:25

yours 269:18 yourself 6:1 yourselves 220:7 you've 10:3 13:21

40:19 64:4

74:15,19 81:24

84:5 86:3 87:4

93:13 97:24 99:7

105:3 119:8

120:1 134:3 144:13

156:12 219:1,6

256:20 265:21

268:18 272:4,5,6

279:20,25 280:1

281:21 288:7,14

293:1,3,21 296:13

Z

zero 31:17,19

128:14 180:25

Zieger 57:3



 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EPA CASAC MEETING 05/01/08 CCR#15905-1	  PAGE  28 

EPA CASAC MEETING 05/01/08 CCR#15905-1	  PAGE  29 

 

 

110

EPA CASAC MEETING 05/01/08 CCR#15905-1	30

 

 

114

EPA CASAC MEETING 05/01/08 CCR#15905-1	31

 

 

118

EPA CASAC MEETING 05/01/08 CCR#15905-1	32

 

 

122

EPA CASAC MEETING 05/01/08 CCR#15905-1	33

 

 

126

EPA CASAC MEETING 05/01/08 CCR#15905-1	34

 

 

130

EPA CASAC MEETING 05/01/08 CCR#15905-1	35

 

 

134

EPA CASAC MEETING 05/01/08 CCR#15905-1	36

 

 

138

EPA CASAC MEETING 05/01/08 CCR#15905-1	37

 

 

142

EPA CASAC MEETING 05/01/08 CCR#15905-1	38

 

 

146

EPA CASAC MEETING 05/01/08 CCR#15905-1	39

 

 

150

EPA CASAC MEETING 05/01/08 CCR#15905-1	40

 

 

154

EPA CASAC MEETING 05/01/08 CCR#15905-1	41

 

 

158

EPA CASAC MEETING 05/01/08 CCR#15905-1	42

 

 

162

EPA CASAC MEETING 05/01/08 CCR#15905-1	43

 

 

166

EPA CASAC MEETING 05/01/08 CCR#15905-1	44

 

 

170

EPA CASAC MEETING 05/01/08 CCR#15905-1	45

 

 

174

EPA CASAC MEETING 05/01/08 CCR#15905-1	46

 

 

178

EPA CASAC MEETING 05/01/08 CCR#15905-1	47

 

 

182

EPA CASAC MEETING 05/01/08 CCR#15905-1	48

 

 

186

EPA CASAC MEETING 05/01/08 CCR#15905-1	49

 

 

190

EPA CASAC MEETING 05/01/08 CCR#15905-1	50

 

 

194

EPA CASAC MEETING 05/01/08 CCR#15905-1	51

 

 

198

EPA CASAC MEETING 05/01/08 CCR#15905-1	52

 

 

202

EPA CASAC MEETING 05/01/08 CCR#15905-1	53

 

 

206

EPA CASAC MEETING 05/01/08 CCR#15905-1	54

 

 

210

EPA CASAC MEETING 05/01/08 CCR#15905-1	55

 

 

214

EPA CASAC MEETING 05/01/08 CCR#15905-1	56

 

 

218

EPA CASAC MEETING 05/01/08 CCR#15905-1	57

 

 

222

EPA CASAC MEETING 05/01/08 CCR#15905-1	58

 

 

226

EPA CASAC MEETING 05/01/08 CCR#15905-1	59

 

 

230

EPA CASAC MEETING 05/01/08 CCR#15905-1	60

 

 

234

EPA CASAC MEETING 05/01/08 CCR#15905-1	61

 

 

238

EPA CASAC MEETING 05/01/08 CCR#15905-1	62

 

 

242

EPA CASAC MEETING 05/01/08 CCR#15905-1	63

 

 

246

EPA CASAC MEETING 05/01/08 CCR#15905-1	64

 

 

250

EPA CASAC MEETING 05/01/08 CCR#15905-1	65

 

 

254

EPA CASAC MEETING 05/01/08 CCR#15905-1	66

 

 

258

EPA CASAC MEETING 05/01/08 CCR#15905-1	67

 

 

262

EPA CASAC MEETING 05/01/08 CCR#15905-1	68

 

 

266

EPA CASAC MEETING 05/01/08 CCR#15905-1	69

 

 

270

EPA CASAC MEETING 05/01/08 CCR#15905-1	70

 

 

274

EPA CASAC MEETING 05/01/08 CCR#15905-1	71

 

 

278

EPA CASAC MEETING 05/01/08 CCR#15905-1	72

 

 

282

EPA CASAC MEETING 05/01/08 CCR#15905-1	73

 

 

286

EPA CASAC MEETING 05/01/08 CCR#15905-1	74

 

 

290

EPA CASAC MEETING 05/01/08 CCR#15905-1	75

 

 

294

EPA CASAC MEETING 05/01/08 CCR#15905-1	76

 

 

298

EPA CASAC MEETING 05/01/08 CCR#15905-1	  PAGE  85 

