Appendix XX.

NATURAL Background Visibility Conditions

Considerations and Proposed Approach to the calculation of Natural
Background Visibility Conditions at MANE-VU Class I areas



Natural Background Visibility Conditions

1.  Introduction

The long-term visibility conditions that would exist in absence of
human-caused impairment are referred to as natural background visibility
conditions.  Accurate assessment of these conditions is important due to
their role in determining the uniform rate of progress that must be
considered when setting reasonable progress goals for each mandatory
Federal Class I area subject to the Regional Haze Rule.  Baseline
visibility conditions – based on monitored visibility during the five
year baseline period (2000-2004) – and estimated natural background
visibility conditions will determine the uniform rate of progress to be
considered when setting reasonable progress goals for any Class I site. 

The U.S. Environmental Protection Agency (EPA) issued draft
methodological guidelines for the calculation of natural background and
baseline visibility conditions at each site as well as methods for
tracking progress relative to the uniform rate of progress that these
values determine. This draft guidance, issued in September 2001 was
subsequently finalized in September 2003.  The final guidance recommends
a default method and allows for certain refinements that states may wish
to pursue in order to make these estimates more representative of a
specific Class I area that may be poorly represented by the default
method.  

This appendix provides a description of the default method for
calculation of natural background conditions.  In addition, the default
method is applied to each Class I area in or near the MANE-VU region in
order to establish default natural background conditions on the twenty
percent best and worst days.  A discussion of potential refinements to
the default method is presented along with rationale for their
consideration.  The uncertainty associated with each potential
refinement is then considered in the context of the overall uncertainty
of the default estimates. Finally, a recommendation for estimating
natural visibility conditions to be included in this SIP is provided. 

Based upon these analyses, as well as comments received on the draft
MANE-VU proposal, it appears that while some aspects of the default
calculation method are understood well enough that they could be
considered as potential refinements, MANE-VU does not feel these
refinements are warranted in light of the very large uncertainties
associated with the most basic elements of the default estimates
(naturally occurring ambient concentrations).  The identified
refinements would result in substantial differences relative to default
estimates without significantly improving the accuracy of our estimate
relative to the default.  Rather, MANE-VU advocates a proposed approach
that is based on use of the default estimates while a program of
research is undertaken to refine those elements which are most uncertain
(natural concentrations) in order to reduce the overall uncertainty as
better scientific understanding of these issues evolves.  Refinements to
other aspects of the default method (e.g. refinements to the assumed
distribution or treatment of Rayleigh extinction, inclusion of sea salt,
and improved assumptions about the chemical composition of the organic
fraction) may be warranted prior to submissions of SIPs depending on the
degree to which scientific consensus is formed around a specific
approach and will be reconsidered at a later point. 

2. The Default Method

The default method is explained in detail in Estimating Natural
Background Visibility Conditions (U.S.EPA, 2003).  Summary information
is provided here but the reader should consult the original guidance
documents for any question as to how this method is applied.

Estimates of natural visibility impairment due to fine and coarse
particles were derived using the 1990 National Acid Precipitation
Assessment Program reported average ambient concentrations of naturally
present particles (Trijonis, 1990). Separate concentration values were
given for the Eastern and Western United States, no finer spatial
resolution is available. Average natural background light extinction due
to particles was then calculated using the IMPROVE methodology and site
specific ANNUAL f(RH) values. Worst visibility levels are derived using
the work of Ames and Malm (2001), who estimated the standard deviation
of visibility in deciviews in the eastern US as 3 dv. By assuming a
roughly normal distribution of data, the default method adds (subtracts)
1.28*(3 dv) to the average estimated natural background to calculate the
90th (10th) percentile level which is taken by EPA to be representative
of the mean of the twenty percent worst (best) conditions. 

Thus in the East, the default method for calculating best and worst
natural background visibility conditions (in deciviews) for any area in
the Eastern U.S. would use the following formulae:

P90 = HI +1.28 sd

P10 = HI – 1.28 sd

Where The Haze Index (HI) represents annual average visibility in units
of deciview and sd is the standard deviation of daily average visibility
values throughout a year, defined by the guidance as 3.0 for the Eastern
U.S.  The Haze Index is calculated as shown:

HI =10 ln (bext/10)

where the atmospheric extinction, bext, is given by the familiar IMPROVE
equation (IMPROVE, 2000) in inverse megameters:

bext = (3)f(RH)[sulfate] + (3)f(RH)[nitrate] + (4)[OMC] + (10)[LAC] +

(1)[SOIL] + (0.6)[CM] + 10

 for species assumed to be non-hydroscopic), and refinement to the
assumed distribution of visibility conditions throughout the year
(including the width, amplitude and potentially shape of the
distribution).  Potential refinements are considered in section 4.

Table 1 below provides the default values to be applied at all Eastern
U.S. Class I areas.  The result of using these default values in the
above equation with an assumed annual average f(RH) value of 3.17 (the
average of 11 Northeastern U.S. sites) default estimated visibility in
the Northeastern U.S. is approximately 3.6 dv on the twenty percent best
days and 11.3 dv on the twenty percent worst days.

3. Application of the Default Method

The Class I areas in the MANE-VU region that are subject to the
requirements of the regional haze rule are: Acadia National Park, Maine;
Brigantine Wilderness (within the Edwin B. Forsythe National Wildlife
Refuge), New Jersey; Great Gulf Wilderness, New Hampshire; Lye Brook
Wilderness, Vermont; Moosehorn Wilderness (within the Moosehorn National
Wildlife Refuge), Maine; Presidential Range – Dry River Wilderness,
New Hampshire; and Roosevelt Campobello International Park, New
Brunswick.  In addition to these Class I areas, we consider several
nearby Class I areas where MANE-VU states may be contributing to
visibility impairment.  These Class I areas include:  Dolly Sods
Wilderness and the Otter Creek Wilderness in West Virginia as well as
Shenandoah National Park and the James River Face Wilderness in
Virginia.  MANE-VU understands that it is the responsibility of the
appropriate VISTAS states to establish estimates of natural visibility
conditions and reasonable progress goals for these areas.  It is
anticipated, however that subsequent consultations will occur with those
MANE-VU states which may be affecting visibility in these areas. 
MANE-VU has therefore calculated estimates of natural background
visibility conditions at the nearby sites using MANE-VU approved methods
in order to facilitate future consultations.  

The only factor in the default method that varies by site is the
climatological annual mean relative humidity adjustment factor.  Table 2
lists this value for the Class I sites of interest and the resulting
best 20 percent and worst 20 percent estimates of natural visibility
conditions.  The variation among sites using the default method is
purely a function of differences in climatological annual mean relative
humidity, with southern and coastal sites being more humid than inland
or elevated sites.

 

4. Potential Refinements

According to the guidance (U.S. EPA, 2003), “… the default approach
to estimating natural visibility conditions presented in this document
is adequate for the development of progress goals for the first
implementation period under the regional haze rule.” However, the
guidance does leave the door open for individual states or RPOs to adopt
their own methods for calculating natural background if they can
demonstrate that the change from the default represents a significant
refinement that better characterizes natural visibility conditions at a
specific Class I site.

The five Regional Planning Organizations have identified a number of
areas for potential improvement and have hired a contractor to refine
the understanding of natural background levels of particulates.  The
statement of work for this project (managed by the WRAP) includes the
following text: “There are three broadly different ways to refine the
default natural aerosol concentrations that are briefly discussed in the
guidance document.  The default annual estimates of species
concentrations for the best and worst 20% haze conditions can be
replaced by better annual estimates, by seasonally varying estimates, or
by event-specific estimates (e.g. in the case of forest fire and dust
storm impacts).  Any technically defensible combination of these
different ways to refine the natural aerosol concentration is
acceptable.  It is likely that refinement will be a multi-step process
over a period of many years as the information required to justify
changes are developed and reviewed.”  The three methods of refinement
noted in this statement of work, mirror those listed in the guidance,
however, the guidance also states that, “states may identify other[
refined approaches] that are more appropriate to their own
situations.”

As noted in section 2, in addition to different ways to adjust ambient
concentration estimates, the relative humidity adjustment factor and the
shape of the distribution would also affect the resulting estimates of
naturally occurring visibility.  The VISTAS RPO has commissioned a
consultant to investigate potential refinements to natural background
(Tombach, 2003). In addition, a white paper developed by EPRI on this
topic and a recent presentation by Bill Malm of CIRA (a principal
investigator of the IMPROVE program) all serve to inform the multitude
of ways that calculations for natural background conditions could be
refined (Malm, 2004; Kumar, 2004).  A synopsis of several potential
refinements and the rationale for their consideration are presented
here.  For more detailed discussion of the scientific merit of each
potential refinement, the original references cited above (or those
contained in the brief explanations below) should be consulted.    

Increase the value of the organic multiplier 

The estimates of organic carbon mass that are used in the guidance are
derived from Trijonis (1990), however his original estimate (1.5) has
been adjusted to be consistent with the ratio of organic carbon
mass/organic carbon that is used in the IMPROVE program.  This value,
1.4, is uncertain and several review articles and studies (Watson 2002,
Turpin and Lim 2001, Malm 2004) have suggested higher values between 1.8
and 2.1 are more appropriate values.  If a higher value were to be used
for the organic carbon multiplier, the estimate of natural background
organic carbon mass would be similarly affected since the original
Trijonis estimate was based on organic carbon, [OC],  and a
multiplicative factor which relates [OC] to organic carbon mass, [OCM]. 


Adjust the factor used to translate average visibility conditions into
twenty percent worst or best conditions

The guidance recommendation for calculating the twenty percent worst and
best visibility conditions by multiplying the average by 1.28 times the
standard deviation of 3.0 assumes a normal distribution and is designed
to return the 90th percentile value in that distribution.  The Regional
Haze Rule requires improvement on the average of the twenty percent
worst days.  This value is not equivalent to the 90th percentile of a
normal distribution.  The 92nd percentile is closer to the simple
average of the top twenty percent of values, if you assume a normal
distribution (Lowenthal et al., 2003).  In this case, a factor of 1.40
is more appropriate for calculation of the 92nd percentile, or the mean
of the top twenty percent of values.  However, it is clear that the
distributions of visibility conditions at most Class I sites are not
perfectly normal.  In fact, the 90th percentile may be closer to the
average of the top twenty percent of visibility conditions at sites that
do not experience as many extreme visibility conditions as a normal
distribution would predict (Malm, 2004). 

Account for visibility impairment due to sea salt at coastal sites

Many Class I sites are located along the coast and are significantly
affected by coarse mode sea salt particles.  The tail of the coarse mode
sea salt particle size distribution is within the sub-2.5 micron size
fraction and should properly be included the IMPROVE equation.  This
would be a straightforward refinement if we assume that all sea salt is
in the form of sodium chloride (NaCl).  However, significant evidence
suggests that a substantial portion of the sodium along the Gulf Coast
is associated with sodium nitrate (NaNO3) (Malm, 2004).  As sea salt
particles age, atmospheric chemical processes appear to replace chloride
with other ions, altering both the chemical composition and the
scattering efficiency.

Account for hygroscopicity of sea salt

Research to date reflects a substantial degree of uncertainty regarding
the appropriate scattering efficiency and hygroscopic growth of sea salt
particles.  Refined estimates hold the potential to significantly change
natural background estimates depending on assumed composition and
concentrations.

Account for organic PM of oceanic origin

Observational evidence exists to support the hypothesis that significant
levels of organic precursor gases are emitted over the open ocean which
could potentially increase the natural background levels of organics,
particularly at coastal sites.

Review soil concentrations

Tombach (2003) suggests that fine soil contributions in the Southeast
U.S. are underpredicted by the Trijonis estimate of 0.5 ug/m3.  He bases
this on the estimated impact of Saharan dust and Asian dust that are
subject to inter-continental transport.  The contribution of these
sources to Northeast and Mid-Atlantic sites is estimated to be
significantly less than for Southeast and Western U.S.

Account for episodic inter-continental dust contributions

In addition to contributing on an annual average basis, the Saharan and
Asian dust impacts are likely to be highly variable in time and should
not necessarily be applied on an annual average basis.  Given the
extreme uncertainty in predicting the frequency of occurrence at any
specific site, there is no reliable means of estimating the temporal
frequency for these effects.

Review organic and sulfur emissions from forests

Observational evidence exists for the tropics (mostly the Amazon River
Basin) to suggest that the estimated natural background concentrations
of organics and sulfur compounds may be significantly higher than the
Trijonis values in those areas.  Application of these data to U.S. areas
remains highly uncertain and requires further research.

Improve estimates of organic and elemental carbon released by natural
fires

Global modeling studies have produced estimates for organic and
elemental carbon released by natural fires that are consistent with the
Trijonis estimates used by EPA.  Nonetheless, these studies as well as
the Trijonis estimates remain uncertain and could be refined through
further research efforts.

Account for inter-continental sulfate and nitrate contributions

Techniques to account for the fraction of light scattering and absorbing
PM that results from extra-jurisdictional anthropogenic emissions (i.e.
Canada, Mexico, Asia) could be developed with the same rough level of
uncertainty that is used in the current default method for calculating
natural visibility conditions.  This is less a refinement of natural
background, however and more of a policy decision as to how natural
background conditions are defined and what is an appropriate planning
goal.  The definition in statute and planning goal supported by the
courts should remain as described in EPA guidance.  MANE-VU feels that
international contributions to Class I fine particulate burdens should
be considered in setting reasonable progress goals, not natural
condition estimates.

Use global chemical transport models to refine estimates of natural
ambient concentrations

The use of global models will certainly prove to be a useful tool for
future research into the topic of natural background conditions, but
MANE-VU does not feel that these tools provide a consistent framework to
serve as the basis for a national program.  The uncertainties within the
model structure mirror the uncertainties in observational evidence for
deducing ambient emission levels of specific PM components.  

Refine temporal resolution of relative humidity adjustment
factors/consider observed relative humidity data instead of
climatological average data.

The use of different averaging periods and different relative humidity
data certainly does affect the resulting estimates of visibility
conditions.  Climatologically average data serves to remove inter-annual
variability of humidity from the process of tracking progress.  This
ensures that measured progress is based on changes in pollution, rather
than meteorological variability.  Further research into the most
appropriate averaging period is still warranted.

In addition to the many potential refinements listed above, NESCAUM has
considered one other possible refinement to the default method; the use
of a higher Rayleigh scattering estimates for coastal sites (12Mm-1 are
approximate Rayleigh conditions at sea level; 10 Mm-1 is used for all
sites in the IMPROVE equation). 

Of the multitude of ways that natural background visibility conditions
could be refined, MANE-VU believes that very few can be justified as
significantly improving the accuracy on the basis of current scientific
understanding.  That is not to say that MANE-VU feels that the default
estimates of natural conditions are truly representative of natural
conditions at each site or that each of the potential refinements listed
above does not bear further investigation, but rather that alternative
methods or values for use in calculating more precise values for most of
the refinements listed above are not readily available at this time.   

Research into many of the potential refinements above should be
continued and  MANE-VU intends to continue research on many of these
questions.  However, MANE-VU feels that only a very few of these
potential refinements can justifiably be considered at this time.  These
include an alternative value for the carbon multiplier, the calculation
of the 92nd percentile of a normal distribution to represent the mean of
the top twenty percent worst visibility conditions, and the inclusion of
sea salt at coastal locations and refined estimates of Rayleigh
scattering.  Calculations were performed to evaluate the effect of these
potential refinements to better understand the effect of such changes on
resulting rates of progress and are shown in Tables 3 and 4.

The uniform rate of progress as determined by baseline and natural
background conditions is most sensitive to absolute changes in natural
background estimates (as opposed to baseline conditions), given the
logarithmic structure of the haze index. For example, using data from
Brigantine Wilderness Area 1999-2002 (a four year period that overlaps,
but does not correspond to the baseline period as described by EPA
guidance) the default estimate for baseline visibility conditions is
27.92 dv. If sea salt is included in the reconstructed extinction
calculation, the baseline estimate increases by 0.24 dv to 28.16 dv. 
Changes in natural background resulting from the addition to sea salt at
Brigantine are from 11.28 to 13.40 dv, a difference of 2.12 dv.  The end
result is a decrease of approximately 10.5 percent in the required rate
of progress slope during the initial period. Although the annual rate
decreases by less than 0.03 dv/year, the change over the course of 14
years is 4 tenths of a deciview which is a substantial difference.  The
estimated impact of adding sea salt to Brigantine has the largest effect
of any of the refinements considered here, thus all refinements
considered (on an individual basis) have less than 10.5 percent impact
on the 1st period progress goal.

While the changes in the rate of progress resulting from these
refinements are substantial, the decision to refine baseline conditions
must be based on whether the refinements are statistically significant. 
In order to meet that test, a potential refinement must alter the rate
of progress to the point that the refined value lies outside the range
of uncertainty of the default value.  To implement refinements that do
not meet this test would result in new values that are substantially
different, but not significantly more accurate.   

Very large uncertainties are associated with most of the parameters that
go into the default natural background calculation and many of the
potential refinements.  For example, In the case of a change to the
organic multiplier, different values ranging from 1.4 to 2.1 or higher
have been proposed, however the uncertainty bounds of these estimates
are large  and overlapping (i.e. most estimates are within the
uncertainty bounds of the others and thus are not statistically
different). 

In the case of sea salt, it certainly represents an improvement in
accuracy to include a term for sea salt scattering when we know it to
exist.  Given the potential for complex chemical interaction of sodium
and chlorine with other components of particulate matter, estimates of
uncertainty are difficult to quantify and large (on the order of 50
percent).  While estimated values that would be appropriate for MANE-VU
coastal Class I sites are statistically different from zero, the
resulting improvement in the overall accuracy of the final natural
visibility estimates relative to the default estimates must be
calculated using standard error propagation techniques.  

Following standard error propagation techniques (Taylor, 1982),
estimates of the contribution of each parameter (see table 1) to the
overall accuracy of natural extinction estimates have been derived.  The
fractional contribution of each parameter to total extinction is
presented in Figure 1 for coastal and inland sites in the MANE-VU
region.  As this figure demonstrates, the overwhelming, dominant
contributor to the accuracy of the estimate of total extinction is the
uncertainty in organic carbon mass.   

range of uncertainty, so the true contribution is probably closer to the
“high” estimate than the “low” estimate, but cannot be
quantitatively determined in an easy way.

Based on this general review of uncertainties associated with the
refinements and the potentially substantial effect on rate of progress
slopes that could result from implementing such changes, it is
appropriate to accept the default natural background visibility
estimates as provided in U.S. EPA guidance.  The default estimates
provide a sound, nationally consistent framework on which to base the
regulatory structure of the haze rule that is justified based upon the
current state of scientific understanding of these issues.

Further, EPA recommendations on potential refinements (Pitchford,
personal communication, 2004) suggest that such refinements be broadly
accepted by the scientific community, substantial, practical to
implement and not create arbitrary inconsistencies.  In addition, these
recommendations request that state efforts to refine the default
estimates should not side-track technical efforts on other aspects of
the regional haze program.  Hence, it is appropriate to adopt the
default natural background conditions at present time until broad
consensus on refined estimates of the individual species concentrations
(in particular, organic carbon) is established. 

5. Recommendation

This document reviews potential refinements to EPA guidelines for
calculating natural background visibility conditions and explores how
such refinements are likely to affect calculated rates of progress.  
Based on the currently available literature on naturally occurring fine
particulate matter over the coastal and continental U.S. and a detailed
analysis of the error propagation of such refinements on the resulting
estimates of natural visibility conditions, changes to the default
methods for calculating these conditions will not be undertaken by
MANE-VU at this time.  

MANE-VU recognizes the simplicity of the default approach and supports
future adjustments which better reflect true natural background
visibility levels as the science surrounding this issue evolves and more
accurate information is available to support such changes. In
particular, efforts to reduce the uncertainties associated with
estimates of organic carbon, sulfate and coarse mass are most important
to pursue through future research activities aimed at improving
estimates of natural visibility conditions.  Potential refinements
investigated in this document including the addition of sea salt,
revision of the organic carbon multiplier and improved understanding of
the distribution of naturally occurring visibility conditions rank as a
second tier set of priorities to be addressed through future research.

Based on this review, MANE-VU proposes to adopt the default estimates at
this time, to actively participate in further research efforts on this
topic, and to reconsider our position with respect to natural background
visibility conditions as future scientific understanding warrants.

References

Ames, R. B., and Malm, W. C. (2001). Recommendations for Natural
Condition Deciview

Variability: An Examination of IMPROVE Data Frequency Distributions.
Proceedings (CDROM) of A&WMA/AGU Specialty Conference on Regional Haze
and Global Radiation Balance -- Aerosol Measurements and Models:
Closure, Reconciliation and Evaluation, Bend, Oregon, 2-5 October.

Day, Malm, Kreidenweis  Aerosol Light Scattering Measurements as  a
Function of Relative Humidity  JAWMA 50: 710-716 May 2000

Eldred, Feeney, Wakabayashi  The Major Components of PM2.5 at Remote
Sites across the United States   V1 of AWMA Proceedings of an
International Specialty Conference Jan 1998 Long Beach CA  pp13-26

Haywood Ramaswamy and Soden   SCIENCE V283 26Feb99 1299-1303  
Tropospheric aerosol climate forcing in Clear-Sky Satellite Observations
over the oceans.

Hegg, Livingston, Hobbs, Novakov and Russell  JGR vol. 102 D21
P25293-303   1997  Chemical Apportionment of aerosol column optical
depth off the Mid-Atlantic coast of the United States

IMPROVE, (2000), Malm, W. C., Principal Author, Spatial and Seasonal
Patterns and Temporal Variability of Haze and Its Constituents in the
United States: Report III. Cooperative Institute for Research in the
Atmosphere, Colorado State university, Ft. Collins, CO. May.

Kumar, N. (2004). Recommendations for Natural Background Conditions and
Potential Refinements. Proceedings (CDROM) of MANE-VU/MARAMA Science
Meeting on Regional Haze – Organic Aerosols and Natural Background
Conditions, Baltimore, MD, 27-29 January.

Lowenthal, D. H., and Kumar, N. (2003). PM2.5 Mass and Light Extinction
Reconstruction in IMPROVE. Journal of the Air and Waste Management
Association, Vol. 53, pp. 1109-1120.

Malm, Molenar, Eldred and Sisler  Examining the relationship among
atmospheric aerosols and light scattering and extinction in the Grand
Canyon area  JGR V101 D14 pg19251-65  1996

Malm, Day, Kreidenweis,  Light Scattering Characteristics of Aerosols
as a function of Relative Humidity: Part 1- A comparison of Measured
Scattering and Aerosol Concentrations Using the Theoretical Models. 
JAWMA 50:686-700 May 2000

Malm, W. C. (2004). Recommendations for Natural Background Conditions
and Potential Refinements. Proceedings (CDROM) of MANE-VU/MARAMA Science
Meeting on Regional Haze – Organic Aerosols and Natural Background
Conditions, Baltimore, MD, 27-29 January.

U.S. EPA (2003). Guidance for Estimating Natural Visibility Conditions
under the Regional Haze Rule. EPA-454/B-03-005. September.

Taylor, J. R. (1982). An Introduction to Error Analysis, University
Science Books, Oxford University Press, Mill Valley, CA.

Tombach, I., 2003. On Refining Estimates of Natural Background Light
Extinction in the VISTAS Region, A report prepared for VISTAS, December
04.

Trijonis, J. C. (1990). Characterization of Natural Background Aerosol
Concentrations. Appendix A in Acidic Deposition: State of Science and
Technology. Report 24. Visibility: Existing and Historical Conditions --
Causes and Effects. J. C. Trijonis, lead author. National Acid
Precipitation Assessment Program, Washington, DC.

Turpin, B. J., and Lim, H-J. (2001). Species Contributions to PM2.5 Mass
Concentrations:

Revisiting Common Assumptions for Estimating Organic Mass. Aerosol
Science and Technology,

Vol. 35, pp. 602-610.

 

Watson, J.G. (2002). Visibility: Science and Regulation. J. Air Waste
Manag. Assoc., 52, 628-713.

MANE-VU SIP Template		Natural Background Visibility

APPENDIX XX		June 10, 2004

-  PAGE  18 -

Table 1. Default Parameters Used in Calculating Natural Background
Visibility for Sites in the Eastern U.S.

σNaCl	2.5 m2/s	16%	Haywood, 1999

f(RH)NaCl	~3.2	33%	Assumed same as S, N

Note: the mass estimates presented above are based are on estimates of
fine particulate concentrations that would exist in absence of any
manmade pollution (including Mexican and Canadian emissions) consistent
with planning requirements of the regional haze rule.  MANE-VU accepts
this as an appropriate planning goal and intends to consider the
contribution of international transport in deciding what controls are
“reasonable” under the regional haze program. 

Table 1. Default Parameters Used in Calculating Natural Background
Visibility for Sites in the Eastern U.S.

σNaCl	2.5 m2/s	16%	Haywood, 1999

f(RH)NaCl	~3.2	33%	Assumed same as S, N

Note: the mass estimates presented above are based are on estimates of
fine particulate concentrations that would exist in absence of any
manmade pollution (including Mexican and Canadian emissions) consistent
with planning requirements of the regional haze rule.  MANE-VU accepts
this as an appropriate planning goal and intends to consider the
contribution of international transport in deciding what controls are
“reasonable” under the regional haze program. 

Table 2. Site Specific Relative Humidity Adjustment Factors, Best and
Worst (Default) Estimates of Natural Background Visibility Conditions

MANE-VU Mandatory Federal Class I Area	F(RH)	Best Visibility

(dv)	Worst Visibility

(dv)

Maine



	    Acadia National Park	3.34	3.77	11.45

    Moosehorn Wilderness	3.15	3.68	11.36

    Roosevelt Campobello International Park, New Brunswick	3.16	3.68
11.37

New Hampshire



	    Great Gulf Wilderness	3.01	3.63	11.30

    Presidential Range – Dry River Wilderness	3.02	3.65	11.30

New Jersey



	    Brigantine Wilderness	2.97	3.60	11.28

Vermont



	    Lye Brook Wilderness	2.91	3.57	11.25



Nearby Mandatory Federal Class I Area



	Virginia



	    James River Face Wilderness	2.93	3.56	11.26

    Shenandoah National Park	2.95	3.57	11.27

West Virginia



	    Dolly Sods Wilderness	3.06	3.64	11.32

    Otter Creek Wilderness	3.06	3.65	11.32



Table 3. Default and Refined Estimates of the Twenty Percent Worst
Natural Background Visibility Conditions at MANE-VU and Nearby Sites.
Default values are provided for comparison, estimates labeled
“[OMC]=[OC]*1.8” are calculated using 1.8 as the organic multiplier,
“P90=HI+1.40*sd” values are calculated using the 92nd percentile
instead of the 90th percentile of the visibility distribution,
“w/seasalt” values show the effect of adding the measured value of
sea salt mass at coastal sites, and “Rayleigh 12Mm-1” values show
the effect of using alternate Rayleigh scattering at coastal sites.

Assumption tested at MANE-VU Mandatory Federal Class I Area	Default
Visibility

dv	[OMC]= [OC]*1.8

dv	P90=HI +1.40*sd

dv	w/              sea salt

dv	Rayleigh 12 Mm-1

dv

Maine





	Acadia National Park	11.45	12.17	11.81	12.87	12.34

Moosehorn Wilderness	11.36	12.09	11.72	12.88 	12.26

Roosevelt Campobello International Park, New Brunswick	11.37	12.09	11.73
12.88	12.27



New Hampshire





	Great Gulf Wilderness	11.30	12.03	11.66	 

	Presidential Range – Dry River Wilderness	11.30	12.03	11.66	 

	New Jersey





	Brigantine Wilderness	11.28	12.01	11.64	13.40	12.19

Vermont





	Lye Brook Wilderness	11.25	11.99	11.61	 

	Nearby Mandatory Federal Class I Areas



 

	Virginia





	James River Face Wilderness	11.26	11.99	11.62	 

	Shenandoah National Park	11.27	12.00	11.63	 

	West Virginia





	Dolly Sods Wilderness	11.32	12.05	11.68	 

	Otter Creek Wilderness	11.32	12.05	11.68	 

	

Table 4. Estimated baseline visibility conditions,† Uniform Rates of
Progress (ROP) to be considered for first implementation period, and
Effect of Natural Background Refinements on ROP at MANE-VU and Nearby
Sites.  “1.8*OC” values are the percent change in uniform progress
(relative to the default ROP) resulting from the substitution of the 1.8
carbon multiplier, “1.4*sd” values are the percent change in uniform
progress when the 92nd percentiles are used to represent the average of
the worst twenty percent visibility conditions, “sea salt” values
are the percent change in uniform progress when extinction due to
measured sea salt at coastal sites is included and “Rayleigh” values
are the percent change in uniform progress when 12Mm-1 of Rayleigh
extinction is used at sea-level sites.

MANE-VU Mandatory Federal Class I Area	Baseline Visibility

dv	Default ROP 

dv/14 yrs	1.8*[OC] 

%change	1.40*sd 

%change	sea salt 

%change	Rayleigh 

¬

®

æ

(

¯

K

kd÷

  hÅs

 hÅs

Nearby Mandatory Federal Class I Area







Virginia







James River Face Wildernessп	28.41	4.00	-2.4%	-2.1%



Shenandoah National Park‡	27.55	3.80	-2.8%	-2.2%



West Virginia







Dolly Sods Wilderness	27.72	3.83	-3.1%	-2.2%



Otter Creek Wilderness	27.72	3.83	-3.1%	-2.2%



† Note that EPA guidance requires at least 3 complete years out of 5
to calculate baseline conditions.  Routine year-round monitoring did not
begin at Camp Dodge (IMPROVE site for Great Gulf/Presidential Range)
until June 2000 so estimates of baseline conditions (and thus a uniform
rate of progress) will not be possible for these sites until data are
available through June 2003.

‡ Only 4 years of data was used in the calculation of estimated
baseline conditions and uniform rates of progress at these sites since
1998 did not meet completeness criteria at these sites.

eteness criteria at this site.

Figure 1.  Relative contribution to overall uncertainty of natural
background visibility extinction at MANE-VU Class I Areas.  Several
potential refinements that have been proposed are highlighted in red in
the legend.

Figure 2. Range of relative contribution to overall uncertainty in
natural background deciview estimates (high and low extremes derived
using extreme values of extinction range)

