_____________________

Ex Parte Meeting Summary:  Ozone NAAQS

Date:  December 14, 2007, 12:00 – 1:00 pm. 

Attendees: 

Visitor: 

Dr. Anne E. Smith

Vice President

CRA International

1201 F Street, N.W., Suite 700

Washington, DC  20004-1204

202-662-3800

asmith@crai.com

EPA attendees: 

George Gray, ORD

Pamela Williams, ORD

Summary of discussion:  

Dr. Smith discussed several of the technical aspects of the risk
assessment that EPA performed as part of the proposed NAAQS rule for
ozone.  This discussion was within the context of the comments that she
had submitted during the public comment period (which she distributed
during the meeting).  

In particular, she discussed the differences between EPA’s methodology
for estimating mortality risks above policy relevant background using
2002 and 2004 air quality data (Figure 5-15 from EPA Staff Paper) and
estimates that she has derived of mortality risk when the maximum 8-hour
average exceeds 70 ppb using an average of 2002 and 2004 air quality
data (Figure 4 in her comments).  Based on her re-analysis of the data,
she concluded that the percent of total non-accidental mortality under
the current standard that is attributable to days when the maximum
8-hour average ozone rises above 70 ppb is relatively small, so that the
current standard already provides significant protection against ozone
below the standard down to 70 ppb.  She also commented that it is more
appropriate to use the posterior credible intervals from the Bell et al.
study from each city, rather than the national average estimate, when
performing these calculations.

Additionally, Dr. Smith discussed some of the limitations of EPA’s
analysis for estimating the Policy Relevant Background (PRB).  For
example, she indicated that the GEOS-CHEM model that was used by EPA is
a global model with a very large grid that does not account for the
variability in ozone concentrations across sites, which results in an
underestimate of the average and peak concentrations.  She also
indicated that EPA’s approach does not include potential sources of
ozone from Canada or Mexico.  Aside from some technical concerns she has
with the new approach used by EPA (versus that used in 1987), she also
indicated that EPA’s risk estimates are very sensitive to the PRB
assumption (Appendix C in her comments), and that the finding of
increased risk over time is just due to changing assumptions.  She also
presented an analysis she performed of the percent reduction in EPA’s
risk estimates when computed using PRB assumptions based on Trinidad
head monitored ozone levels (she indicated that this site is viewed as
not having any US contribution) rather than based on GEOS-CHEM model
results (Figure 5 in her comments).  Based on the sensitivity of the
risk estimates to the PRB assumption, she indicated that the tightening
the current standard might be very difficult or impossible to attain in
some areas.  

