Bryan Hubbell/RTP/USEPA/US 

03/06/2008 10:00 AM

	

To

Tricia Crabtree/RTP/USEPA/US@EPA

cc

Subject

For inclusion in the both the Ozone NAAQS and Ozone NAAQS RIA dockets









Thanks!

Dr. Bryan J. Hubbell

Senior Advisor for Science and Policy Analysis

Health and Environmental Impacts Division

U.S. EPA/OAQPS (C504-02)

RTP, NC 27711

(919) 541-0621

----- Forwarded by Bryan Hubbell/RTP/USEPA/US on 03/06/2008 09:57 AM
-----

Bryan Hubbell/RTP/USEPA/US 

03/05/2008 01:11 PM

	

To

"Schwab, Margo" <Margo_Schwab@omb.eop.gov>

cc

Harvey Richmond/RTP/USEPA/US@EPA, Karen Martin/RTP/USEPA/US@EPA, Lydia
Wegman@EPA

Subject

Paragraph on ozone averaging times in the RIA vs Risk analysis









Margo-

As I indicated in my voice mail, here is the paragraph that I have
drafted for inclusion in the ozone NAAQS RIA.  Let me know if you have
any edits.  I will also make sure that we have conforming language on
the overall issue based on the Locomotive Marine RIA writeup.  As I
mentioned in my voice mail, we are not planning on including this
writeup as part of the preamble or other rule documents, only as part of
the RIA.  Thanks,

Bryan

"In EPA's risk analysis for the ozone NAAQS rule, mortality risks were
estimated for 8 urban areas based on application of city-specific effect
estimates derived from single city studies and from the Bell et al
(2004) and Huang et al (2005) multi-city studies.  These effect
estimates were based on 24-hour average daily ozone concentrations. 
While it may have been preferable to use shorter averaging times,
conversions from daily averages to shorter averaging times was not
appropriate due to the lack of city-specific conversion factors.  In our
benefits analysis for the ozone NAAQS, we applied national effect
estimates based on the pooled multi-city results reported in Bell et al
(2004) and the three meta-analysis studies.  Bell et al (2004), Bell et
al (2005), Levy et al (2005), and Ito et al (2005) all provide national
conversion ratios between daily average and 8-hour and 1-hour maxima,
based on national data.  However, these conversions were not specific to
the ozone "warm" season which was the period used in the health risk
assessment.    As such we were able to convert the national C-R function
parameters from daily average to 8-hour average, albeit with the
introduction of additional uncertainty due to the use of effect
estimates based on a mixture of warm season and all year data in the
epidemiological studies.  Given the heterogeneity in ratios of daily
average to 8-hour and 1-hour maxima that exists between cities, it would
be inappropriate to use national conversion ratios to adjust C-R
functions for individual cities.  However, it should be noted that the
mortality risks estimated in the ozone NAAQS risk analysis will be
insensitive to the specification of the averaging time, because unlike
the air quality modeling of specific emissions reductions scenarios used
in the RIA,  the air quality rollback technique used to simulate
attainment with alternative NAAQS in the risk analysis does not result
in increases in ozone during nighttime hours.  As such, specification of
the effect estimate in terms of 24-hour averages rather than 8-hour
maximum is simply a scale translation, and will not result in large
differences in risk estimates."  

Dr. Bryan J. Hubbell

Senior Advisor for Science and Policy Analysis

Health and Environmental Impacts Division

U.S. EPA/OAQPS (C504-02)

RTP, NC 27711

(919) 541-0621

