AQMS REVIEW OF CAA 812
PROSPECTIVE STUDY OF COSTS AND
BENEFITS: HEALTH AND ECOLOGICAL
EFFECTS INITIAL STUDIES
SEPTEMBERS, 1998
-------
Septembers, 1998
EPA-SAB-COUNCIL-ADV-98-002
Honorable Carol M. Browner
Administrator
U.S. Environmental Protection Agency
401 M Street, SW
Washington, DC 20460
Dear Ms. Browner:
Subject: Prospective Study I: Advisory by the Air Quality Models Subcommittee
(AQMS) on the Air Quality Models and Emissions Estimates Initial Studies
The Air Quality Modeling Subcommittee (AQMS) of the Science Advisory
Board's (SAB) Advisory Council on Clean Air Compliance Analysis (ACCACA, or "the
Council") met on January 22-23, 1998 in Washington DC, to review materials related
to the 812 Prospective Study Air Quality Modeling and receive briefings from EPA staff
and consultants.
The general charge for the meeting addresed whether a) the input data used for
each component of the analysis, and b) the models and methodologies employed are
sufficiently valid and reliable for the intended purpose; and if the answer is negative to
either a) or b) then c) what specific alternative assumptions, data and or methodologies
would the Council recommend the Agency consider using for the first prospective
analysis. Although the above charge defines the general scope of the advice re-
quested from the Council, specific modeling-oriented questions and issues were
identified for individual analytical components in the briefings and discussions engaged
with the AQMS at the public meeting.
This Advisory report is the product of that meeting, and summarizes the AQMS'
advice to EPA regarding the prospective study design, implementation and future
planning. Our detailed comments are included in the enclosed report.
Overall, we concluded that the prospective study team's approach to the air
quality assessment is comprehensive, but needs to be described more clearly and
concisely. The strategy of using model results and observations is an appropriate,
sound approach for the current prospective study. We offer the attached Figure 1 in
-------
this advisory to the prospective study team as an illustration of an approach that might
be helpful.
In looking ahead to the future prospective studies, we suggest that the prospec-
tive study team consider use of the more comprehensive modeling platform provided by
EPA's Models-3, which has recently become available for use. This platform would
make it possible to apply one method of analysis throughout the U.S.. We also suggest
use of more advanced interpolation schemes and we have provided a number of
references and contacts for the staff. Finally, we strongly advise development and use
of a more flexible and user-friendly emissions modeling system that provides the ability
to better diagnose data problems and easily examine multiple scenarios. The Models-3
emissions work provides a good starting point for further development of improved
user-friendly emissions processing.
We realize that the development of new emissions management modeling
systems is an EPA-wide need and goes beyond the needs of the prospective studies.
For that reason, we are suggesting that EPA as a whole consider a thorough review of
how emissions are currently being dealt with in the Agency and recommend develop-
ment of a detailed strategy for how the process could be improved in the near and long
term. All of us in the AQMS would be willing to serve on the review committee since
we all feel very strongly about the very key need to develop a nation-wide flexible,
efficient and reliable emissions management modeling system. We also encourage the
prospective study team to coordinate with other related assessment activities and
reviews currently underway at EPA.
We encourage the prospective study team to explore links between air quality
and climate variability and possibly climate change in future prospective studies. The
current prospective analysis assumes that meteorology remains constant for the
assessment years and no climate variability is considered. It will be important to
consider how best to take potential climate variations into account when doing the
prospective modeling in the future. It may also be important to examine emissions
change strategies based on climate policy. The study team included one of these
scenarios for our consideration as a possible supplemental analysis to the current
prospective study. The possibility of doing an analysis including climate variability
during the next iteration of the study should be given serious consideration.
The Subcommittee's most serious concern with the current prospective study
involves the predictions for particulate matter (both PM10 and PM25) in future years.
Over recent years, a downward trend has been observed in the concentration of
airborne particulate matter. The PM criteria document shows downward trends for all
regions of the country for PM10 during the period of 1988-1994. More recently, in a
paper by Darlington et al. (1997) (see References in the enclosed report documents in
more detail the downward trend in PM10. The EPA's Air Quality Criteria for Particulate
-------
Matter, (see References in the enclosed report, U.S. EPA/ORD, 1996 ) shows that
regional reductions in PM10 from 1988 to 1994 at EPA trend sites range from 17 percent
to 33 percent. The 1997 paper shows that PM10 reductions follow similar patterns at
rural, suburban, and urban locations. Although the monitoring sites are primarily in
urban areas, the trend also seems to be substantiated in non-urban areas as well.
In contrast, the current prospective study pre-CAAA90 (Clean Air Act Amend-
ments of 1990) scenario results shows an average increase in PM and the
post-CAAA90 scenario shows a decrease significantly less than the decrease already
observed during the initial 5 years of the prospective study I analysis period. Given the
very large effect of PM on projected health outcomes, this obvious discrepancy
between the documented trend in ambient PM concentrations and the simulated
outcomes would raise serious doubts in the minds of many readers of the report and
has the potential to undermine the credibility of the entire prospective study effort.
The AQMS suggests further review of agricultural tilling, road and other sources
of dust, livestock management, and industrial point source emissions. In addition to
changes in directly emitted PM, there have been substantial decreases in emissions of
man-made precursors of particulate matter formed in the atmosphere and the extent to
which these decreases have been accounted for needs to be checked. The AQMS
also notes that the recent EPA sponsored Grand Canyon Visibility Transport
Commission Assessment of Scenarios to the year 2040 (Western Governors
Association Reports, 1996. See References in the enclosed report, Middleton and
associates, 1996) provides an opportunity for cross-checking emission projection
assumptions and techniques. We strongly advise that this trends problem be resolved
first before conducting any new scenario runs in the current prospective study.
We appreciate the diligence of the prospective study team on this difficult and
timely assessment. Given our particularly strong concerns over the PM emission
trends, we suggest that our subcommittee provide an informal consultation on new PM
emission projections as soon as they become available. We look forward to your
response, particularly to the main points outlined in this advisory letter to you, and to
continued interaction with your professional staff.
Sincerely,
Dr. Maureen L. Cropper, Chair
Council
enclosure
Dr. Paulette Middleton, Chair
Air Quality Models Subcommittee
Council
-------
NOTICE
This advisory report has been written as a part of the activities of the Science
Advisory Board, a public advisory group providing extramural scientific information and
advice to the Administrator and other officials of the Environmental protection Agency.
The Board is structured to provide a balanced, expert assessment of scientific matters
related to problems facing the Agency. This advisory report has not been reviewed
for approval by the Agency; hence, the comments of this advisory do not necessarily
represent the views and policies of the Environmental Protection Agency or of other
Federal agencies. Any mention of trade names or commercial products does not
constitute endorsement of recommendation for use.
-------
ABSTRACT
The Science Advisory Board's Air Quality Models Subcommittee (AQMS) of the
Council, has reviewed precursors to the first Prospective Study: Report to Congress.
Overall, the AQMS concludes that the strategy of using model results and observations
is found to be an appropriate, sound approach for the current prospective study, but
needs to be described more clearly and concisely.
For future prospective studies, the AQMS suggests that the study team consider
use of the more comprehensive modeling platform of EPA's Models-3 platform which
would make it possible to have a more consistent analysis of areas throughout the
U.S.. In addition, the AQMS also suggests use of more advanced interpolation
schemes. Finally, the AQMS strongly advises development and use of a more flexible
and user-friendly emissions modeling system that provides the ability to better
diagnose data problems and more easily examine multiple scenarios.
The Subcommittee's most serious concern involves the predictions for
particulate matter (both PM10 and PM25). Recently, a downward trend has been
observed in the concentration of airborne particulate matter. In contrast, the current
prospective study pre-CAAA90 scenario results shows an average increase in PM and
the post-CAAA90 scenario shows a decrease significantly less than the decrease
already observed during the initial 5 years of the prospective study I analysis period. .
The AQMS suggests several strategies that might help address this issue and strongly
advises that this discrepancy in predicted and observed trends be understood or
resolved first before conducting any new scenario runs in the current prospective study.
Keywords: Clean Air Act, Air Quality Models, Emissions Estimates, Prospective Study
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD
AIR QUALITY MODELS SUBCOMMITTEE OF THE COUNCIL
CHAIR
Dr. Paulette Middleton, Deputy Director, RAND Center for Environmental Sciences &
Policy, Boulder, CO
MEMBERS AND CONSULTANTS
Dr. Philip Hopke, Professor, Department of Chemistry, Clarkson University, ,
Pottsdam, NY
Dr. Harvey Jeffries, Professor, Department of Environmental Sciences & Engineering,
University of North Carolina, Chapel Hill, NC
Dr. Timothy V. Larson, Professor, Environmental Engineering & Science, Department
of Civil Engineering, University of Washington, Seattle, WA
Dr. Peter K. Mueller, Technical Manager, EPRI, Palo Alto, CA
Dr. James H. Price, Jr., Senior Scientist, Texas Natural Resource Conservation
Commission, AU.S.tin, TX
Dr. George T. Wolff, Principal Scientist, Environmental and Energy Staff, General
Motors Corp Detroit, Ml (Past Chair, Clean Air Scientific Advisory Committee & Past
Chair of the AQMS)
SCIENCE ADVISORY BOARD STAFF
Dr. K. Jack Kooyoomjian, Designated Federal Officer, Science Advisory Board
(1400), U.S. Environmental Protection Agency, 401 M Street, SW, Washington, DC
20460
Mrs. Diana L. Pozun, Management Assistant, Science Advisory Board (1400), U.S.
Environmental Protection Agency, 401 M Street, SW, Washington, DC 20460
-------
U. S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD (SAB)
ADVISORY COUNCIL ON CLEAN AIR COMPLIANCE ANALYSIS
(THE COUNCIL)
CHAIR:
Dr. Maureen L. Cropper, Principal Economist, The World Bank, Washington, DC
MEMBERS:
Dr. Ronald G. Cummings, Professor of Economics and Noah Langdale, Jr. Professor of
Environmental Policy, Policy Research Center, Georgia State University, Atlanta, GA
Dr. A. Myrick Freeman, Professor, Department of Economics, Bowdoin College, Brunswick,
ME (Also Vice-Chair of the Health and Ecological Effects Subcommittee, HEES of the
Council)
Dr. Lawrence H. Goulder, Associate Professor, Department of Economics & Institute for
International Studies, Stanford University, Stanford, CA
Dr. Jane V. Hall, Professor of Economics, Department of Economics and Institute for
Economic and Environmental Studies, California State University, Fullerton, CA
Dr. Paul Lioy, Deputy Director-EOSHI & Director Exposure Measurement & Assessment
Division, Environmental & Occupational Health Sciences Institute, Robert Wood Johnson
School of Medicine, Piscataway, NJ (Also Chair of the Health and Ecological Effects
Subcommittee, HEES of the Council)
Dr. Paulette Middleton, Deputy Director, RAND Center for Environmental Sciences & Policy,
Boulder, CO (Also Chair of the Air Quality Models Subcommittee, AQMS of the Council)
Dr. Richard Schmalensee, Deputy Dean, Sloan School of Management, Massachusetts
Institute of Technology, Cambridge, MA
Dr. Thomas H. Tietenberg, Professor, Dept. of Economics, Colby College, Waterville, ME
CONSULTANTS:
Dr. Alan J. Krupnick, Senior Fellow, Resources for the Future, Washington, DC
SAB COMMITTEE LIAISON:
Dr. William H. Smith, Professor of Forest Biology, School of Forestry & Environmental
Studies, Yale University, New Haven, CT (Liaison from the Environmental Processes and
Effects Committee)
SCIENCE ADVISORY BOARD STAFF:
Dr. K. Jack Kooyoomjian, Designated Federal Officer, Science Advisory Board (1400), U.S.
Environmental Protection Agency, Washington, DC 20460
IV
-------
Mrs. Diana L. Pozun, Management Assistant, Science Advisory Board (1400), U.S.
Environmental Protection Agency, Washington, DC 20460
-------
TABLE OF CONTENTS
1. EXECUTIVE SUMMARY 1
1.1 Overview 1
1.2 Modeling Strategy 812 - Prospective One 2
1.3 Modeling Strategy 812 -Prospective-Two and Beyond 2
1.4 Trends Concerns 3
2. INTRODUCTION 3
2.1 Background 3
2.2 Charge 3
3. MODELING STRATEGY - 812 PROSPECTIVE ONE 5
3.1 Overall Strategy Presentation
3.2 Comparing Retrospective and Prospective Studies 5
3.3 Scope of Study 6
3.4 Approach 6
3.5 Rationale for and Choice of Models 6
3.6 Model Evaluation 7
3.7 Use of Model Results 7
3.8 Clarification of Scenarios 10
3.9 Background Aerosols 10
3.10 Presentation of Results 10
3.11 Visibility Assessment 10
3.12 Local Analysis for Western Sites 10
3.13 Description of Secondary Organic Aerosols 11
4. MODELING STRATEGY 812-PROSPECTIVE-TWO AND BEYOND 12
4.1 Consider U.S.e of MODELS-3 Platform 12
4.2 Use of Advanced Spatial and Temporal Interpolation Schemes 12
4.3 Consider More Flexible Emissions Management Modeling 13
4.4 Consider Climate Variability 13
4.5 Coordinate with Other Agency-Wide Modeling Reviews 13
4.6 Reconciling Emission and Concentration Trends 14
4.6.1 The Concern 14
4.6.2 Possible Reasons for the Discrepancy 15
4.6.3 Possible Steps 16
4.7 Overall Quality of Emissions Inventories 17
4.8 Additional Scenario Analysis 17
5. SUMMARY 19
REFERENCES CITED R-1
VI
-------
TABLE OF CONTENTS: CONTINUED:
APPENDIX A - Modeling Draft Review Materials A-1
APPENDIX B - Additional Information on Recommendation to
Consider Use of EPA's Urban and Regional Community Air Quality
Modeling System B-1
APPENDIX C - Glossary of Terms and Acronyms C-1
VII
-------
1. EXECUTIVE SUMMARY
1.1 Overview
The Air Quality Modeling Subcommittee (AQMS) of the Science Advisory
Board's (SAB) Advisory Council on Clean Air Compliance Analysis ("Council") met on
January 22-23 in Washington DC, to review materials related to the 812 Prospective
Study Air Quality Modeling and receive briefings from EPA staff and consultants. This
advisory report outlines the AQMS' advice to EPA regarding the prospective study
design, implementation and future planning as developed at the meeting The AQMS
cites EPA Documents by referring to their Document Number as described in the
January 2, 1998 memorandum "Review Materials and Charge for SAB Council AQMS
Meeting to Discuss Section 812 Prospective Study Air Quality Modeling". This is
document 175 cited as [175] and included a document "Section 812 Prospective Study
Publicly Available Documents" [100].
Overall, the AQMS concludes that the prospective study team's approach to the
air quality assessment is comprehensive, but needs to be described more clearly and
concisely. The strategy of using model results and observations was found to be an
appropriate, sound approach for the current prospective study.
We approached this review by examining the validity of the study assumptions
and clarity of presentation. As a result of our review and discussions we are providing
what we believe are constructive and feasible suggestions for improvements within the
current context and constraints of the current prospective study, as well as
improvements in future design for the subsequent prospective studies. Our comments
focus on the modeling approach, along with detailed suggestions for improvement in
the current and future prospective studies; recurrent difficulties with air quality trends
and emissions estimates that still need resolution; and selection of emission change
scenarios that can help address the discrepancies observed and predicted
concentration trends. Specific findings are:
a) Modeling Strategy 812 - Prospective One: Regarding the current
prospective study (referred hereafter as prospective I or the current
prospective study), the AQMS suggests clarification be provided on the
following topics: comparison of the prospective and retrospective studies;
overall prospective strategy and scope; choice, evaluation and use of
models; background aerosols; visibility estimates; treatment of organic
aerosol processes; treatment of western cities; and characterizing
scenarios.
b) Modeling Strategy 812 -Prospective-Two and Beyond: In looking
ahead to the future prospective studies, the AQMS suggests that the
1
-------
prospective study team consider use of EPA's more comprehensive
Models-3 modeling platform which would make it possible to apply one
method of analysis throughout the U.S.. In addition, the AQMS also
suggested use of more advanced interpolation schemes. Finally, the
AQMS strongly advised development and use of a more flexible and
user-friendly emissions modeling system that provides the ability to better
diagnose data problems and easily examine multiple scenarios.
c) Trends Concerns: The Subcommittee's most serious concern with the
current prospective study involves the predictions for particulate matter
(both PM10 and PM25) in future years. Recently, a downward trend has
been observed in the concentration of airborne particulate matter. In
contrast, the current prospective study pre-CAAA90 scenario results
shows an average increase in PM and the post-CAAA90 scenario shows
a decrease significantly less than the decrease already observed during
the initial five years of the prospective study I analysis period. The AQMS
suggests further review of primary particle emissions from agricultural
tilling, road and other sources of dust, and industrial point source
emissions. In addition the AQMS suggests reexamining precursor gas
emission estimates and cross-checking projections with previous studies.
The AQMS strongly advises that this trends problem be resolved first
before conducting any new scenario runs in the current prospective study.
-------
2. INTRODUCTION
2.1 Background
The Air Quality Models Subcommittee (AQMS) of the Council conducted a
review meeting on January 22 and 23, 1998 to review draft documents pertaining to the
emissions modeling assumptions, methodology, results and documentation
components of the Clean Air Act (CAA) prospective study, and provide advice to the
Council to transmit to the Administrator regarding the reasonableness, technical merits,
and appropriate interpretations of the modeling results at this stage of the prospective
study analysis. Earlier advice on the prospective study modeling assumptions,
methodology, results and documentation was provided by the AQMS through its
Council ( U.S. EPA/SAB, 1997). This report has been produced as a result of the
Subcommittee's deliberations at, and following, the public meeting.
2.2 Charge
The charge to the AQMS and the Council is to review the draft materials and
provide advice to the Agency pursuant to general charge questions, consistent with the
review responsibilities of the Council, as defined in Section 812 of the Clean Air Act
Amendments of 1990 (CAAA90). Specifically, subsection (g) of the CAA Section 312,
as amended by Section 812 of the amendments states the following:
"(g) The Council shall - (1) review the data to be used for any analysis required
under this section and make recommendations to the Administrator on the use of
such data, (2) review the methodology used to analyze such data and make
recommendations to the Administrator on the use of such methodology; and (3)
prior to issuance of a report required under subsection (d) or (e), review the
findings of such report, and make recommendations to the Administrator
concerning the validity and utility of such findings."
The general charge questions posed by the Agency to the Council's AQMS
include the following:
a) Are the input data used for each component of the analysis sufficiently
valid and reliable for the intended analytical purpose?
b) Are the models, and the methodologies they employ, used for each
component of the analysis sufficiently valid and reliable for the intended
analytical purpose?, and
c) If the answers to either of the two questions above is negative, what
specific alternative assumptions, data or methodologies does the Council
recommend the Agency consider using for the first prospective analysis?
-------
Although the above charge defines the general scope of the advice requested from the
Council, specific modeling-oriented questions and issues were identified by the Ageny
staff and AQMS M/C (members and consultants) for individual analytical components
during the public meeting.
-------
3. MODELING STRATEGY - 812 PROSPECTIVE ONE
3.1 Overall Strategy Presentation
We strongly endorse the four study goals provided in "Analytical Strategy For
the First Section 812 Prospective Study." However, the goals could be more clearly
stated and important terms could be more clearly defined. We suggest that a summary
of terms be developed. The document should be updated as needed and should be
supplied to the Council, the Council Subcommittees, and others as appropriate, during
the prospective study review process. In particular, the trend estimates and how
estimates differ from model predictions and observations needs to be described. The
use of terms like uncertainty analysis and sensitivity studies as well as monetized and
incremental benefits and costs also needs to be clearly outlined in the context of the
prospective study.
3.2 Comparing Retrospective and Prospective Studies
We suggest that the comparison be presented but the current presentation be
replaced with a comparison that focuses on the pollutants included and the type of
treatment needed to meet intended uses of model results. The substances included in
the prospective study include both primary (i.e., directly emitted) and secondary (i.e.,
formed by chemical and physical processes in the atmosphere) substances. Some
suggestions follow below.
The primary substances included in study are :
a) Sulfur Dioxide (S02) assumed linearly proportional to emissions
b) Nitroge Oxides (NOX) assumed linearly proportional to emissions
c) Carbon Monoxide (CO) assumed linearly proportional to emissions
d) Particulate Matter assumed linearly proportional to emissions
(PM)-primary
The secondary pollutants included are:
a) Ozone (03) non-linearly related to NOX and VOC via photochemistry
b) Acid Deposition non-linearly related to NOX, S02, ammonia (NH3) emissions,
Ozone, other oxidants via photochemistry, aqueous
chemistry
c) PM-secondary related to NOx, S02, NH3, volatile organic compounds
(VOC) emissions, free radical levels in the NOX, VOC, 03
system; particle composition
d) visibility related to composition and size of PM
-------
3.3 Scope of Study
We suggest that EPA's future descriptions of the scope of the prospective study
state clearly that the CAA applies to the whole nation, while the prospective study will
include only the contiguous 48-states. Because the implementation period and time for
effectiveness periods of many parts of the CAA and CAAA90 are so long, the temporal
domain of the Study is 20 years. Rather than simulate all years, three target years: the
beginning of period, 1990; the mid-point of period, 2000; and the end of period, 2010
are selected using similar meteorology for each, thereby for the initial analysis,
sidestepping the complexity of including realistic interannual climate variability.
3.4 Approach
The scope of the prospective study has determined its technical design. To
further justify the study approach, the team should discuss the fact that, at present,
there are no operational air quality models that simulate the entire U.S. for all
substances of interest nor even for any one substance. The approach to developing
annual distributions, using a variety of models for different areas and different
pollutants, needs to be more carefully outlined.
3.5 Rationale for and Choice of Models
The project study team needs to provide detailed rationale underlying the choice
of the specific models used in the current prospective study. The project study team
also needs to comment on the quality of the model's formulation (especially the degree
of completeness in representing processes believed to be important, as well as the
degree of generalization, deletion, and distortion in the representation that are
included) relative to the intended uses and range of the model's predictions.
Similarly, for the inputs to the models, the prospective study project team needs
to specify carefully the quality of input data. This is important for two reasons: a) given
the relatively large deviation from observations of several of the models' predictions for
the base case, the prospective study project team needs to make a judgment as to
potential sources of error: flaws in the model itself (or in its particular configuration,
e.g., too few vertical layers); or inaccurate inputs, such as incorrect mixing heights; b)
given that the model's predicted values may be inaccurate, the project study team
needs to demonstrate that differences between two scenarios are acceptable estimates
of the consequences of the differences between the inputs to the models in the two
scenarios.
The current prospective study focuses on incremental concentration differences
between two basic scenarios. The prospective I study asserts that the differential
responses of the model are essentially independent of the concentration levels in the
-------
region of the models' predictions. This should be explicitly stated and some evidence
should be offered to support the assertion.
3.6 Model Evaluation
Very little is said about model performance evaluation in the current prospective
study documents. In the Document 137 description given in [100, p. 5] attached to
document 175, the Review Materials and Charges memo, for example, the description
reads "Report documenting fundamental elements of the criteria pollutant air quality
modeling, including models, methodologies, assumptions, and draft results" with no
mention of "model performance evaluation" and which was not included in document
137a. The only information in 137a is given is in Fig 4-2 and 4-3, and the statement on
page 4-14, "Please note that for both the 1993 and 1995 simulation periods, the
maximum simulated ozone concentration was greater than observed, especially for the
high ozone days."
The Ozone Transport Assessment Group (OTAG) simulations were not intended
to simulate attainment demonstrations, but instead were designed to estimate the
effects of regional ozone transport for the purposes of eventually helping with boundary
conditions for more localized attainment modeling. We strongly suggest that Figs 4-2
and 4-3 of the subject documents be eliminated as they provide virtually no useful
information. They should be replaced by the standard scatter diagram of simulations
vs. observations, which will make it much clearer the extent to which the whole set of
simulations were consistent with all of the observations. Reasons for the high
simulated values need to be offered and arguments advanced as to why differences
among different emission scenarios are sufficiently reliable "estimates" of changes in
air quality for this model.
Another advantage of using these scatter diagrams is that they give some idea
of the ratio of the modeled ozone to observed ozone that can be compared with the
ratio of pre-CAAA90 to post-CAAA90 values that will be used later to scale the ambient
observations. Similarly Fig 4-8 and 4-9 provide little information about the quality of the
modeling and need to be replaced with scatter diagrams of simulations vs. observations
for these two attainment style cases. Fig 4-10 and 4-11 are also not very useful.
3.7 Use of Model Results
The prospective study project team should more clearly articulate the overall
concept of the modeling study. In thinking about this we have produced a schematic
which the team may want to use. Figure 1 illustrates how model results and
observations are used to produce the frequency distributions for the impact analysis.
Glossed over in the figure is the issue of "aggregation" of various model runs to
produce a single set of "ratios" of future-to-base case. The figure shows the five model
-------
runs at the top. The middle shows the four sets of "ratios" of model results in space.
The bottom shows the frequency distributions of pollutant monitor concentrations and
the space-dependent scaling of these by the ratios of the model predictions. The
Figure glosses over the issue of "aggregation" of various model runs to produce a
single set of "ratios" of future to base case.
8
-------
Figure 1 - Concentration and Concentration Distributions of Air Quality Model Predictions and Air Quality
Monitor Observations. [ NOTE: Figure illustrates how model results and observations are used to
produce the frequency distributions for the impact analysis. F. Figure illustrates five model runs at the
top; four sets of "ratios" of model results in space in the middle; and, frequency distributions of pollutant
monitor concentrations and the space-dependent scaling of these by the ratios of the model predictions
on the bottom.!
-------
3.8 Clarification of Scenarios
Communication regarding the scenarios would benefit from a more explicit
description of what was included and excluded in the pre-CAAA90 and post-CAAA90
scenarios. For example, in the EPA presentation to the Subcommittee, the PM-primary
inventory in the post-CAAA90 scenario is indicative of this problem. We are still
unclear on the issue of what is meant by freezing at 1990, but having growth overtake a
downward trend.
3.9 Background Aerosols
The assumed levels of background aerosols (those directly emitted along with
those that are formed in the air) were not clearly presented anywhere in the EPA
documents. Background concentrations for several components by aerosol species
need to be provided for our review for each major part of the country. Assumptions
about background (i.e., aerosols transported into an assessment domain) need to be
described and justified, because background determines the effectiveness of the
projected emissions reductions within a domain.
3.10 Presentation of Results
The prospective study results presented in terms of ambient concentrations tend
to focus attention on the attainment issues. We suggest using differentials and to
avoid discussions regarding attainment, especially in light of the uncertainties still
surrounding the emissions projections (which we will discuss later in this report).
3.11 Visibility Assessment
The approach to estimating visibility from particulate concentrations needs to be
clearly outlined. Given the large differences in the contribution of different aerosol
components to light extinction, it is important to provide the assumptions used in the
study. The derivation of key assumptions concerning the specification of particle size
distribution, aerosol composition, individual extinction efficiencies for each aerosol
substance and the variation of efficiencies in space and time should be briefly
described with proper citations to the published literarure.
3.12 Local Analysis for Western Sites
Given the resource limitations and current readiness of SAQM (SARMAP Air
Quality Model) for aerosol application, we recommend that the detailed analysis of the
San Joaquin Valley done for the Retrospective Study: Report to Congress be omitted
from this current prospective study. In addition the treatment and descriptions of
Denver and Salt Lake City need to be carefully outlined, since these areas are quite
10
-------
different from Los Angeles and Phoenix with respect to altitude, terrain and related
factors.
3.13 Description of Secondary Organic Aerosols
The study does not adequately document the treatment of secondary organic
aerosols. Most importantly, the AQMS M/C could not find any information on how
secondary organic particles are related to VOC emissions. Given uncertainties
surrounding the relative importance of biogenic and anthropogenic emissions to
secondary organic particle formation, it is particularly important to describe carefully
how the relevant emissions are being translated into multi-phase organics.
11
-------
4 MODELING STRATEGY 812-PROSPECTIVE-TWO AND BEYOND
4.1 Consider Use of MODELS-3 Platform
EPA's Office of Research and Development (EPA/ORD) has released the first
version of its MODELS-3 framework this summer. MODELS-3 is intended to be a
community modeling system supporting not only air quality simulation, but also
simulations of processes involving lakes, rivers, estuaries and bays. A brief description
of the MODELS-3 framework is included in Appendix B of this report (U.S. EPA, 1998a
and1998b).
The current Community Air Quality Model (CAQM) included in the release will
permit prediction of all air quality state variables that were included in the Prospective I
study in a single model. That is, CO, NOx, VOC, 03, SOX, acid deposition by species,
PM10, PM25 with composition, and visibility. By the time the next prospective modeling
will be needed, it is expected that EPA/ORD will already have produced the
meteorological files and other inputs to simulate the continental U.S. with a 36-km
resolution. In addition, much of the Eastern U.S. will have been simulated with a 12-km
resolution, and many urban areas with a 4-km resolution. Furthermore, the 30 case
aggregation data set now used with RADM (Regional Acid Deposition Model) will have
been converted to MODELS-3 for the U.S. by then.
The MODELS-3 system is expected to have the capability to run many
simultaneously nested simulations occurring at different locations in the U.S., all
sharing a large scale meteorological and chemical data set. Thus, modelers in
different locations could be making consistent contributions to the prospective study at
the same time. Advancements in the science can be introduced into the modeling
framework without needing to replace the entire model; thus investments in input file
sets and in meteorological scenarios are to be protected while still providing advances
in the modeling science. Because of this flexibility, use of such a tool would greatly
simplify the prospective study, permitting more time to focus on the disaggregation of
benefits and costs requested by Congress rather than the initial estimation of the
U.S.-wide benefits and costs by pollutant as is the case in the present prospective
study.
4.2 Use of Advanced Spatial and Temporal Interpolation Schemes
There are a set of spatio-temporal interpolation schemes that have been
developed in recent years (Christakos etal., 1993 through 1998, Bogaert and
Christakos, 1997) that provide simultaneous interpolation in space and over time.
These interpolation schemes include space-time cross-terms that enhance the
accuracy of spatial interpolation. Furthermore, these techniques automatically include
an estimation error variance in space and time. The AQMS believes that these
12
-------
advanced schemes provide a good format for space-time interpolation and evaluation.
4.3 Consider More Flexible Emissions Management Modeling
Throughout our review and discussions, the topic of emission uncertainties and
emission scenario development continued to dominate. We appreciate the level of
frustration facing the prospective study team as it cannot easily turn around new
emissions modeling experiments to either test their assumptions regarding future
emissions or to try new scenarios. We strongly recommend that in the next round of
prospective analysis, the study team consider the development of a more flexible
approach to emissions modeling. As we envision this, such an approach would
compartmentalize each step in the emissions processing such that different
assumptions could be made and implemented easily in the system. The Grand Canyon
Visibility Transport Commission's Assessment of Scenarios (Middleton, 1996; Western
Governors Association Reports, 1966) used a system that had many of these desirable
components. Ongoing emissions work associated with the Models-3 effort also
provides a good starting point for improved user-friendly emissions processing.
We realize that development of new emissions management modeling systems
is a need that is EPA-wide and goes beyond the needs of the current and near-future
prospective studies. For that reason, we are suggesting that EPA consider an Agency-
wide thorough review of how emissions are currently being addressed , with a goal of
recommending a detailed strategy for how the process could be improved in the near
and long term. All of the AQMS' M/C would be willing to serve on the review
committee since we all feel very strongly about the imperative to develop a nation-wide
flexible, efficient and reliable emissions management modeling system.
4.4 Consider Climate Variability
The current prospective analysis assumes that meteorology remains constant for
the assessment years. No variability is considered. It will be important to consider how
best to take the potential climate variations into account when doing the prospective
modeling in the future. It may also be important to examine emissions change
strategies based on climate policy. The study team included one of these scenarios for
our consideration as a possible supplemental analysis for the current prospective
study. The possibility of doing an analysis, including climate variability, in the next
iteration should be given serious consideration.
4.5 Coordinate with Other Agency-Wide Modeling Reviews
We encourage the prospective study team to coordinate with other activities and
reviews. Given the resource and time constraints, the more synergism that can be
developed across Agency programs, the better.
13
-------
4.6 Reconciling Emission and Concentration Trends
The Subcommittee's most serious concern with the current prospective study
involves the predictions for particulate matter (both PM10and PM25) in future years.
Recently, a downward trend has been observed in the concentration of airborne
particulate matter, while the prospective study documents shows an increase.
4.6.1 The Concern
The PM criteria document shows downward trends for all regions of the country
for PM10 during the period of 1988-1994. More recently, an October 1997 paper by
Darlington, etal., 1997) documents in more detail the downward trend in PM10. The
EPA's Air Quality Criteria for Particulate Matter ( U.S. EPA/ORD, 1996) shows that
regional reductions in PM10 from 1988 to 1994 at EPA trend sites range from 17 percent
to 33 percent. The 1997 paper shows that PM10 reductions follow similar patterns at
rural, suburban, and urban locations. Although the monitoring sites are primarily in
urban areas, the trend also seems to be substantiated by data from non-urban areas.
In contrast, the prospective study pre-CAAA90 scenario results shows an
average increase in PM and the post-CAAA90 scenario shows a decrease significantly
less than the decrease already observed during the initial five years of the current
prospective study analysis period. Given the current attention to PM related to
associated health outcomes, this discrepancy between the well-known trend in ambient
PM concentrations and the simulated outcomes would raise serious doubts in the
minds of many readers of the report and has the potential to undermine the credibility
of the entire prospective study effort.
This concern was brought up in the September 9, 1997 letter from Dr. Maureen
Cropper and Dr. Paulette Middleton to EPA Administrator Carol Browner. The letter
transmitted the review by the Air Quality Models Subcommittee (AQMS) of the Advisory
Council on Clean Air Compliance Analysis (ACCACA) ("the Council") of the Clean Air
Act Amendments of 1990, Section 812 prospective study emissions modeling and
associated air quality modeling issues (U.S. EPA/SAB, 1997). The letter states:
"The particulate matter (PM) emission trends provided in the prospective study
increase regardless of assumptions of growth, while recent PM concentration
trends are apparently going down. This important discrepancy still needs to be
examined and explained. Subsequent discussions occurred in the Council's
public teleconference meetings of May 15 and June 30, 1997, which emphasized
the need for the Agency to have clear text discussions on PM trends in the
Prospective Study Report to Congress."
14
-------
The most complete response the AQMS has seen to our concerns is contained
in the June 13, 1997 memorandum "Information for SAB Council on Section 812
Prospective Study" ( U.S. EPA/OPAR, , 1997) documented the Agency's revaluation
of PM emissions projections. This memorandum acknowledged that the National Air
Quality and Emissions Trends Report, 1995 (U.S. EPA, 1995) included data that
showed a 22 percent decrease in the national average of annual mean PM
concentrations from 1988 to 1995. The response also provided an initial analysis to
identify problems with the emissions inputs into the models and several useful
corrections were made. However, at that time no model results were available and
there was no way to gauge the effect of these changes on the model outcomes. Now
that these results have been obtained, it is clear that further refinement is needed.
At the AQMS public meeting, it waspointed out to Agency staff that there has
been a 30-year downward trend in total suspended particulate matter data and that PM
fine aerosol concentrations in the East have been decreasing over the 1988 to 1995
period. The AQMS M/C also noted that the trends for PM10 levels are down in both
attainment and nonattainment areas. The draft final report Prospective Analysis of Air
Quality in the U.S.: Air Quality Modeling (ICF Kaiser/SAIC, 1998) shows the modeled
ambient concentration trends for PM10. In contrast, the year 2000 Post-CAAA90
scenario shows that PM10 would increase at over one-fourth of the PM10 monitoring
sites and that the mode would be a 6-to-8 percent decrease in PM10.
Both in the June 13, 1997 memo noted above, and at the January 22-23, 1998
AQMS public meeting, EPA staff pointed out that the current Section 812 prospective
study will model costs and benefits based only on the difference between the
Pre-CAAA90 and Post-CAAA90 scenarios. The AQMS understands this point, but
the actual differences could be larger or smaller if the trend simulation were more
accurate. It appears that the largest part of the costs and benefits calculated in
the Section 812 Prospective Study may come from controls of PM. Because of the
obvious importance of PM to the overall current and future prospective study findings,
the AQMS is concerned that the discrepancy between the modeled change and the
substantial PM10 decreases already achieved will damage the credibility of the
prospective study. Therefore the apparent discrepancy between the projections
and the actual trends up to 1995 remain an important issue.
4.6.2 Possible Reasons for the Discrepancy
The AQMS suggests further review of agricultural tilling, sources of road dust,
and industrial point source emissions. There has been an extensive movement toward
the use of reduced and no tilling production, field mapping, and reduced numbers of
trips through fields for pest control. We also suspect that road dust emissions may well
have been reduced because of actions by local agencies to control urban runoff that
degrades surface water quality. In addition, economic growth will not necessarily lead
15
-------
to increased emissions from industrial point sources since the nature of industries
continues to change. Specifically, the areas of active growth have been in high
technology goods that produce relatively low emissions per unit of contribution to the
economy. In addition, much of the heavy industries that might represent significant
emissions have initiated aggressive control strategies for reasons other than air quality
regulations (e.g., a focus on pollution prevention as well as improved community
relations). Further, another significant fraction of heavy industries have moved out of
the United States to locations with cheaper labor and less rigorous environmental
regulations. In total, all of these changes need to be considered in the emissions
estimates.
In addition to changes in directly emitted PM, there have been substantial
decreases in emissions of man-made precursors of particulate matter formed in the
atmosphere. For instance, the CAAA90 acid deposition program requirements have
led to substantial reductions in S02 and NOX emissions. Further reductions are
required. In addition, ozone controls are leading to substantial reductions in volatile
organic compound (VOC) and nitrogen oxide (NOX) emissions. Both assessments and
ambient data indicate that these factors have been major contributors to decreasing
PM10 concentrations.
The AQMS also advised that the recent Grand Canyon Visibility Transport
Commission Assessment of Scenarios (Middleton and associates, 1996) provides an
opportunity for cross-checking emission projection assumptions and techniques. That
study primarily focused on visibility changes in national parks and wilderness areas in
the Grand Canyon and nearby parks. The impacts of emission changes throughout the
entire western U.S., from 1990 to 2040, were assessed. The emissions projections for
the baseline, which was developed assuming current regulations were in place, showed
an increase in visibility in 2000 and then a decrease from 2010 to 2040. Overall
region- wide directly emitted fine organic carbon, other fine and coarse particles were
projected to increase from 2000 through 2040. These increases were mainly
associated with the steadily projected increases in road dust emissions throughout the
assessment period. This emission category also was recognized to be the most
uncertain. Other point sources, area sources and transportation sources were
projected to have slight increases in emissions for the region, but only beyond 2010.
The emissions for sulfur and nitrogen oxides, VOC and elemental carbon all showed
decreases through 2010. For 2020 through 2040, emissions of NOX, VOC, and all
particles showed increases for the U.S. western regional totals.
4.6.3 Possible Steps
Considering the emission modeling tools now available and the resources to
prepare the current prospective study, we are wondering what options are available for
dealing with the discrepancy between observed PM air quality trends and the
16
-------
projections of air quality changes. Since the emission modeling problems appear to be
associated primarily with direct emissions of PM, one possibility is to look only at
modeled concentrations of secondary PM (i.e., the particles formed in the atmosphere).
As another possibility, some diagnostic information may be available by
examining the contribution of crustal material to PM2 5 and by examining the RADM
results before adding primary crustal materials and secondary organic particles. These
results can help to suggest if it is the fine particle tail in the crustal particle size
distribution that is contributing to the simulated increases in PM concentrations in both
size ranges and if the secondary particles in the eastern U.S. follow the expected
decreasing trend.
We also suggest that there could be problems in operating the modeling system.
It may be worthwhile to obtain an expert review of the modeling effort by someone
outside of the program who can visit the modelers and carefully review the process by
which the modeling is performed. There have been previous instances when this
approach has uncovered problems that the modelers could not detect because of their
proximity and familiarity to the process. We hope that these steps will lead to a better
understanding of the reasons for the divergent outcomes and permit modifications
which will lead to more credible results.
4.7 Overall Quality of Emissions Inventories
The apparent results of the PM modeling exercise lead to a major question
regarding the specification of the quality of emission inventories. Emission data
continue to be the most substantial problem in all aspects of air quality modeling. No
matter how sophisticated the effort to model the atmospheric chemistry and
meteorology, the effort is doomed to failure if there are large errors in the input values
of the emissions. For example, even in Los Angeles where extensive efforts have been
made to produce an accurate emissions inventory, there remain important problems
that need to be resolved (Fujita et al., 1992). Many other instances of such problems
exist. Therefore, the AQMS recommends that there be strong support within the EPA
to examine this problem and take action to improve the current process for assembling
inventories and to specify the quality of the emissions information.
4.8 Additional Scenario Analysis
We strongly advise that the trends problem be resolved first before moving on
the new scenario runs. Consequently, we recommend that no additional scenarios be
run until the issues surrounding the PM trends have been adequately resolved. The
Agency should use the available resources to investigate the underlying problems in
the analysis.
17
-------
When the trends discrepancy is resolved, we recommend that future emission
scenarios be simulated. The disaggregation approach (i.e., identifying the individual
impact of a variety of proposed emission management activities) suggested by some of
the Council Members is important, but does not seem feasible within the context of the
current prospective study. Multiple assessment runs would be needed in order to
isolate the individual effects for different policies. The current prospective modeling
and emissions system is not set up for quick and inexpensive turnaround. In response
to these obvious problems, the prospective study team has proposed an alternative,
that is, looking at additional reasonable scenarios to demonstrate the potential effects
of additional controls. The AQMS believes that, given the current resource and data
limitations, this is a viable alternative.
The new scenarios, if conducted, should start with a combination of the
scenarios presented by the EPA staff. This would provide an upper bound on possible
effects. If significant changes are indeed noted, then additional scenarios should be
considered.
18
-------
5 SUMMARY
The AQMS looks forward to continued interaction with the prospective study
team. As a committee, and as individuals, we especially would like to assist the study
team toward a successful resolution of the trend issue in the near-term. We also
anticipate continuing to provide review of future prospective study designs, strategies
and scenarios. Each of us will be available as individuals for consultation should be
study team need more detailed assistance in the months ahead.
In developing plans for the future, the prospective study team's emphasis
probably ought to be on the need to consider more comprehensive, integrated
modeling approaches such as provided by the emerging Models-3 Platform. We also
emphasize the need for a more flexible, comprehensive emissions modeling system
that will provide the ability to better diagnose emissions data problems and to more
easily examine multiple emissions management scenarios.
Again, we appreciate the professionalism of the study team and the excellent
work accomplished to date on this difficult, but important, assessment.
19
-------
REFERENCES CITED
Bogaert, P. and G. Christakos, 1997,:"Spatiotemporal Analysis and Processing of
Thermometric Data over Belgium," Jour. Geophys. Res.-Atmospheres, v. 102, n.
D22, pp. 25831-25846
Christakos, G. and G.A. Thesing, 1993, "The intrinsic random field model in the study
of sulfate deposition processes, Atmos. Env. v. 27A, n. 10, pp. 1521-1540
Christakos, G. and D. Hristopulos, 1996, "Characterization of atmospheric pollution by
means of stochastic indicator parameters," Atmos. Env., v. 30, n. 22, pp. 3811-
3823
Christakos, G. and J. Lai, 1997, "A study of the breast cancer dynamics in North
Carolina," Soc. Sci. Mediicine, v. 45, n. 10, pp. 1503-1517
Christakos, G. and V. Vyas, 1997, "A novel method for studying the health impacts of
spatiotemporal ozone distribution," Soc. Sci. Medicine, in review
Christakos, G. and X. Li, 1998, "Bayesian maximum entropy analysis and mapping: A
farewell to kriging estimators?" Mathematical Geology, v. 30, n. 3
Darlington, Thomas L., Dennis F. Kahlbaum, Jon M. Heuss and George T.Wolff,
"Analysis of PM10 Trends in the United States from 1988 through 1995," J. Air &
Waste Manage, Assoc., v. 47, pp. 1070-1078, October, 1997
Fujita, E.M., B.E. Croes, C.L. Bennett, D.R. Lawson, F.W. Lurmann, H.H. Main, 1992, "
Comparison of Emission Inventory and Ambient Concentration Ratios of CO,
NMOG, and NOX in California's South Coast Air Basin," J. Air & Waste Manage.
Assoc., vol 42, pp. 264 - 276, 1992
Middleton, Paulette and Associates,, Grand Canyon Visibility Transport Commission
Assesment of Scenarios, Western Governors Association Reports, Science &
Policy Associates, (Complete reports are available from Western Governors
Association), Volumes I-IV, 1996
ICF Kaiser, Systems Applications International Inc., 1998, "Prospective Analysis of Air
Quality in the U.S.: Air Quality Modeling,"Draft Final Report, SYSAPP98-97/69
U.S. EPA , 1998a, EPA Third-Generation Air Quality Modeling System, Models-3
Volume 9a: System Installation and Operations Manual. EPA/600/R-98/069(a),
National Exposure Research Laboratory, Research Triangle Park, NC, 131 pp.
R-1
-------
U.S. EPA, 1998b, EPA Third-Generation Air Quality Modeling System, Models-3
Volume 9b: User Manual. EPA/600/R-98/069(b), National Exposure Research
Laboratory, research Triangle Park, NC, 833 pp
U.S. EPA/OPAR, June 13, 1997, Memorandum from Mr. James DeMocker, OPAR to
Dr. Jack Kooyoomjian, SAB/Council DFO entitled "Information for SAB Council
on Section 812 Prospective Study"
U.S. EPA/ORD, 1996, Air Quality Criteria for Particulate Matter, Vol III of III, Office of
Research and Development, Washington, D.C., EPA/600/P-95/001cF, 1996
U.S. EPA/SAB, September 9, 1997, "Letter Report of the Air Quality Models
Subcommittee (AQMS) of the Council's Science Advisory Board on Review of
the Clean Air Act Amendments (CAAA) of 1990, Section 812 Prospective Study
Emissions Modeling and Associated Air Quality Modeling Issues," EPA-SAB-
COUNCIL-LTR-97-012
Vyas, V. and Christakos, G., 1997, "Spatiotemporal analysis and mapping of sulfate
deposition data over eastern U.S.A.," Atmos. Env., v. 31, n. 21, pp. 3623-3633
Yang, Y-j., Wilkinson, J. and Russell, A., 1997, "Fast, Direct Sensitivity Analysis of
Multidimensional Photochemical Models," Env. Sci. Technol., v. 31, pp. 2859-
2868
R-2
-------
APPENDIX A- Modeling Draft Review Materials1
1) DeMocker, James, Jan. 22, 1998, "Air Quality Modeling Overview for the First
Section 812 Prospective Study,"Briefing for SAB AQMS (Item 179)
2) DeMocker, James, Jan-Feb. 1998, "Analytical Strategy for the First Section 812
Prospective Study,"Briefing for SAB Council, HEES, and AQMS (Item 136a)
3) Dennis, Robin L, January 5, 1998, "Regional P'articulate Model: Background
Analysis for 812 Prospective Study,"Draft White Paper, Atmospheric Modeling
Division, National Exposure Research Laboratory, Office of Research and
Development, RTP, NC
4) Dennis, Robin L., October, 1995, "Estimation of Regional Air Quality and
Deposition Changes Under Alternative 812 Emissions Scenarios Predicted by
the Regional Acid Deposition Model, RADM, "A Report Prepared for the 812
retrospective Study in Coordination with Jim DeMocker, OPAR, U.S. EPA
5) Douglas, Sharon, undated, "812 Prospective Air Quality Modeling, "SAI, Inc., San
Rafael, CA (A Presentation to the AQMS) (Item 138a)
6) ICF Kaiser/SAIC (Science Applications International, Inc.), January, 1998,
"Prospective Analysis of Air Quality in the U.S.: Air Quality Modeling," Draft Final
report, SYSAPP98-97/69
7) ICF Kaiser/SAIC (Science Applications International, Inc.), undated, "REMSAD:
The regulatory Modeling System for Aerosols and Deposition, Modeling Primary
and Secondary Fine Particle Distributions."
8) ICF Kaiser/SAI (Systems Applications International, Inc.), Decermber, 1995,
Technical Memorandum, "Development and Preliminary Testing of the
Regulatory Modeling System for Aerosols and Deposition (REMSAD), "SYSAPP-
96/11
9) Langstaff, John E., September, 1991, "Overview of the SAI Urban Airshed
Model," Prepared for Barry Elman, OPPE, U.S. EPA
10) Neumann, Jim and Mike Hester, June 23, 1997, "Evaluating Benefits of Marginal
Emissions for the 812 Prospective Analysis: Overall Approach and Immediate
1 These documents are available from the U.S. Environmental Protection Agency
(U.S. EPA), Office of Air and Radiation (OAR)/Office of Policy Analysis and Review
(OPAR), Mail Code 6103, U.S. 401 M Street, SW, Washington, DC 20460
A-1
-------
Action Items from June 17 Teleconference,"Industrial Economics, Inc. (lEc)
11) Neumann, Jim and Mike Hester, July 14, 1997, "Proposed Modeling Scenarios
for Evaluating Benefits of marginal Emissions reductions Beyond the 812
Prospective Analysis Post-CAAA Scenario," Industrial Economics, Inc. (lEc)
12) U.S. EPA/OPAR, August 29, 1997, "Analytical Strategy for the First Section 812
Prospective Study" (DRAFT)
13) U.S. EPA/OPAR, Jan-Feb., 1998, "Analytical Strategy for the First Section 812
Prospective Study /'Briefing for SAB Council, HEES and AQMS
A-2
-------
APPENDIX B - Additional Information on Recommendation to
Consider Use of EPA's Urban and Regional Community Air Quality
Modeling System
Atmospheric modeling practice has evolved rapidly in the last 6 years. The
EPA/ORD/NOAA program has recently released a third generation modeling system,
known as MODELS-3, which is the result of more than five years of collaborative and
in-house work, that has a unique formulation permitting various "science modules" to
be plugged and un-plugged easily. An attractive feature of MODELS-3 is that it can
grow with scientific advances without having to replace the model wholesale as occurs
with present models. MODELS-3 is actually a modeling system that is capable of
supporting air quality, as well as, river and ground water, and estuary models to allow
a multi-media assessment to be performed similar to that done for atmospheric nitrogen
deposition into the Chesapeake Bay. The present implementation of this modeling
system includes the use of: a) advanced non-hydrostatic meteorological models that
use four-dimensional observational data assimilation; b) multiple chemical reaction
mechanisms each of which can be expanded to include new chemicals; c) primary and
secondary modal aerosol sub-model; d) aqueous and cloud chemistry sub-models with
wet deposition processes; e) visibility post-processors; and f) analysis and data
visualization packages. The model can be operated in a telescoping, one-way nesting
mode that permits it to simulate the entire United States at resolutions of 108 and 36
km, to simulate a nested region at 12 km resolution, or to simulate urban areas at
resolutions of 4 or 1.3 km. Currently the model extends from the surface to 12 km with
27 logarithmically-spaced vertical layers having as many as 15 or more layers in the
boundary or surface mixed layer. The model development team has produced
informative scientific and operational documentation and the model will be put through
at least one major evaluation exercise with field data before or just after its release to
the public.
Recent studies, including the OTAG program have shown that for regional scale
analyses, simulation time periods as long as 18 days are probably necessary and that
to simulate ozone formation for a week period or longer over any of the mid-size urban
areas in the eastern U.S., the outer 36 km grid needs to cover at least the entire
eastern U.S. This sets new standards for good modeling practice. The prototype of
MODELS-3 is being used in a 1995 full-summer season simulation for the eastern U.S.
The EPA has committed to re-doing the so-called 30 case "aggregation" data set used
to estimate annual deposition using the Models-3/CAQM system and to extending the
36 km meteorological simulations to include the continental U.S.
The model is intended to be a "community" model in that it would become freely
available. Because of its design, it could be expanded or have additional functionality
added to the basic system. It is intended that at regular intervals the new additions and
advances will be evaluated and certified for use with the rest of the system and will
B-1
-------
become readily available for use by the entire user community.
The widespread dissemination and use of such a community tool for both the
regional and urban atmospheric chemistry modeling research community, as well as for
meeting the EPA regulatory requirements to produce State Implementation Plans
(SIPs) by the states is expected to result in a well-documented and well-understood set
of simulations conditions and model file sets. These should permit individual
investigators or private companies to conduct high quality modeling exercises in a
nested mode without having to bear the full expense and effort to produce their own
analyzed and well-understood scenarios.
Additional advances in the MODELS-3/CAQM system that are currently planned
include:
a) The enhancement and modification of the North Carolina Supercomputing
Center's SMOKE emissions processor system for use in the MODELS-3
framework. The SMOKE system can produce an OTAG-domain
emissions inventory file set in only 12 minutes of clock time. Work to
expand the SMOKE system includes the addition of the processing
options to include special calculations for reactivity that permit placing the
emissions in the correct locations at the correct times for different classes
of VOC, e.g., industrial VOC emissions would be incrementally added to
the correct SCC (Source Classification Codes), alternative fuel emission
would be associated with on-road VMT (Vehicle Miles Traveled).
b) The addition of the direct differential method for simultaneously
computing local sensitivity of ozone (or other model state variable) to as
many as 30 different VOC emissions changes at the same time as the
species are being computed (or other relevant parameters). This method
has recently been refined and implemented (Yang et al., 1997) in another
modeling system and it needs to be integrated into the CAQM
(Community Air Quality Model) system.
c) The integration of a fully "study planner" support system into the
MODELS-3 decision support framework.
B-2
-------
APPENDIX C - GLOSSARY OF TERMS AND ACRONYMS
ACCACA
AQMS
ADV
AUSPEX
CAA
CAAA
CAAA90
CAQM
CO
EPA
LTR
M/C
MODELS-3
NH3
NOAA
NOX
03
OAR
OPAR
OPPE
ORD
OTAG
PM
PM-Primary
PM-Secondary
RADM
SAB
SAQM
SARMAP
SCO
Advisory Council on Clean Air Compliance Analysis (the Council)
Air Quality Models Subcommittee (of the Science Advisory Board,
SAB/Council, U.S. EPA)
SAB Advisory Report
Atmospheric Utilities Signatures Predictions and Experiments (A
Cross-Tracer Study/Experiment)
Clean Air Act
Clean Air Act Amendments
Clean Air Act Amendments of 1990
Community Air Quality Model
Carbon Monoxide
Environmental Protection Agency (U.S. EPA)
SAB Letter Report
Members/Consultants
EPA's Third-Generation Air Quality Modeling System Advanced
Object-Oriented and Network Aware Models Program Developed
by EPA's National Exposure Research Laboratory, Research
Triangle Park, NC
Ammonia
National Oceanic and Atmospheric Administration
hJitrogen Oxides
Ozone
Office of Air and Radiation (OAR, U.S. EPA)
Office of Policy Analysis and Review (OAR/OPAR, U.S. EPA)
Office of Policy Planning and Evaluation (OPPE, U.S. EPA)
Office of Research and Development (ORD, U.S. EPA)
Ozone Transport Assessment Group
Particulate Matter (PM10 and PM 2.5; that is, particulate matter that
is less than or equal to 10 micron and 2.5 micron in size,
respectively)
Primary Particulate Matter
Secondary Particulate Matter
Regional Acid Deposition Model
Science Advisory Board (U.S. EPA/SAB)
SARMAP Air Quality Model
SJVAQS/AUSPEX Regional Modeling Adaption Project
Source Classification Codes
C-1
-------
SCCSMOKE Supercomputing Center SMOKE (in North Carolina)
SIPs State implementation Plans
SJVAQS San Joaquin Valley Air Quality Study
SMOKE Sparse Matrix Operator Kernel Emissions Processor (The North
Carolina Supercomputing Center's Emission Processor System for
Use in the MODELS-3 Framework)
S02 Sulfur Dioxide
SOX Sulfur Oxides
U.S. United States
VMT Vehicle Miles Traveled
VOC Volatile Organic Compounds
C-2
-------
DISTRIBUTION LIST
Administrator
Deputy Administrator
Assistant Administrators
Regional Administrators
Office of the Administrator
Office of Cooperative Environmental Management (OCEM)
Deputy Assistant Administrator for Air and Radiation
Director, Office of Policy Analysis and Review (OPAR)
Director, Office of Air Quality Planning and Standards (OAQPS)
Deputy Assistant Administrator for Policy, Planning and Evaluation (OPPE)
Director, Office of Policy Analysis (OPA)
Director, Office of Regulatory Management and Evaluation (ORME)
Director, Office of Strategic Planning and Environmental Data (OSPED)
Deputy Assistant Administrator of Research and Development:
Deputy Assistant Administrator for Science - ORD
Deputy Assistant Administrator for Management - ORD
Director, Office of Science Policy
Director, Mega Laboratories and Centers
Director, Research Laboratories
EPA Headquarters Library
EPA Regional Libraries
National Technical Information System (NTIS)
Library of Congress
-------
United States Science Advisory EPA-SAB-COUNCIL-ADV-98-002
Environmental Board September 1998
Protection Agency Washington, DC www.epa.gov/sab
&EPA AN SAB ADVISORY: THE CLE/
ACT SECTION 812 PROSPECTI1
STUDY OF COSTS AND BENEF
AIR QUALITY MODELS AND
EMISSIONS ESTIMATES INITIAI
STUDIES
BY THE ADVISORY COUNCIL ON CLEf
COMPLIANCE ANALYSIS OF THE SCIE
ADVISORY BOARD
------- |