United States
Environmental
Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA-450/4-91-013
July 1991
AIR
    GUIDELINE FOR REGULATORY

       APPLICATION OF THE

     URBAN AIRSHED MODEL

-------

-------
 GUIDELINE FOR REGULATORY APPLICATION OF THE
             URBAN AIRSHED MODEL
    U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air Quality Planning and Standards
         Technical Support Division
       Source Receptor Analysis Branch
     Research Triangle Park, NC   27711
                 July  1991

-------

-------
 ;                          DISCLAIMER
                                               I                  '
 |    This report has been reviewed by the Office of Air Quality

Planning and Standards, U.S.  Environmental Protection Agency, and

has been approved for publication.  Any mention of trade names or

commercial products is not intended to constitute endorsement or

recommendation for use.

-------

-------
                             CONTENTS
                                                             Page
DISCLAIMER	

TABLES	   v
                                                |

ABBREVIATIONS 	 .	  ............ vii

ACKNOWLEDGEMENTS	. . .	!.   	ix

1    INTRODUCTION	   i

2    MODELING PROTOCOL	   5
                                                I
     2.1  Protocol Development Process  .... 	   6
     2.2  Contents of Protocol Document .... 	   8

3    DOMAIN AND DATA BASE ISSUES	i........  11
                                                i
     3.1  Episode Selection	!	11
     3.2  Size of the Modeling Domain	;.	  15
     3.3  Horizontal Grid Cell Size	  15
  ,   3.4  Number of Vertical Layers  .  .	17
     3.5  Meteorological Data	  19

  j        3.5.1   Wind fields	  19
          3.5.2   Data needs for wind field     i
                  development	23
          3.5.3   Mixing heights	;	24
          3.5.4   Clear-sky assumption for photolysis rate
                  calculations	25

     3.6  Air Quality .'....'	26

          3.6.1   Initial  and boundary conditions 	  26
          3.6.2   Performance Evaluation Data	  .  31

     3.7  Emissions	............  32

          3.7.1   VOC speciation	35
  |        3.7.2    Spatial  gridding of area sources  	  35
  |        3.7.3    Mobile sources  .	36
          3.7.4    Episode-specific adjustments  .......  37
  ;        3.7.5    Biogenic emissions  	 	  38
          3.7.6    Point-source and plume-rise cutoff levels  .  39
  !        3.7.7    Consistency with national inventories ...  40
                               111

-------
CONTENTS f Continued)

                                                             Page

4    DATA QUALITY ASSURANCE AND MODEL DIAGNOSTIC ANALYSES .  .   41

     4.1  Step 1:  Quality Assurance Testing of Component
                   Fields	43
     4.2  Step 2:  Diagnostic Testing of the Base Case
                   Meteorological episodes	   44
     4.3  Step 3:  Additional Base Meteorological Episode
                   Sensitivity Testing  	   46

5 MODEL PERFORMANCE EVALUATION  	 ........   49

     5.1  Performance Measures  	 .....   49

          5.1.1   Graphical performance procedures  .....   51
          5.1.2   Statistical performance measures  .....   53

     5.2  Assessing Model Performance Results ........   55

6    ATTAINMENT DEMONSTRATION 	 .....   59

     6.1  Developing Attainment-Year Model Inputs  ......   59
     6.2  Construction of Attainment Year Emission Control
          Strategies  	 	 .....   60
     6.3  Performing Attainment-Year Simulations to Assess
          Various Control Strategies   	 .....   62
     6.4  Using Modeling Results in the Attainment
          Demonstration 	 	 .....   63
     6.5  Exceptions to Guidance Document ..........   64
REFERENCES   	 .

APPENDIX A  RECOMMENDED MODELING PROTOCOL CONTENTS  ,

APPENDIX B  IDENTIFICATION OF METEOROLOGICAL REGIMES
            CORRESPONDING WITH HIGH OBSERVED OZONE
65

69


77
APPENDIX C  PERFORMANCE MEASURE FORMULATIONS   .  .  	  81
                                IV

-------
                         TABLES
Number
          Example Table of Contents for Protocol
          Document	
   2      Default Boundary Condition Concentrations for
          Carbon Bond-IV Species   ..........    28

-------
VI

-------
                          ABBREVIATIONS
AIRS      Aerometric Information Retrieval System
BEIS      Biogenic Emissions Inventory System
CAAA      Clean Air Act Amendments
GARB      California Air Resources Board
 i
CMSA      Consolidated Metropolitan Statistical Area
CO        Carbon Monoxide
CSC       Computer Sciences Corporation
DWM       Diagnostic Wind Model (UAM preprocessor program)
EKMA      Empirical Kinetic Modeling Approach  !
EPA       U.S. Environmental Protection Agency
EPS       Emissions Preprocessor System for the UAM
GMISS     Gridded Model Information Support System
MSA       Metropolitan Statistical Area
NAAQS     National Ambient Air Quality Standard(s)
NO        Nitric Oxide
N6X       Nitrogen oxides
NO2       Nitrogen Dioxide                     ;
NSR       New source review                    j
NTIS      National Technical Information Service
 I                            -
NWS       National Weather Service             !
OAQPS     EPA Office of Air Quality Planning and Standards
OMS       EPA Office of Mobile Sources         f
ORD       EPA Office of Research and Development
PWD       Predominant wind direction
RACT      Reasonably available control technology
RFP       Reasonable further progress (a type of tracking
             required under Section 182 of the CAAA)
ROM       EPA Regional Oxidant Model
SAI       Systems Applications International
                               VI1

-------
ABBREVIATIONS f Continued ~>

SCC       Source Category Code
SCRAM BBS EPA Support Center for Regulatory Air Models
             Bulletin Board System
SIP       State Implementation Plan
UAM       Urban Airshed Model
UV        Ultraviolet
VOC       Volatile Organic Compound
VMT       Vehicle Miles Traveled
                               Vlll

-------
                         ACKNOWLEDGEMENTS
     Mr. Dennis C. Doll is the principal author of this document.
Dr. Richard D. Scheffe, Dr. Edwin L. Meyer, and Mr. Shao-Hang Chu
also contributed significantly to the guidance document's content.
Mrs. Cynthia J. Baines was responsible for typing and assembling
the manuscript.

     A draft version of this guidance document Has been subject to
p'ublic review and comment through inclusion in pocket A-88-04 for
 '
public comment on  the Fifth Conference on Air Qusility Modeling held
                                               f
in March 1991.  In addition, the  authors would like to acknowledge
the efforts of  a  working group set  up  to  review and advise them
regarding the issues addressed in the guidance document.  While all
the recommendations do not necessarily reflect j the views of each
working  group  member  nor  their  organizations,  members'  free
exchange of ideas and  the generoxis  amount  of time they spent in
reviewing  various  drafts  added  materially  to  the  guidance
document's content.  Working group members  (in alphabetical order)
ijncluded  Dr.   C.S.  Burton  (Systems Applications  International
Incorporated  (SAI, Inc.)), Mr. C. Durrenberger (Texas Air Control
Board), Mr. G. Gipson (U.S. Environmental Protection Agency (EPA),
Office  of  Research and  Development (ORD)),  Mr. T.  Helms  (EPA,
Office  of  Air Quality Planning  and Standards),  Dr.  H. Jeffries
('university of North Carolina), Mr. M. Koerber (Lake Michigan Air
 i
Directors  Consortium),  Dr.  C.   Liu  (South  Coast  Air  Quality
Management  District),  Mr.  T.  McGuire  (California  Air  Resources
Board),  Mr.  R.  Morris  (SAI, Inc.),  Dr.   S.  T.  Rao  (New  York
Department  of  Environmental Conservation), Dr.
P. Roth  (private
consultant),  Mr.  K.  Schere  (EPA, ORD),  Dr. T.  Tesche  (private
consultant),  Mr.  J.  Vimont (U.S. National  Park Service), Mr. D.
                                IX

-------
Wackter (Connecticut Department of Environmental Protection), and
Dr. S. Ziman (Chevron USA, Inc.).
     Finally,  the authors  would  like to  thank  Ms.  Jeanne  R.
Eichinger (CSC) for her thorough review and many helpful editorial
comments.

-------
                            CHAPTER 1
                           INTRODUCTION         |

     The Clean  Air Act Amendments  (CAAA)  of 1990  require ozone
nonattainment areas  designated as  extreme,  severe,  serious,  or
multi-State  moderate  to   demonstrate  attainment  of  the  ozone
National Ambient Air Quality Standard (NAAQS) through photochemical
grid modeling or  any  other  analytical  method determined  by the
Administrator to  be at least  as effective.*   The  Environmental
Protection Agency (EPA) has adopted the Urban Airshed Model (UAM)
as the guideline model for photochemical grid modeling applications
involving  entire   urban  areas.    Procedures described  in  this
guidance document are intended to  satisfy  the CAAA attainment
demonstration requirements,   foster technical  credibility,  and
promote consistency among UAM regulatory applications.
     This guidance document  provides  recommendations and general
procedural guidance for exercising the UAM (described in
References 1-5)  in regulatory applications. However, methodologies
and procedures discussed in this guidance document generally apply
fjor  other   urban-scale  photochemical  grid  jmodels   as  well.
 J     The CAAA does not specify the method for demonstrating
attainment for within-State moderate areas.  Thus, the EPA has
determined that the use of the Empirical Kinetic Modeling
Approach (EKMA) may be sufficient for demonstrating attainment
fpr these areas.  However, the use of a photochemical grid model
is preferred.

-------
Acceptance criteria for alternative models is beyond the scope of
this   guidance  document.     Use  of   alternative  urban-scale
photochemical  grid models as well as regional-scale photochemical
grid models other than the EPA Regional Oxidant Model (ROM) must be
addressed on a case-by-case basis through the EPA Regional Offices.

     The  UAM  source  code is maintained  and distributed  by the
Source Receptor Analysis Branch, Technical Support Division of the
EPA Office of  Air Quality Planning and Standards (OAQPS).  Users
will  be informed of  modifications  or enhancements  to  the UAM
through the Support Center for Regulatory Air Models  Bulletin Board
System  (SCRAM BBS).   Additionally,  the UAM  source code, user's
guide,  and  test case  data  base are available  from the National
Technical  Information  Service   (NTIS)(703-737-4600).   The  NTIS
document numbers  are noted in the reference list.

     This guidance document provides recommendations  and procedures
for conducting an ozone analysis with the  UAM for ozone attainment
demonstrations.    Some  of  the  recommendations  and  procedures
described were adopted from the California  Air  Resources Board
(GARB)  Technical  Guidance Document; Photochemical  Modeling6, and
will be referenced as  such throughout the text.

     Steps needed to conduct an urban-scale photochemical modeling
study using  the UAM typically consist of the  following:

     1.   Establish a  protocol  for the modeling study.

     2.   Compile air  quality,  meteorological, and  emissions data
          to  develop  UAM  input files  for  each   meteorological
          episode to be used in the attainment demonstration model
          simulations.
  introduction

-------
3.   Execute the UAM for each meteorological episode.
                  • -    .                 •  • I
4.   Conduct  diagnostic  ancilyses  on  each  meteorological
     episode simulation.  The principal purpose of diagnostic
     analyses   is   to  ensure   that  the  model   properly
     characterizes physical and chemical phenomena (e.g., wind
     fields,   spatial  and   temporal  emission   patterns)
     instrumental in leading to observed ozone concentrations.
     The visible product is enhanced model performance (i.e.,
     better  spatial  and  temporal agreement  with  observed
     data).  Diagnostic model  simulations  uncover  potential
     model input data gaps that, when corrected,  may lead to
     improved model  performance.
     Exercise the  UAM for  each  meteorological episode  and
     conduct a series  of performance measures  to  determine
     overall model performance in replicating observed ozone
     concentrations and patterns.  Model performance evalua-
     tion should  also be  done for ozone precursors (e.g.,  NO,
     NO2)  if  suitable monitoring  data are available.
                                                •
                  •
     For each meteorological episode,  estimate emissions  and
     air quality for the projected  attainment year  required
     under the  CAAA.    Perform model  simulations  for each
     episode to determine whether the  ozone NAAQS can be  met
     in the attainment year.
     If the model simulations for the attainment year do not
     show  attainment  for  each  modeled  episode,   develop
     emission control measures on selected source categories
     (for example, volatile  organic compound  (VOC) and/or NO
     controls on  selected source categories, alternative fuel
     scenarios, etc.).

-------
r
                8.   Perform  model  simulations  for  the emission  control
                     measures to demonstrate attainment of the ozone NAAQS for
                     each meteorological episode.  If the control measures do
                     not show attainment, repeat steps 7 and 8 as an iterative
                     process  until  attainment  is  shown  for  each  modeled
                     episode.

                These  steps are  addressed  in  subsequent chapters  of this
           guidance document as follows::

                Chapter 2:  Modeling Protocol
                Chapter  3:
Domain and Data Base Issues
   •Episode selection
   •Domain selection/grid spatial allocation
   •Meteorological/air quality data
   •Emission inventories
                 Chapter  4:   Data Quality Assurance and Model Diagnostic
                             Analyses

                 Chapter  5:   Model Performance Evaluation

                 Chapter  6:   Attainment Demonstration
             .Introduction '•"'"

-------
                             CHAPTER 2
                         MODELING PROTOCOL
      Regulatory application of the UAM potentially affects a broad
 spectrum  of  society.    The UAM  modeling domains  may  encompass
 multiple  geopolitical boundaries  (counties,  cities,  and  States)
                                                   Therefore, the
with  a potentially  large regulated  community.
development of a Modeling Protocol is required, j This Protocol is
necessary to (l) promote technical credibility, (2) encourage the
participation of all interested parties,  (3) provide for consensus
building among all interested parties concerning modeling issues,
and  (4)  provide  documentation for  technical  decisions made  in
applying the model as well as the procedures followed in reaching
these decisions.                                '
     The  Protocol  should detail  and  formalize procedures  for
conducting all phases of the modeling study, such as  (1) describing
the  background  and objectives  for the  study,  (2) creating  a
schedule and organizational structure for the study,  (3) developing
the  input  data,  (4) conducting diagnostic  and model performance
evaluations,  (5)  interpreting  modeling results,  (6)  describing
procedures  for  using the model to demonstrate  whether proposed
strategies  are  sufficient  to  attain  the  ozone  NAAQS,  and  (7)
producing documentation and data analyses  that must be submitted
for EPA Regional Office review and approval.
     All issues concerning the modeling  study must be thoroughly
addressed during the Protocol development.   Modifications to the
Protocol as  the  study progresses  should not  be needed unless

-------
significant, unforeseen  procedural  and/or  technical  issues  are
encountered.  All  parties involved in the  study  should agree to
Protocol  modifications  through  the Modeling  Policy  Oversight
Committee (see below).   It is especially important that the State/
local agencies  and EPA  Regional  Office(s)  overseeing the  study
concur on Protocol modifications.

2.1  Protocol Development Process

     Ordinarily, the State  agency responsible for developing the
ozone State Implementation  Plan  (SIP)  is  also  the  lead agency
responsible  for developing  the Modeling Protocol.   For domains
encompassing parts  of  more  than one State,  the responsible State
agencies  need  to  develop  the  Modeling  Protocol jointly.   The
Protocol  should  describe   the modeling  policy and technical
objectives of the study.  This will require  input  from various EPA
and State/local personnel dealing with regulatory policy issues and
from others  with modeling expertise.  It is  likely that  Modeling
Policy  Oversight  and  Technical  Committees  will be  needed for
addressing these issues.  The composition and responsibilities of
the Committees  should  be  defined  in the  Modeling  Protocol.

     Responsibilities  of the Modeling Policy Oversight Committee
may be, at  a minimum,  to set the  objectives of  the study, set the
schedule, determine resource needs,  and implement any modifications
to the  Protocol as the modeling study  proceeds.  The Committee
should  include  representatives from the  appropriate  EPA  Regional
Office(s),  State/local  agencies,  the  regulated community, and
public  interest groups.   It is  important that appropriate policy-
oriented  personnel be  identified for membership on the Committee.

     Responsibilities  of  the   Technical  Committee may  be,  at  a
minimum,  to develop   the  Protocol's  technical specifications
  Protocol

-------
 concerning emission inventories,  meteorological data,  air quality
 data,  data  quality  assurance,  development  of emission  control
 strategies, model diagnostic analyses, model performance evaluation
 procedures,  and interpretation of  model results.   The Technical
 Committee should include appropriate technically-oriented members
 from  the  EPA  Regional  Office(s),  State/local  agencies,   the
 regulated community,  and public interest groups.

     For  some  areas,   regional  modeling  is  being  planned  to
 establish initial  and boundary conditions for urban-area modeling
 attainment  demonstrations.    The  urban-area
 development  should  be  coordinated with  the
Modeling  Protoco1
regional  Modeling
 Protocol.    Some  members of the  urban-area Policy Oversight  and
 Technical Committees would probably also be members of the regional
 Policy Oversight  and Technical Committees.
 i    The Modeling Protocol must be submitted to the appropriate EPA
Regional  Modeling  Contact for  review  and approval.    The EPA
Regional  Modeling  Contact  should  be  a  member  of  the Policy
Oversight  and/or Technical Committees  so that  rapid  review and
approval of the  Protocol  is assured.
     Recommendations
     A Protocol Document is required for each UAM application used
     for an ozone attainment demonstration.  This -Protocol should
     describe the methods and procedures to be used for conducting
     the photochemical modeling study.
     Additionally, it is recommended that both a Policy Oversight
     Committee and a Technical Committee be established to develop
      A regional  Modeling Protocol  is  being prepared by the EPA
OAQPS and will be available by late 1991.  This regional Modeling
Protocol will facilitate coordinating regional model applications
to support the nonattainment area SIP UAM applications.

-------
r
                the Modeling Protocol.  The composition and responsibilities
                of the Committees should be defined  in the Protocol.

                The Modeling Protocol  and  any modifications to it should be
                agreed upon -by all parties involved in the study,  through the
                Policy Oversight Committee.   It is especially  important that
                the State/local agency participants and EPA Regional Office(s)
                overseeing  the  modeling   study  concur  on  any  Protocol
                modifications.   Protocol  modifications should be  documented
                for subsequent public  review.

                For  some nonattainment areas,  regional  modeling  is being
                planned to provide initial and boundary conditions as  well as
                other   inputs  for   the  urban-area  modeling   attainment
                demonstrations.  Procedures for coordinating the  development
                of the urban-area Modeling  Protocol with the regional Modeling
                Protocol  should be clearly described.

                It is especially important that close technical coordination
                be maintained during  the  Protocol  development among  nearby
                urban-area  domains   within  a  regional  modeling   domain.
                Procedures should be established for coordinating the Modeling
                Protocols among these areas, and these coordination procedures
                should  be  clearly  specified  in each  nonattainment  area's
                Modeling  Protocol.

                The Modeling Protocol must be  submitted to the appropriate EPA
                Regional  Modeling Contact  for review and  approval.
            2 . 2  Contents of Protocol
                 The recommended contents of the Protocol Document  (Table  1)

            are patterned after those described in  a GARB Technical  Guidance

            Document . 6


                 Recommendations

                 It is recommended that, at a minimum, the components listed in
                 Table   1 be  included  in  the Protocol  Document  for  each
                 attainment demonstration modeling  study.   A  description  of
                 each component is presented in Appendix A.
              Protocol

-------
                    TABLE 1
                                      j
EXAMPLE TABLE OF CONTENTS FOR PROTOCOL DOCUMENT
              •

 UAM Modeling Study Design            '
      Background and Objectives
      Schedules
      Deliverables
      Management Structure/Technical Committees
      Participating Organizations
      Relationship to Regional Modeling Protocols
      Relationship to Other Urban-Area Modeling Protocols
      Relationship to Planning/Strategy Groups

 Domain and Data Base Issues
      Applicable Preprocessor Programs (e.g.,  ROM-UAM
        Interface System)
      Data  Bases:
           •Air quality
           •Meteorology
      Base  Meteorological  Episode Selection
      Modeling Domain
      Horizontal Grid Resolution
      Number of  Vertical Layers
      Emission Inventory
      Specification  of Initial  and Boundary Conditions
      Wind  Field Specification
      Mixing Depths
      Sources of Other Input Data     [

 Diagnostic  Analyses                   :
      Quality Assurance Tests of  Input Components
      Diagnostic Tests of  Base  Case  Simulation
      Test Results/Input Modifications
Model Performance Evaluation
     Performance Evaluation Tests
                                      f
Attainment Demonstrations
     Identification of Attainment-Year Mandated Control
       Measures
     Methodologies  for   Generating  ! Control  Strategy
       Emission Inventories           i
     Procedures for Attainment Demonstration

Submittal Procedures
     Data Analyses Review
     Documentation Review and Approval

-------
Protocol
                              10

-------
                            CHAPTER 3

                   DOMAIN AND DATA BASE ISSUES
  |   Described in this chapter are the following topics:  episode
  i                                              I
selection, domain selection, meteorological data, air quality data,
and emissions inventories.   Choices made in each  topic area are
often interrelated.  Accordingly, decisions concerning a particular
topic area  probably will  be based  on consideration  of several
areas.  In several topic areas, recommendations  are made concerning
minimum requirements for data availability and modeling resolution.
To reduce uncertainties in modeling inputs and outputs, users are
encouraged  to  exceed  these  minimum  recommendations  whenever
possible.                                       i    •
3.1  Episode Selection
     A major component  of the Modeling Protocol  is  selection of
meteorological episodes.  In genercil, episode selection involves a
review (described below) of several multiday periods during which
high ozone was monitored.  At least 1 day is chosen as the day of"
primary interest  for each  selected  episode.    Model simulations
typically begin  at  least one day  prior to  the day  of primary
interest.  This minimizes the effects of assumed  initial conditions
on predicted concentrations for the critical dayL  The length of a
modeled episode is generally a minimum of  48  hours,  and the last
day in this  period-—the  day of primary  interest—ris referred to as
the "primary day."                              j
                                                    Data Issues
                                11

-------
     Episodes that have a  high  probability of covering different
sets of meteorological  conditions corresponding with  high ozone
concentrations should be modeled.  Clearly,  a trained meteorologist
familiar  with  local and  regional  weather  patterns  should  be
consulted  in the  selection  process.    Conditions resulting  in
distinctly different source-receptor configurations should be the
prime  consideration   in  distinguishing  different  meteorological
regimes.  Generally,  conditions reflecting both poorly defined wind
flow (stagnation) and better defined flow (transport) will need to
be included.  It is important to coordinate episode selection with
those  responsible for  Modeling   Protocols  in  nearby  domains,
particularly   when   observed  exceedances   may   result   from
"overwhelming transport."7

     The following,approach  is  recommended for selecting episode
days for use in modeling:

     1.   Identify the meteorological regimes associated with high-
          ozone   episodes.     The  procedure   recommended   for
          identifying  meteorological  regimes  is  described  in
          Appendix B.

     2.   Select  candidate  episode  days  for modeling  from the
          period  from  1987   to  the  present   time.    Place  each
          candidate  episode  day in the appropriate meteorological
          regime  (see Appendix B).

     3.   Rank  each  candidate  episode  day  within each regime
          according  to  the magnitude of the  peak observed ozone
          concentration (ranked highest  are days with the highest
          observed daily maximum  ozone from among all sites in or
          near  a nonattainment  Consolidated Metropolitan Statis-
          tical Area/Metropolitan Statistical  Area  [CMSA/MSA]).
  Data Issues
                                12

-------
                                      I
Select the episode days for modeling from among the three
highest  ranked episode  days from  each meteorological
                     .                 j
regime.   In choosing from among the top-ranked episode
days,  consider  the  availability  and  quality of  air
quality and meteorological data bases; the availability
of supporting regional modeling analyses, the  number of
monitors  recording daily maximum  ozone concentrations
greater than  0.12 ppm (i.e., pervasiveness),  number of
hours for which ozone in excess  of  0.12 ppm is  observed,,
frequency with which  the observed meteorological condi-
tions correspond  with observed exceedances,  and model
performance (discussed in Chapter 5).  For example, the
top-ranked episode day within a meteorological regime may
have only  routine air quality  and meteorological data
bases available  for  use  in  the modeling.   The third-
highest  day,  however,  may   have  occurred   during  an
intensive field study, so that a more comprehensive data
base is available.   Thus,  the third-highest day may be
more desirable for modeling than the top-ranked day.  As
another example,  the  three highest-ranked episode days
may have  air  quality and meteorological  data bases of
similar  quality   and quantity,   and  the   number  of
                                      i
monitoring sites  recording daily maximum ozone greater
than 0.12 ppm may also be similar.  If|model performance
on  the  initially  chosen  day  is  questionable,  the
Technical Committee  may  wish to consider a  second- or
third-ranked day from the three highest-ranked days for
a  regime.    The  day  with  the  overall  best  model
performance may   be  selected as  the  primary day  for
modeling in the attainment demonstration.   Note that a
more comprehensive model  performance  evaluation may be
needed for the selected day,  as  described in Chapter 5.
                                         "Data
                      13

-------
     5.    At least 1 day  should be modeled from  each  identified
          meteorological regime.  Further, a minimum of 3 days from
          among all meteorological regimes  should  be modeled for
          the attainment demonstration  (e.g., three meteorological
          regimes each  containing  1  primary episode day,  or two
          meteorological regimes with at  least 2 primary days from
          one of those  regimes).   Using  the model  results in the
          attainment demonstration is described in Section 6.4-.

     States may want to consider a technique other than  the one
outlined in steps 1-5 for  selecting  modeling  episodes.   Any such
techniques  should be   described  in  the  Modeling Protocol  and
approved by the appropriate EPA Regional Office.

     Consideration   of   several   meteorological   regimes  that
correspond to observed daily maximum ozone levels above  0.12 ppm is
important,  because  certain emission control  strategies  that  are
effective   in  reducing peak  ozone  under  some   meteorological
conditions  may be less  so under others.   The  goal is  to  develop
strategies that are robust with respect to  effectiveness over most
scenarios.

     Recommendations
     It is recommended that episodes for modeling be selected from
     the period from 1987  to the present time.  Selected  episodes
     should represent different meteorological regimes  observed to
     correspond with ozone >0.12 ppm (as described above).  When
     selecting episodes, both  stagnation and transport conditions
     should be examined.   A  minimum  of  3 primary episode days
     should be simulated.
     Primary episode days  falling within each meteorological regime
     are  ranked according to  the  highest  observed daily maximum
     ozone  concentration measured within or near the nonattainment
     CMSA/MSA.  Episodes may be chosen to include any of the  three
     top-ranked  days in each regime.  In  addition to  considering
     the  magnitude of  the highest  observed  daily maximum  ozone
  Data Issues
                                14

-------
      concentration  in making this  choice,  data availability and
      quality, model performance, availability of regional modeling
      analyses,   pervasiveness,   frequency  with  which  observed
      meteorological  conditions  coincide  with  exceedances,  and
      duration of observations >0.12 ppm may be  considered.
                                                I            ' .
      Other techniques for selecting episodes should be described  in
      the  Protocol Document  and  approved by  the appropriate EPA
      Regional Office.                           i
3-2  Size of the Modeling Domain

     The size and location of the modeling domain define the data
requirements for  the modeling.   In  selecting a modeling domain,
consideration should be given to (l) typical wind patterns,  (2) the
location of major area and point emission sources, (3)  the location
of air quality monitors and  important receptor  locations, and
(4) the need to mitigate effects of uncertainty  in upwind boundary
conditions.   Generally,  the domain  should be  set as  large  as
feasible  in order  to  reduce  the dependence of predictions  on
uncertain boundary  concentrations  and to  provide  flexibility in
simulating different meteorological episodes.  It is  generally much
easier to subsequently reduce the size of a modeled area than it is
to subsequently increase it.                    !
     Once UAM input data for a sufficiently large domain have been
assimilated and processed, the size of the modeling.domain can be
reduced for modeling purposes by specifying domain boundary values
in the UAM.   Procedures for reducing the size jof  the domain are
described in Reference 2.  This could save resources in simulating
modeling episodes  in which  light  or  poorly defined  wind fields
 .
result in a smaller domain being adequate.  In contrast, expanding
domain dimensions  would require reconstructing  most of  the UAM
input files.
                                                   Data Issues
                                15

-------
    Recommendations
    It  is recommended that  the domain's downwind boundaries be
    sufficiently far from the CMSA/MSA that is the principal focus
    of  the  modeling  study  to  ensure that emissions  from the
    CMSA/MSA occurring  on  the primary  day for  each  selected
    episode remain within the domain until 8:00 p.m.  on that day.
    The extent  of the  upwind boundaries  will depend  on the
    proximity  of large upwind  source areas and the adequacy of
    techniques7   used   to   characterize   incoming   precursor
    concentrations.  Large upwind emission source areas should be
    included in the modeling  domain to the extent  practicable.
    Also,  if large uncertainty  is anticipated for domain boundary
    conditions,  the upwind boundaries  should  be  located at  a
    distance sufficient to minimize boundary effects on the model
    predictions in the center of the domain.  Sensitivity analyses
    described  in Section 4.3 assist in determining the effects  of
    boundary conditions  on predicted values.
3.3  Horizontal Grid Cell Size


     The horizontal dimension of each model  grid square is based

upon (1) the sensitivity of predicted concentrations to horizontal

grid size,  (2) the resolution of  observed meteorological and air
quality data and/or estimated emissions data, and (3) limitations

imposed by other considerations such as a required minimum domain
size.  Generally,  large grid  square dimensions result in smoothing

of  the emission  gradients,  wind  fields,  and  spatially varying
mixing heights, which in turn leads to a smoothing of the predicted

concentration field.   Also, larger grid cell dimensions reduce both

computer storage space and computational time.


     The  following   should  be  considered  when  selecting  the

horizontal  grid cell  size:
     1.
The grid cells should be small enough to reflect emission
gradients  and  densities  in urban  areas,  particularly

those  resulting  from large  point  sources and  major
terrain or water features that may affect air flow
  Data issues
                                16

-------
     2.    Sensitivity studies conducted  by the EPA suggest  that
          peak  ozone  predictions  may  increase  as   grid  size
          decreases                             j
                                                1
     3.    Practical  limitations  on the  grid cell  size are  the
          resolution of the emission inventories eind the density of
          meteorological and air quality monitoring networks
  i
     Previous modeling studies have used horizontal grid cell sizes
generally in the range of 2 x 2 km to 8 x 8 km.  A grid size of
5 ;x  5 km  has generally been  compatible with  computer  resource
requirements and emission inventory development.

  1   Recommendations                            j
     It is recommended that the size of the horizontal grid cells
  i   should not be greater than 5x5 km.  Grid  cell sizes coarser
     than  this should  be justified and should,  at  a  minimum,
  1   address  items  1-3 above.    Smaller  grid  cell sizes  are
  i   encouraged because they allow more accurate gridding of area
  :   and mobile sources.  Additionally,  emissions  from major point
     sources are better characterized by smaller grid cell sizes.
     However,  grid cell  sizes smaller  than  2 x  2  km  are  not
  !   recommended   because   of   potential   model   formulation
     inconsistencies  for those grid  sizes.      J
3.4
            of Vertical Layers
  !    In specifying the number of vertical layers]  issues  analogous
to  those  raised for  horizontal  grid  cell  dimensions  must  be
addressed.   Again,  a compromise  is generally needed between  the
number of vertical layers and the adequacy of available data bases
and computer resources.  It  is important that sources  with tall
stacks or  sources having plumes with high buoyancy be assigned to
an  appropriate altitude. Pollutants in elevated,  buoyant,  point-
source plumes often  have effective release heights in layers well
above the  surface.    Increased vertical  resolution allows  more
                                                    Data
                                17

-------
accurate  representation  of the  vertical  layer  at  which these
emissions  interact with emissions  occurring closer to the ground.
Also,  increased vertical resolution  minimizes  dilution that may
result from placing  emissions into  artificially  large vertical
layers.    Finally,  increased  vertical  resolution  improves  the
simulation of when and where  plumes are mixed  to ground level.
Simulation of  the chemistry  between individual plumes  and the
environment can be greatly affected by how well the model simulates
mixing of  these plumes  with the ambient air.

     Previous  applications  of  the  UAM have  generally used four or
five vertical  layers, with  two layers between the surface and the
morning mixing height (diffusion break in the UAM) and three layers
between  the mixing height  and the  top of the  modeling domain.
Sensitivity studies suggest that  using fewer than  three layers
above  the  mixing height may artificially  dilute elevated point-
source plumes, which may  cause  the  model  to underpredict near-
surface ozone  and  precursor concentrations.

     Users  of the UAM should consider specifying greater detail for
the  horizontal  and vertical  grid  cell  size  than the  minimum
recommended in  this  guidance document.     This  is  encouraged
particularly  in modeling domains  containing complex terrain  or
land/water  interfaces.   Wind  field  models  can  typically produce
wind fields for many more vertical layers than the minimum number
given  here.3  The number of vertical  layers considered in the UAM
is more likely to be constrained by the time-consuming calculations
needed to simulate atmospheric chemistry.
     Recommendations
     Based on previous model applications, it is recommended that
     a  minimum  of five vertical  layers be used  in  the modeling
     study,  with  at least three layers  above  the morning mixing
     height  (diffusion  break in the  UAM).   Additionally,  it is
  Data Issues
                                18

-------
     recommended that the top of the modeling domain (region top in
     the UAM) be specified above the mixing height  by at least the
     depth of one upper-layer cell.   This can be done by setting
     the region top value  equal  to  the maximum mixing depth plus
     the minimum depth of the upper-layer cells.

     Previous applications have typically used 50 m for the minimum
     depth of the vertical  layers below the diffusion break and 100
     or 150 m for the vertical  layers above the diffusion break.
     It is recommended that 50 m be used as the minimum thickness
     for layers below the diffusion  break  and 100 m as the minimum
     thickness for layers above the diffusion break.
3l5  Meteorological Data
 '    The availability of meteorological  data varies widely among

prospective  modeling domains.    Also,   there  are  a variety  of

techniques  available  for   developing wind  fields,  temperature

fields,  and  mixing  heights.    Although  high  resolution  and

confidence  for  all meteorological  data  are desirable,  time and

resource  constraints force  a compromise  between  desirable and

acceptable  methods.   Historically, measured meteorological data

have been interpolated for most UAM applications.  More recently,

diagnostic and prognostic meteorological modeling techniques have

been   explored   as  possible  mecins  to  develop  input  fields

(particularly wind fields)  for air quality models.
 ,    Wind fields and mixing heights are two of fche.most important
 (
meteorological inputs that significantly affect photochemical model

predictions.   Methodologies and recommendations  for determining

these inputs are described below.
     3.5.1  Wind fields
 !         Methodologies  to construct  wind  fields  for the  UAM

applications have historically fallen into three categories:6
                                                         Issues
                                19

-------
     1.    Objective analyses that  interpolate observed surface and
          aloft data throughout the modeling domain

     2.    Diagnostic wind models in which physical constraints are
          used in conjunction with objective analyses to determine
          the wind field

     3.    Prognostic models  based on  numerical solution  of  the
          governing  equations  for mass,  momentum,  energy,  and
          moisture conservation along with  numerical solutions for
          thermodynamic processes

     More recently, an additional methodology has been developed in
areas where the EPA ROM  has  been  applied.   Computer software has
been developed to  map a  ROM  diagnostic gridded wind field into a
nested UAM domain.5

     Objective  analysis  -  These procedures  generally  involve
straightforward interpolative techniques.   They have the advantage
of being  relatively simple and inexpensive to  use.   The primary
disadvantages  are that  these analyses contain  limited physical
concepts, and  results are highly  dependent upon the temporal and
spatial  resolution  of  the  observed values.   Thus,  in domains
containing sparse observational data or complex topography, results
may be unsatisfactory.

     Diagnostic wind models - These models improve mass consistency
for the flow fields.  This may  be  addressed through pararaeteriza-
tions for terrain blocking effects  and upslope and downslope flows,
as in the UAM Diagnostic Wind Model.3  Diagnostic models generally
require  minimal  computer  resources  and  can  produce  a three-
dimensional  wind  field.     However,   diagnostic  models  need
  Data Issues
                                20

-------
representative observational data to generate fesitures such as land
and sea breezes.
  I   Prognostic models - These models simulate relevant atmospheric
physical  processes while  requiring minimal  observational data.
Prognostic  models  require a specification  of the synoptic-scale
flow.   Reliability of these approaches  is  usually  enhanced if
sufficient observations are available to "nudge" solutions closer
to  observations.    Since  these  models can  simulate  temperature
fields in addition to the wind field, it is possible to determine
stabilities  and mixing heights,  thus eliminating  the  need to
generate these from sparse observational data. Another significant
advantage  is  that interdependencies  of various  meteorological
inputs with  one  another are considered in  prognostic models.  A
major disadvantage is  the  extensive computational resources needed
to  run  a prognostic  model.  Additionally,  the  availability of
evaluated models and  expertise  needed to apply  them  for general
application with photochemical grid models is limited.

  ;   The ROM-UAM Interface System - This system can develop a UAM
gtidded wind field  from a diagnostically derived wind field used in
  i
the ROM.  Such  a  ROM-derived wind field can be applied  for a UAM
domain that  is nested  within a ROM domain,  provided  ROM data are
available for identical episode periods.  Use of ROM data has the
advantage  of  being  easy  to  implement  and  also  provides  a
  1   .                                          i
consistency  between  ROM model  predictions  used  to  specify  UAM
boundary conditions and the corresponding wind j fields.   The ROM
data are based on an approximately  18 x 18 km  horizontal grid cell
size.  Thus, one disadvantage is that ROM gridded wind fields may
not  sufficiently  describe  detailed  features such  as  land/sea
                                                    Sata Issues
                                21

-------
circulations.   A more finely resolved wind field may be obtained

by using the ROM-gridded winds  as  the initial wind field for the

UAM's Diagnostic Wind Model (DWM) preprocessor (see Reference 3).

This provides a means for mass consistency when using" the ROM data

as boundary conditions in conjunction with another wind model.


     The  selection of  a specific  technique for  generating the

domain wind field depends largely on (1) availability of concurrent
ROM diagnostic wind fields,  (2) the spatial and temporal resolution

of  surface and  upper-air   observations,   (3)  available  modeling
expertise  in applying alternative  meteorological models, and (4)

available  computer  resources.    However,  some  guidelines  on

preferences for generating the wind fields are as follows.


     Recommendations

     The ROM-UAM Interface System should be used  to derive the UAM
     gridded wind fields when the UAM domain is nested within  a ROM
     domain for  concurrent time periods  and ROM predictions are
     used to derive the hourly UAM boundary  conditions.  If  it is
     judged  by the  Technical Committee  (and identified  in the
     Protocol) that a wind field derived  from the UAM DWM is more
     representative of the  domain-scale flow, then this wind  field
     may be  used in  lieu  of  the ROM diagnostic  wind  field.  To
     minimize mass  inconsistency problems, the ROM-gridded winds
     may  be used as  the   initial  wind  field  in  the  DWM  (see
     Reference 3) when generating the UAM gridded wind field.

     For   cases   in   which  concurrent   ROM   applications   are
     unavailable,  it  is  recommended  that  the  DWM  be used  to
     generate  the UAM  gridded  wind  fields.   The use  of  other
     techniques  for deriving the wind  field,  such as prognostic
     wind models or other objective techniques, may be employed on
     a case-by-case basis,  subject to approval from the appropriate
     EPA Regional Office.
     *The ROM data for use in the ROM-UAM Interface System can be
accessed through the EPA UAM Subsystem of the Gridded Model
Information Support System  (GMISS) .
  Data. Issues
                                22

-------
  j   3.5.2  Data needs for wind field development:


     The development  of a wind  field for each  modeling episode

depends upon ground-level and elevated wind observation data.  It

is!preferred that a surface-based monitoring network report wind

speed and direction as hourly averages, because an. hour is the time

period commensurate with most  UAM concentration output analyses.

Th4 surface monitoring  network should be  broad and dense so that
diagnostic models  (if that is the technique  chosen)  can depict

major  features  of the  wind field.,   Data representing vertical

profiles  of  wind speed and direction  are  required  in  order to

establish upper-level  wind fields.  Preferably, data should provide

adequate spatial (horizontal)  and temporal resolution.  Results of

UAM  applications  are  often   criticized  because of  inadequate

meteorological  data,  and lack of  sufficient meteorological data

often prevents definitive diagnostic analyses. Thus,  the need for

adequate meteorological data cannot be overstated.
  i   Time  constraints  imposed  by  the 1990  CAAA  will  probably

preclude consideration of new meteorological monitoring stations.
Thus, it is likely that the base case to be used in  the attainment

demonstration will be from an historical  episode for which model

performance has been  deemed acceptable.


  ;   Recommendations

  j   Meteorological data routinely available for  a UAM modeling
  ,   demonstration  usually consist  of National  Weather Service
     (NWS)  hourly surface and  upper-air  observations (for winds
  i   aloft).  If these data are the only data available for use in
  |   a modeling demonstration, they may have  to suffice.  However,
     the  NWS data consist  of observations made  over very short
  I   periods  rather  than hourly averaged values.   An assumption
     that wind velocity measured over a very  short period persists
     unaltered  over  an  hour  may lead  to  an   overestimate of
     transport.   Therefore,  whenever  possible,  hourly averaged
     meteorological  data  (e.g.,  from an intensive field study)
                                                         Issues
                                23

-------
      should  be  used.   Additional  meteorological  data  may be
      available from  other  sources  in the  domain  (e.g.,  an  on-site
      meteorological monitoring program at an industrial  facility).
      These data  may  be used to supplement the NWS data, provided
      the data have been adequately quality assured.  Additionally,
      the  EPA guideline  entitled  On-Site Meteorological  Program
      guidance	for  Regulatory  Modeling  Application"should be
      consulted to assess whether  the  supplementary data  reflect
      proper  siting of meteorological instruments and appropriate
      data reduction procedures.

      In planning a special field study to provide a more spatially
      and temporally dense meteorological  data base, the number of
      surface   meteorological  monitoring  stations   should  be
      sufficient  to  describe  the predominant wind  flow features
      within  the  modeling  domain.   An  experienced meteorologist
      familiar  with  local  climatic patterns should  be consulted
      concerning  the  location  and  suitability   of   the   surface
      meteorological stations.  Vertical sounders  or profilers are
      highly encouraged in  a special  field study to resolve winds
      aloft  and  mixing heights.   Any special  field  study and
      monitoring program should be planned in consultation with the
      appropriate EPA Regional Office before implementing the study.


      3.5.3  Mixing heights


      Predictions from the  UAM have been shown to be sensitive to

the mixing height field.6   Therefore,  the temporal variations in

the mixing height field over the  UAM domain should be described as

realistically as  possible.   The UAM modeling system contains a

methodology for  deriving mixing heights  (diffusion  break  in the

UAM) based on surface temperatures, vertical  sounding measurements

of  temperature,  and  cloud  cover  (see Reference 2).   However,

because of the diversity of techniques and data bases that may be

available on a case-by-case basis, we cannot recommend  a specific

procedure for deriving the mixing height  field in all cases.
     Recommendations

     It  is  recommended  that,  at  a  minimum,  the  techniques
     described in Reference  2  be  used in establishing the mixing
     height field in the domain.
-Data
                                24

-------
     The choice of  upper-air stations to  be  used in the  mixing
     height calculations should be based on prevailing wind fields
     and location of the upper-air  stations relative to the  UAM
     domain. If there are no upper-air stations within the domain,
     stations outside the domain may need to be  used.  A  trained
     meteorologist should be consulted on the selection of upper-
     air stations for use in determining mixing heights.

     The techniques for generating the mixing height field should
     be described in the Protocol Document.  Techniques other than
     that  described in Reference  2  should be  documented  and
     justified.


     3.5.4  Clear-sky assumption for photolysis rate calculations
  I   For regulatory  UAM applications, clear-sky  conditions have

typically been  assumed for  photolysis  rate calculations  in the

MEJTSCALARS processor.  The UAM's current structure does not allow

for spatial  variation in  cloud cover,  so  the choice  is  either

uniformly clear  or a uniform  cloud  cover based on  a mean cloud

cover over the domain.  Use of mean cloud  cover could significantly
understate reaction  activity .in  "clear" patches  of  the domain.

Potentially,  this could be  a  more  serious error than assuming

cl'ear-sky conditions and simulating an overall excess of "domain-

wi,de"  insolation.   Additionally, the  ROM-UAM  Interface  System
IMETSCL processor assumes clear-sky conditions  for  photolysis rate

calculations.
     Recomroendat i ons
     For applications  involving the current regulatory version of
     the  UAM,  it  is  recommended that clear-sky  conditions be
     assumed  for  calculating photolytic  rate constants  in the
     METSCALARS processor.
                                                    Data Issues
                                25

-------
3.6  Air Quality

     Ambient air quality data are generally used for two purposes:
(1) to specify initial- and boundary-condition concentrations, and
(2) to assess the model performance for each meteorological episode
to  be used  in  the attainment  demonstration.   These  topics are
addressed in the following  two  subsections.

     3.6.1  Initial and boundary conditions

      Three general approaches  for specifying boundary conditions
for UAM simulations are as  follows:   (l) use  objective/interpola-
tion techniques with  a  sufficient amount of measured data  (i.e.,
data from an intensive field program),  (2) use default background
values and expand the upwind modeling domain and  simulation period
to mitigate uncertainties due to paucity of measurements, and (3)
use  regional-scale  model  predictions  of ozone   and  precursor
concentrations.  Initial conditions for UAM simulations are handled
in one of two ways:   (1)  use regional-scale model predictions to
derive initial  conditions, and/or  (2)  begin  the  UAM  simulation
sufficiently  far  in  advance  of the primary day  to  eliminate
sensitivity of results to arbitrary assumptions regarding initial
conditions.

     Clearly,  the  nature  of  case-specific  applications  will
determine what approaches  should be taken for  establishing initial
and boundary  conditions  for  particular domains.    Ideally,  the
preferred technique would be based on an intensive field program
with regional-scale modeling used to fill in spatial and temporal
gaps.   This approach is  seldom feasible, however, particularly for
historical  episodes.    Presented below  are  recommendations  for
implementing each  of the three techniques identified above  for
deriving  boundary   conditions,  including   discussion   of  the
  Data Issues
                                26

-------
advantages and disadvantages of  each technique.  !Default boundary-
condition values for ozone  and  precursor concentrations are also
provided.  Finally, recommendations are  provided on the approach
most likely to be feasible for specifying the initial and boundary
conditions for modeling historical episodes in most locations.
  !                                •              j
  I                     '            '      '       :       .'     '   "
  i   Use of measured data - All sources of air quality data for a
  s
particular modeling domain should be evaluated for applicability in
establishing initial and boundary condition's.  Unfortunately, most
ongoing monitoring programs  have been designed (xinderstandably so)
with a receptor-based orientation.  While available monitoring data
are useful for evaluating model performance, they usually are not
adequate for establishing initial and boundary concentrations.
  !                                              I
  !   Recommendations
  i   To develop initial and  boundary conditions,  it  is recommended
  !   that one  or more  monitoring stations be sited upwind of the
     central  urban area along  prevailing  wind  trajectories that
  j   give rise to  ozone exceedances.
  !   The  sampling and  analysis program  should provide  data to
  !   calculate  hourly  values for ozone,  NO,  NO2,  and speciated
     hydrocarbons.
  !                                 .
  j                                              ,| •
     At the inflow boundaries,  air quality data  at the surface and
  I   aloft  should  be   used whenever  available  to specify  the
  !   boundary  conditions.   Default values  (Table  2)  may be used
  j   where necessary.
  i                                              !'
  i                                              i^
     Use  of  default values  - Some urban areas  may lack adequate
data  suitable for establishing initial  and boundary conditions.
Section  3.2  on domain  selection and  Chapter  4  on  diagnostic
analyses  recommend constructing  domains  and  simulation periods
large enough to minimize the sensitivity of inner core and downwind
concentrations to  assumed initial and boundary conditions.
                                                    Data- Issues
                                27

-------
                            TABLE 2


         DEFAULT BOUNDARY CONDITION CONCENTRATIONS FOR
        CARBON-BOND-IV SPECIES (SEE REFERENCES 1 AND 7)
    Species

    OLE
    PAR
    TOL
    XYL
    FORM
    ALD2
    ETH
    CRES
    MGLY
    OPEN

    PNA
    NXOY

    PAN
    MONO
    H202
    HN03
    MEOH
    ETOH
    03
    N02
    NO
    CO
    I SOP
Species Name
Concentration fppbC)*
Olefins
Paraffins
Toluene
Xylene
Formaldehyde
Higher Aldehydes
Ethene
Cresol,Higher Phenols
Methyl Glyoxal
Aromatic ring fragment
   acid
Peroxynitric acid
Total nitrogen
   compounds
Peroxyacyl nitrate
Nitrous acid
Hydrogen peroxide
Nitric acid
Methanol
Ethanol
Ozone
Nitrogen Dioxide
Nitric oxide
Carbon monoxide
Isoprene
       0.60
      14.94
       1.26
       0.78
       2.1
       1.11
       1.02
       0.01
       0.01
       0.01

       0.01
       0.01

       0.01
       0.01
       0.01
       0.01
       0,
       0,
      40.0 (ppb)
       2.0 (ppb)
       0.0 (ppb)
     350.0 (ppb)
       0.1 (ppb)
    ppbC, parts per billion Carbon
Data Issues
                              28

-------
  I   Initial- and boundary-condition concentrations are influenced

by large- and small-scale  weather  patterns and emissions distri-

butions that  are unique to each modeling domain.    Thus,  case-

specific attributes should be used in estimating these concentra-

tions whenever feasible.  For example, boundary concentrations of

hydrocarbons, particularly  those species (or intermediate products)

emitted from vegetation, are likely to be  higher in urban areas
  !                                              :
surrounded by dense vegetation than in areas surrounded by sparse

vegetation.
     Recommendations          '       •           !

     It is  recommended that  use  of default values  to establish
     boundary conditions be  limited to  areas surrounded by large
     expanses of low-density anthropogenic emissions.  Accordingly,
     the modeling domain may need to envelop rural areas.

     Those choosing to use default  values should plan to perform
     diagnostic/sensitivity simulations  (see Chapter 4) to evaluate
     the sensitivity of domain-interior model  predictions to the
     boundary conditions.
                                                i
     Table 2 lists the  recommended default boundary values for the
     chemical species used in the model.  Use of default boundary
     values under regional transport conditions should be closely
     evaluated.   When  using default values, the  boundary of the
     domain should extend as far upwind as practicable.

     To diminish dependence on arbitrary specification of initial
     conditions, a simulation should begin at least 1  day prior to
     the primary day.
  i   Use of regional model concentration predictions - Output from

regional-scale models  such as the EPA  ROM  provides  estimates of

initial and boundary conditions (as well as certain meteorological

inputs) for urban-scale models.  This is especially important under

regional  transport  conditions.    The  ROM-UAM Interface  System
  i                                              '
referred to in Section 3.5.1 can use ROM concentration predictions

to  develop UAM input  files of initial and  boundary conditions.

This interfacing  software should be used for  UAM domains nested
                                                    Data
                                29

-------
within  more  extensive  ROM  domains.     Using  the  ROM  is  the
recommended approach for generating boundary conditions. It is the
most technically defensible approach for estimating future boundary
conditions for the attainment year.

     Certain considerations arise when using interfacing methods.
First, selection of  historical  episodes is limited  to those that
have  been  modeled on  a regional  scale.   Second,  there may  be
inconsistencies  in mass conservation  when  applying  ROM-derived
initial and boundary  conditions in conjunction with wind fields not
derived from the ROM wind field (see Section 3.5.1).  The
combinations of concentrations and wind velocities produced by the
ROM-UAM Interface System represent mass fluxes passing through the
urban-scale modeling domain.   In  cases where ROM-derived initial
and  boundary  conditions are  applied without ROM-generated  wind
fields, locally developed wind fields may need to be evaluated for
mass consistency throughout the urban-scale  domain.   Methods for
addressing this problem  will  need to be chosen  on a case-by-case
basis.   A  general  procedure for  enhancing mass consistency  is
described in  Section 3.5.1.   Additionally,  initial  and boundary
conditions  derived from the ROM  data  should  be  compared  with
corresponding monitoring data wherever available.  This will ensure
that  the  ROM  wind fields  adequately represent the  transport  of
ozone and precursors into the domain region.

     Recommendations
     It  is recommended that,  whenever  feasible,  the  ROM-UAM
     Interface System be applied to derive the initial and boundary
     conditions for the episode(s) being modeled.  If the Interface
     System is used to  derive  the initial and boundary conditions,
     it is  also recommended  that  it be used  to derive  the UAM
     gridded wind field, unless there is sufficient justification
     that other  techniques for deriving the wind field  are  more
     accurate.
  Data Issues
                                30

-------
  I   In cases for which ROM predictions are  not  available,  it is
  j   generally recommended that measured data be used to establish
  !   the initial and boundary  conditions,  provided the Technical
  I   Committee identified in the Protocol  determines the data are
     adequate.   If  measured data  are  not adequate,  the default
  i   values may be  used.   To diminish sensitivity  of results to
  i   assumed  initial conditions,  simulations should  begin  1 day
     prior to each primary day.
  ,
  i                                              ...
3.6.2  Performance Evaluation Data              !
                                                    .

     Air quality data are needed to diagnose problems in setting up

model  applications  and  assessing  model  performance  for  the
meteorological  episodes   being   considered   inj  the  attainment

demonstration.    A  lean  air  quality data  base   may introduce
significant uncertainties in characterizing model performance.
  I   Under Title  I,  Section 182 of' the CAAA of  1990,  the EPA is

required to develop  regulations for  enhanced monitoring of ozone

and  ozone  precursors  in  serious,   severe,  and  extreme  ozone

nonattainment  areas.   When promulgated,  these  regulations will

specify criteria  for network design,  monitor  siting,  monitoring

methods,   operating   schedule,   quality   assurance,   and  data

submittal.11  The  enhanced ozone monitoring system is designed to
                                                •I
provide a  more comprehensive  data base  for model  input and to

improve model performance evaluation.


     Recommendations                            [

     It is recommended that the data base used in the attainment
     demonstration   modeling  meet   the   requirements  for  the
     enhanced ozone monitoring system to be promulgated by the EPA.
     However, the EPA recognizes that some historical episodes that
     will be used in the attainment demonstration modeling for the
     November  1994 ozone SIP submittals may have data bases that
     would  not  meet  the  requirements  for an enhanced ozone
     monitoring system.   Under these conditions,  the data bases
     should be scrutinized in detail  by the Technical Committee to
     help  ensure  that  model  performance  tha^.  appears  to  be
                                                    Data Issues
                                31

-------
      acceptable has not actually resulted from compensating errors
      in the data  bases.   Additional  diagnostic analyses may  be
      necessary for lean  data bases from historical  episodes.
      If it is  determined that the  existing air quality monitoring
      program does  not meet the requirements for the  enhanced ozone
      monitoring  system,  responsible  regulatory agencies  should
      begin planning for development of  an enhanced ozone monitoring
      system for potential future modeling  studies.

 3.7   Emissions

      The  credibility  of UAM applications  is  directly  tied  to
 formulating  the best possible emission inputs.   Model  performance
 may  hinge on  how well  emissions  are  estimated.   Also,  in the
 attainment demonstration,  modeling results are  used to determine
 emission  scenarios  that  lead, to improved  air quality  levels
 consistent with the NAAQS.  A faulty emission inventory could  lead
 to erroneous conclusions  about the extent  of  needed controls  and,
 in some cases,  errors in  judgment about the need to control  certain
 classes of precursors  (e.g.,  NOX).

      Developing photochemical model emission input data is  the  most
 intensive task of model applications, and requires consideration  of
many  issues.   The source of  the UAM modeling emission inventory
will  be  the 1990  SIP nonattainment  base year inventory required
under  the CAAA of 1990  for all   ozone  nonattainment areas.   A
further discussion of the 1990 base year inventory is contained  in
Emission—Inventory Requirements for Ozone  State Implementation
Flaps.12  It is important to note that the 1990 modeling inventory
will  not  be identical  to the 1990  nonattainment area inventory
required  for reasonable  further   progress  (RFP) tracking under
Section 182 of  the CAAA.   For example,  the modeling inventory will
probably  have  to  cover  a  larger geographical  area than  that
required for the nonattainment area inventory.  The discussion of
  Data  Issues
                                32

-------
modeling domain and boundary-condition issues in Sections 3.2 and
3.6 makes  it  clear that the modeling  inventory must encompass a
larger area than the nonattainment MSA.  A complete description of
relationships between the modeling inventory and the nonattainment
area inventory  is provided in Procedures for the Preparation of
Emission  Inventories for Volatile Organic  Compounds.  Volume II:
Emission  Inventory  Reauir-ements  for  Photochemical—Air Quality
simulation Models  (Revised").13  Additional guidance for developing
the modeling emission inventory  is found in Reference 4.
   |                          _                    |     ...
   i                  '                            i
     For  use in  regulatory applications   of  the UAM,  the  1990
modeling  inventory  will  have   to undergo  several  adjustments.
First,  the inventory needs to be adjusted  to be consistent  with
meteorological conditions during each selected episode (i.e., "1990
day-specific emissions").  Second, the resulting "1990 day-specific
emissions"  should  be  adjusted  to reflect  control  programs  and
activity levels prevailing during the year(s) of selected episodes.
For example, if a selected episode occurred  in 1988, the "1990 day-
spjecif ic  emissions" would be further adjusted to  reflect controls
and activity levels prevailing in 1988.  This latter adjustment is
needed  to  provide  an  estimate  of  emissions  most  suitable  for
evaluating performance  of the UAM.

     As noted in  Chapter  1, once the UAM's  performance has been
evaluated  and   the  model  has been  determined  to  perform
satisfactorily,  it is used to derive control strategies to attain
the  NAAQS.   This requires another  adjustment to the  "1990 day-
 specific emissions" described  above.  This  adjustment entails use
 of growth factors,  ongoing control  programs and  retirement rates
 for obsolete sources  of emissions to  project  "1990 day-specific
 emissions" to the  years by which the  CAAA  specify that the NAAQS
 must  be  attained.    Reference  14  describes  the  appropriate
 methodology  for  making  emission  projections.I   The  resulting
                                 33

-------
 "attainment year modeling inventory" is used as a  starting point
 from which to  construct "strategy  inventories."    A  "strategy
 inventory" is obtained by superimposing additional control measures
 on   sources  of  emissions  in  the  "attainment   year  modeling
 inventory."

      In summary, a 1990  modeling  inventory is first adjusted to
 evaluate UAM performance.   The 1990 modeling  inventory is then
 readjusted to  reflect emissions most likely at the time the CAAA
 require attainment of the NAAQS.

      Two emission files drive the UAM, a file of emissions that are
 injected into  the first, surface-based  vertical  layer,  and an
 elevated point  source  file  of emissions that  are  injected into
 vertical layers above  ground  level.   The  UAM Emissions Prepro-
 cessing System  (EPS)4 reads county-level  area- and point-source
 files and performs four major functions: (1) area sources  and point
 sources are allocated  to grid cells; (2)  temporal profiles are
 assigned to source categories;  (3)  hydrocarbon speciation profiles
 are  assigned to source  categories,  and  (4) point sources  with
 effective  plume  heights greater than a prescribed cutoff  level are
 assigned to the  elevated point source file and the remaining point
 sources are assigned  to the  surface-layer emissions file.

      Addressed   below  are the following  issues that   arise  in
 developing emission input data:   (1)  use of speciation profiles,
 (2) use of surrogate factors  to grid area sources,  (3) treatment of
mobile  sources  and  top/down  versus bottom/up approaches,  (4)
 episodic adjustment of inventories to day-specific modeling inputs,
 (5)  treatment  of  biogenic emissions,  (6)  cutoff  levels for NOX
 point sources, and  (7)  consistency with national inventories.
  Data Issues
                                34

-------
     3.7.1  VOC speciation
  i              ,                                i
  I   The EPA provides "default" nationwide VOC speciation profiles
for various source category codes (SCCs).13  Use of local speciation
information,  especially  for  major  emitters,   is  preferable  to
national default profiles. If feasible, major VOC point- and area-
source categories should be surveyed to determine appropriate VOC
composition profiles.   In many cases, both  the quantity  and the
composition of emissions change as process operations are modified.
To,the extent feasible,  this should be accounted for when deriving
local speciation,profiles and in simulating control  strategies.
The  emissions  inventory  guidance document" provides  details  on
developing local speciation profiles.           |

     Most current-year applications  are likely to rely on existing
default data  for  speciating mobile-source  emissions.   Projected
future-year mobile-source emissions files may be  based on different
formulations of gasoline and use  of alternative  fuels.  Speciation
guidance  for  these  fuels will be provided by  the EPA Office  of
Mobile Sources (OMS) through the appropriate EPA Regional Office.
  I
  '   Recommendations                            I
  ;   It is  recommended  that local speciation profiles for point-
  I   source and area-source categories be used whenever feasible.
     The Technical Committee should determine the appropriateness
  I   of  using  local or  national  default  speciation  profiles.
  !   Profiles  used   in  the  modeling  demonstration  must  be
  !   documented, and any changes  assumed in profiles as the result
  !   of control strategies must be identified and justified.
     3.7.2
aroa sources
  i   Area-source emission data,  including motor vehicle emission
data, are often supplied on a county basis.  Spatial allocation of
coynty-level emission estimates to grid cells is performed for each
                                35

-------
identified  area-source category; the  allocation  requires use of
surrogate  distribution factors  such  as population distribution,
land use, and road type.  The UAM EPS4 contains a program that uses
gridded surrogate factors to allocate  county-level emissions data
to the grid cell size  of the modeling  domain.

     Recommendations
     It  is  recommended  that  the emission  inventory guidance
     document13  be  consulted  for  alternative surrogate  factor-
     choices and sources of information for assimilating surrogate
     data.  The EPA is currently developing a utility  to provide
     gridded  surrogate data.    States will  be notified  of the
     availability  of  gridded  surrogate  data through the EPA
     Regional Offices.

     3.7.3  Mobile sources

     Development  of gridded,  time-variant  mobile-source  inputs
raises several concerns and  often represents the largest fraction
of  effort  when  assimilating  mobile-source  emissions  inputs.
Mobile-source emissions have been  compiled from original data or
from existing county-level emissions.13  Developing gridded mobile-
source emissions from original data  requires aggregating sub-grid-
cell-level components.  This may require exercising transportation
models that produce  inputs  for the mobile-source  emissions model
(i.e., the latest  EPA MOBILE  model), and then performing the
necessary  spatial   allocations  to   grid   cells   and  temporal
distribution  over  every  hour.    This  practice  is  far  from
standardized.  Also, in certain areas,  execution of transportation
models is restricted by lack of appropriate traffic count and speed
data.

     The  emission inventory guidance document13 provides direction
for developing mobile-source inputs from original data  (referred to
  Data Issues
                                36

-------
as a bottom/up method) or from an existing county-level inventory

(referred to as a top/down method).
     Recommendations

     Bottom/up methods are the  preferred approach for estimating
     vehicle activity  levels  and emission  factors  because these
     methods have potential for resolving variations in speed and
     vehicle miles traveled (VMT) among different grids over hourly
     time slices.  Bottom/up  approaches  are most appropriate for
     addressing  the   inner   urban   core  of  modeling  domains.
     Peripheral,  less  dense  traffic areas  can be  treated  with
     top/down methods.  Exceptions to these recommendations should
     be considered by  the Technical Committee  on a case-by-case
     basis.   Justification for more  extensive use of  top/down
     methods should be sought in discussions with the appropriate
     EPA Regional Office.
     3.7.4  Episode-specific adjustments
  i                                              j
  i                            •                  i

  [   Several source categories  of  VOC emissions are sensitive to

meteorological conditions.   Thus,  it is important  for modeling

inventories to reflect episode-specific meteorological conditions.^

For  example,   biogenic  emissions,   mobile-source  evaporative

emissions,  and solvent categories will  need  to reflect specific

modeling days.   In addition,  known  episode-specific events such as

changes in  process operations  for  point sources affect emissions

rates and should be reflected in the episode modeling inventory.
     Recommendations

     Mobile-source  emissions  should  be  adjusted  for  episode-
     specific temperatures.  This is done by running  the latest EPA
     MOBILE  model  using  episode-specific maximum and  minimum
     temperatures.  Chapter  7  of  the emission inventory guidance
     document13  describes  the  procedures  for deriving  episode-
     specific  mobile-source emissions using  the  latest  MOBILE
     model.  Use of models other than the latest EPA MOBILE model
     should be reviewed  by the Technical Committee  on a case-by-
     case basis,  and  is  subject to approval  by  the EPA Regional
     Office.
                                                         Issues
                                37

-------
     Biogenic  emissions must reflect episode-specific conditions
     (see Section  3.7.5).

     If  available,  episode-specific operating  rates  for point
     sources are  preferable for estimating temporal point-source
     emissions.  Procedures for temporally adjusting point and area
     sources are also provided in the emission inventory  guidance
     document.li
     3.7.5  Bioqenic emissions


     Biogenic emissions can be a significant portion of the overall

VOC emission inventory for a given domain, particularly in areas of

high vegetative density.   The EPA provides the Biogenic Emissions

Inventory System  (BEIS),  which  can develop day-specific, hourly,
gridded, speciated inputs (see  Reference 4),  and also provides a

national data  base of  land use distributions with this system.

Spatial variability is limited to the county level  (i.e., emissions

are evenly spread throughout the grids within a specific county).


     The EPA  is  currently  modifying  the BEIS to  allow  users to
input  user-derived   and  possibly  more   up-to-date  land  use

distribution data.  Users will be advised of the expected delivery
date of the modified processor via the SCRAM BBS and EPA Regional
Offices.
     Recommendations

     Biogenic emissions must be included in the emission inventory
     developed for  each  model simulation  (i.e.,  base case  and
     control strategy).   The biogenic emission  processor (BEIS)
     that is part of the EPA Emissions Preprocessor System4 should
     be used to derive the inventory.   Use of alternative land use
     factors in the  BEIS should be described and documented in the
     Protocol Document.

     Also, methods  other  than the  BEIS  may  be considered  for
     deriving  the biogenic  emissions.   These  methods must  be
     described in the  Protocol  Document along  with  justification
     for using them.
 Data  Issues
                               38

-------
  !   3.7.6  Point-source and plume-rise cutoff levels
  •t                    •  •      •

  j    Guidance for  initiating ozone/CO SIP  emission inventories

pursuant to the 1990  CAAA15 specifies point-source cut-off levels

of [10 tons/yr and 100 tons/yr for VOCs and NOX/ respectively.  Any

source may be treated as a point source as long as stack data are

specified that  allow derivation  of  effective plume  height,  and
source  location  is  provided.
                              4
In  some cases,  the  Technical
Committee may  wish to  treat selected  smaller sources  as point

sources.


  I   The UAM EPS4 requires the specification of a plume-rise cutoff

level for delineating  elevated point sources from area sources.  If

the plume rise that the EPS calculates for a given point source is

below the user-specified level, then the point-source emissions are

placed  in the  area-source emissions file.  If the  plume rise is

above the level, the emissions are treated as  coming from elevated

point  sources   and  are  then  placed  within the  appropriate UAM

vertical layer.
     Recommendations

     Point-source  cutoff levels  of  10 tons/yr  for VOCs  and no
     greater than 100 tons/yr for N(X are recommended for inclusion
     in the modeling emission inventory.  Point sources must have
     the stack data needed to calculate effective  plume height, so
     that  the  heights at  which emissions are  injected  into the
     modeling system can be determined.

     The Technical Committee may consider  using a  lower plume-rise
     cutoff level, particularly in areas where there may be a high
     density of  point  sources.   Additionally, the CAAA specifies
     "major source" definitions  that have lower cutoff limits for
     purposes such as  application of reasonably available  control
     technology  (RACT),  new source review  (NSR)  and creation of
     Emission Statements.15  The Technical Committee may consider
     using these lower cutoff  limits  in  the modeling inventory.
     The Technical Committee should specify the plume-rise cutoff
     level to be used  in delineating point-source and area-source
                                39

-------
     emissions, and the level should be identified  in the Protocol
     Document.
     3.7.7  Consistency with national inventories

      Comparisons should be made between the modeling inventory and
the 1990 SIP and RFP tracking emission inventories reported in the
EPA Aerometric  Information Retrieval  System (AIRS).16   Although
these  inventories  will  not  be  identical,  such a  check can  be
considered  part  of   the  quality  assurance  process.     Major
inconsistences should  be noted and documented.   It is especially
important that those planning to use ROM-derived air quality data
in the model simulations follow applicable guidance/regulations for
reporting  statewide  emissions  data  to  AIRS.    These  national
inventories are used  in the  ROM modeling.   As  noted previously,
using  the ROM  is  the preferred  procedure  for estimating  UAM
boundary  conditions   and  meteorological   inputs.     Attainment
demonstrations will be less consistent if the ROM and the UAM use
significantly different emissions data bases.

     Recommendations
     For  an  acceptable  attainment demonstration,  documentation
     should be  provided that  shows that  the  modeling  emission
     inventory is  consistent with  the  emission inventory  being
     reported in AIRS  in accordance with applicable guidance and
     regulations.16
 Data  Issues
                               40

-------
                            CHAPTER 4
      DATA QUALITY ASSURANCE AMD MODEL DIAGNOSTIC ANALYSES
     This chapter provides general guidance for quality assurance
testing of component data  input  fields  and diagnostic testing of
base case episodes.   These analyses are designed to establish and
improve reliability  of the  input data and proper functioning of the
model.

     Although the UAM has been evaluated on a number of historical
data bases,  measures  of model behavior  with  respect to observed
data are  necessary  for  new  applications.  Model  developers and
users perform diagnostic tests to uncover potential input data gaps
that,  when corrected,  may lead to  improved  treatment  of  model
processes.  Regulators need some indication that the model captures
the key features of  the  base  meteorological episodes being applied
inlthe model simulations in  order to have  confidence  in a model's
   1
ability to predict future ozone (1) after applying projected growth
and planned  emission  controls and (2)  after applying alternative
emission control  strategies.

   j  Important  prerequisites for a  model performance evaluation
 (see Chapter 5) are (1)  quality  assurance  testing  of  model inputs
and  (2) diagnostic  testing  of  the base  meteorological episode
simulation to ensure that the model is functioning  properly and
that apparently accurate model results  are being obtained for the
right  reasons.    For example,  quality  assurance testing of  input
data  helps to ensure  that apparently good model results have not
 resulted  from compensating errors  in input data.
                                                   '"Quality
                                41

-------
     An  excellent  compilation  of  model  performance  evaluation
techniques,  including  diagnostic tests  and related  issues,  is
contained in Reference 17.  This reference serves as the basis for
this chapter and for the model performance evaluation described  in
Chapter  5.   Various  graphical  and  numerical measures  described
below are treated in detail  in Reference  17.

     Two useful graphical displays for both quality assurance and
diagnostic testing are mapping and time-series plots.

     Mapping is  a two-  or three-dimensional spatial  display  of
values  illustrated  with  various contouring  and tiling methods.
These displays may depict political boundaries and monitoring  site
locations  as  well.    Mapping  capability  is  a  multipurpose  tool
applicable  for all forms of  gridded  data,   such as future-year
emission control strategy results and most input data fields  (e.g.,
gridded wind fields, temperatures, and emission densities).  Point-
source  locations  may also  be  depicted to ensure  that they are
properly located.  Spatial displays of predicted  and  observed ozone
patterns are particularly useful as part  of  a  model performance
evaluation.

     Time-series plots displciy hourly measured and predicted ozone
values for  specific locations  such  as monitoring  sites.   Time-
series plots provide an overview of the temporal performance of the
model predictions.  Comparison of time-series  plots across multiple
monitoring sites provides an indication of spatial response.   Even
though measured VOC or NOX species data  may be  limited,  it may
still be  useful  to plot time-series plots  for some  of  these
species, particularly for  cases where ozone predictions do not meet
expectations.    Such  plots  may  provide  insights  to  the  ozone
prediction patterns and also to data base inconsistencies requiring
further investigation.
  Quality
  Assurance
                                42

-------
  i   The  following  sections  describe   recommended  steps  for
conducting diagnostic  testing of  each base  case  meteorological
episode simulation.
  l

4.1  Step 1:   Quality Assurance Testing of Component Fields

     Starting with initial,  quality-assured  data,  input data are
developed for use in various UAM preprocessors.  The first stage of
diagnostic testing should focus on assessing  the accuracy of major
UAM input fields  produced  by the UAM preprocessors.   Generally,
the testing is qualitative in nature and based on. comparing visual
displays, of  preprocessor outputs with patterns exhibited  by the
observed data.    Prior  to  conducting a base case meteorological
episode simulation,  individual air  quality,  meteorological, and
emissions fields  should be reviewed  for  consistency and obvious
omission errors.  Both spatial and temporal characteristics  of the
data should  be  evaluated.   These checks may be only cursory, but
errors  uncovered by  this component  testing might  be extremely
difficult to diagnose  later in the modeling process, when  errors
could  arise   from any  subset  of the  data  inputs.   Examples of
  i                                  •
co'mponent testing include  the  following:
     Air Quality:
      Emissions:
Check   for   correct   order  of   magnitude,
especially  when   using   background  values;
assure reasonable speciation

Plot  various  source types  by grid  cell and
review  major  source  locations  with  local
emissions   patterns;   check  major  highway
routes; generally,  look  for obvious omission
errors; plot VOCs, NOX  and CO by grid cell and
cross-check  with  source  distribution  for
                                                   - Duality
                                43

-------
                     logical  patterns,  such  as  high  NOX  levels
                     associated with major power plants
     Meteorology:
Plot  surface  and elevated  wind vectors  and
compare with monitoring  stations  and weather
maps for consistent patterns;  compare mixing
height  fields  with   sounding  data;   check
temperature fields
     In  quality assurance testing of component input fields, the
emphasis is on  capturing rather large  errors before performing
model simulations.

     Recommendations
     It  is  recommended that quality assurance testing of the air
     quality,  emissions,  and meteorological  data  input files be
     conducted before proceeding to diagnostic testing of the base
     case meteorological  episodes.   At a minimum, emissions data
     should be quality assured by looking at emission distribution
     maps and  known source  locations and  emission strengths.

4.2  Step 2:   Diagnostic  Testing of the Base Case Meteorological
     episodes

     After  confidence has  been  achieved in  producing  UAM input
fields,   the   UAM  should  be   exercised  for  each  base  case
meteorological episode.   The initial  run is  termed a diagnostic
simulation  because review of initial base case simulations usually
uncovers additional  input  errors requiring correction  before an
acceptable  set of base case inputs can  be  derived.   During this
stage of the process,  emphasis is placed  on assessing the model's
ability  to  correctly depict plume  orientation  and  the  timing of
observed ozone maxima. Accordingly, visual methods such as mapping
and time-series  plotting, using measured  data as reference marks,
may be used to assess  model behavior.
  Quality
  Assurance'
                                44

-------
Recommendations

To aid in interpreting simulation results, it is recommended
that  predicted  and  observed  ozone  concentration maps  be
constructed for each base meteorological episode simulation.

Concentration  maps  present   spatial  information  on  the
structure of the ozone plume.
Maps at 1- or 2-hour intervals should be constructed over the
periods of most  interest.   While a typical  period might be
defined as  early morning to  late afternoon for  the  day of
highest  ozone,  it  is  also  useful  to  look  at  most  time
intervals  under  recirculation,   stagnation,   and transport
conditions.

Consideration should also be given to constructing a map that
depicts the  highest predicted daily  maximum ozone value for
each grid cell.   Examples  of various mapping techniques are
described in Reference 17.

It is also recommended that the predicted concentrations used
in the  time-series  plots be  consistent with the method for
deriving predicted  concentrations for  the model  performance
evaluation described  in  Chapter  5.   This method  is based on
Reference 17  and produces  a weighted average using bilinear
interpolation  of the  predictions from the  four grid cells
nearest to the monitor location.

Other methods for deriving predicted  concentrations for time-
series comparisons may be judged appropriate by the Technical
Committee; some suggestions are contained in Chapter 5. These
methods should be described in the Modeling Protocol.

If suitable  data are available,  time-series plots should be
developed  for NO and NO.,,  and for VOC  species  at selected
locations, particularly  for cases in which ozone  time-series
or mapping results  are not  consistent with expectations.

Comparisons   of  ozone   precursors   should   be  done  for
concentration  levels  above  the detection  limits "for the
monitoring equipment.
                           45

-------
4•3  Step 3;—Additional Base Meteorological Episode Sensitivity
     Testing

     In  addition  to  running  the  base  meteorological  episode
diagnostic simulation,  other  episode  diagnostic simulations that
perturb levels of emissions, initial and boundary conditions, and
meteorological  inputs  may  provide   valuable  information  for
identifying critical  input areas and  ensuring  proper  domain and
episode selection.  The following sampling  of simulations,  which
are  equivalent  to  sensitivity  tests on  major  model  inputs,
illustrate the utility of this exercise.

     !•   Zero emissions - To indicate levels  of  sensitivity to
          emissions,   all   emissions  are  ' set   to  zero  and  the
          resulting predicted  concentrations are compared with the
          base meteorologiccil  episode  predictions that  include
          emissions.     A   lack  of  substantial  sensitivity  may
          indicate a need to reexamine the  selection of episodes or
          domains.  Variations  can be performed  by zeroing  out
          emission subsets, such  as biogenic  emissions,  mobile-
          source  emissions, and individual source  categories.

     2«   Zero boundary concentrations - Inflow concentrations at
          the  lateral boundaries and top of the modeling domain are
          reduced to zero or low background levels.  Sensitivity of
          concentrations in the inner core  and downwind portions of
          the  modeling domain provide  a measure of the  boundary
          conditions'  influence.   This simulation can  identify
          transport-dominated  episodes and provide  assurance that
          the  upwind extent of the domain is adequate for  episodes
          where intradomain emissions dominate.  In minimum trans-
          port conditions, the  second- and third-day concentrations
 Quality
 Assurance
                               46

-------
  I        (inner core and downwind locations) should be relatively
  I                                                        • •
  i        insensitive to boundary-condition concentration changes.
  i      .                                        •
  i   3.    Mixing  height  and   wind  speed  variations   -   Much
  i
  j        uncertainty is associated with mixing heights  and wind
  |        speeds, and simulated  concentrations are often sensitive
  |        to these inputs.   Simulations that test the sensitivity
  t                     .
  j        of model estimates to variations in wind  speed and/or
  1
  i        mixing height provide bounds on some of the uncertainty
  i        resulting from these  parameters.   Large sensitivity may
  '        suggest  that  future  model   applications  will   need
 . t                           '                   "
          improvement in  the meteorological  data bases.   Also,
          large  sensitivity  may  suggest  a  need  to  consider
  1        alternative wind field generation techniques.

  i   Certain  numerical  measures,  which are  recommended in the
discussion of model performance evaluation in Chapter 5,  are also
useful diagnostic tools.  For example, consistent underpredictions
usually produce bias  values  less than  zero.  This phenomenon could
be- due to  various factors, such  as overstatement of wind speeds or
mixing heights, or low emission estimates. Modelers are encouraged
toi use numerical as well as  graphical  techniques in the diagnostic
prbcess.

  '   The  diagnostic  analyses  described  in  this  chapter  are
considered to  be  a starting point  for a specific modeling study.
Diagnostic tests  discussed  in  Reference 17  should  be considered
whenever  possible.    Also,  the EPA  is developing  a UAM  Post-
processing System18  to assist in diagnostic testing  of  the base
meteorological  episodes.   Availability of this software will be
announced through the SCRAM BBS.
                                                    Assurance
                                47

-------
   Recommendations

   Diagnostic  testing of the  model should  begin  with quality
   assurance   testing  on  input  data  files  (Section  4.1).
   Diagnostic testing of each base meteorological episode should
   follow (Section 4.2). Additional diagnostic sensitivity tests
   for the base episode should also be  considered (Section 4.3),
   including using zero emissions  and/or  zero boundary condi-
   tions, and varying mixing height and wind speed estimates.

   Agreement should  be  obtained among members of the Technical
   Committee concerning input field modifications arising from
   the quality assurance testing.  These modifications should be
   based on scientific or physical reasoning and not just on what
   will improve model performance.  All changes to  the data that
   result from the diagnostic testing should be documented and
   justified.

   In addition, all  diagnostic  steps should be documented to
   avoid misinterpretation of model performance results.  After
   confidence   is  gained  that  the   simulation   is  based  on
   reasonable   interpretations   of observed  data,  and  model
   concentration  fields  generally track  (both  spatially  and
   temporally) known urban-scale plumes, a performance evaluation
   based  on  numerical  measures  is  conducted  for each  base
   meteorological  episode (see  Chapter 5).
Quality
Assurance
                              48

-------
                            CHAPTER 5
                  MODEL PERFORMANCE EVALUATION

  ;   At some point in the modeling process, an  assessment of model
performance  is  required.  Specifying  rigid rejection/acceptance
criteria has not been supported by model developers nor by decision
makers  participating  in previous  modeling  efforts.    Instead,
performance  measures derived  from  previous  photochemical  model
applications may provide a reasonable benchmark for model perform-
ance .   Also, graphical  procedures reveal qualitative relationships
bejtween predicted and observed concentrations that can be used in
model performance evaluation.

  j   Poor   performance  may   necessitate   (1)   delaying   model
applications until further diagnostic testing and quality assurance
checks  are  reflected  in the input  data  base, or  (2)  selecting
another meteorological episode for modeling.   However, this is not
a valid reason for delaying SIP attainment demonstration submittals
beyond the dates required in the 1990 CAAA.   Also,  cases where good
model  performance  is shown  should .be  reviewed as  well,  because
compensating errors  can induce spurious agreement among observed
and predicted values.                           •

5.1  Performance Measures
     This section describes recommended graphical and statistical
performance measures for ozone predictions.  These measures should
be applied for modeling results beginning on the second day of the
modeled episode.   As described in Section  3.1, the first day is
                                                   ' Evaluation:
                                49

-------
eliminated to mitigate the effects of specifying initial conditions
arbitrarily.   Performance measures should also be considered for
ozone  precursors  wherever  possible,  based  on  availability  of
monitored  data.    Obvious   problems  exist  in  comparing  model
predictions  with  observed  values.    The  UAM  output  represents
volumetric  (e.g.,  25  km3),  l-hour average concentrations, but air
quality  data  represent point  locations  with various  sampling
periods.    This "incommensurability"  may  lead  to  considerable
uncertainty in  the comparisons,  especially for precursor species
that  are  not buffered chemically and may have  been  sampled  at
locations not representative of areawide concentrations.

     As part of the UAM Postprocessing System, the  EPA is currently
developing  a model performance  utility that  will contain  the
performance measures  listed below.   Users  will be able to access
this  utility for  model performance  evaluation  testing.    This
utility is expected to be available in late 1991.  Model users will
be advised on its availability through the EPA SCRAM BBS.

     The measures used in the performance  evaluation should include
both  qualitative   (e.g.,  graphical)  and  quantitative  (e.g.,
statistical) analyses. Statistical measures may provide a meaning-
ful test of model performance for dense monitoring networks, such
as those  for special  field studies.   However, for  some routine
monitoring  networks where  coverage may  be  sparse,  statistical
measures  may  provide a  distorted  view  of model  performance,
especially for paired values.

     Reference 17 provides detailed descriptions of graphical and
statistical measures  available  for assessing  the performance  of
photochemical grid models.  The Technical Committee should consult
this  reference  when  formulating model  performance  evaluation
methods, and may want to use it for developing additional perform-
  Eval&ation;
                                50

-------
ance evaluation  procedures  other than those  recommended  in this
Guidance Document.
     5.1.1  Graphical performance procedures
  i
  i   Graphical, displays  can  provide  important  information  on
qualitative relationships between predicted and observed concentra-
tions.  At  a  minimum,  the following graphical displays should be
developed  for each meteorological  episode:   time-series plots,
ground-level  isopleths, quantile-quantile plots, and scatterplots
of predictions and observations.

     Time-series plots -  The time-series plot, developed  for each
monitoring  station in  the modeling  domain,  depicts  the hourly
predicted  and observed concentrations for the simulation period.
Thfe time series reveals the model's ability to reproduce  the peak
prediction, the presence of any significant bias within the diurnal
cycle, and a comparison of the  timing of the predicted and  observed
ozone maxima.
             17
  ;   Ground-level  isopleths  or  tile maps  -  Ground-level  isopleths
 or  tile  maps  display  the  spatial  distribution  of  predicted
 concentrations  at  a selected hour..  Isopleths of  predicted  daily
 maxima may also be constructed.  The isopleths provide information.
 on  the magnitude  and location of  predicted pollutant  "plumes."
 Superimposing observed hourly or daily maximum concentrations  on
 the predicted   isopleths  reveals   information   on  the  spatial
 alignment of  predicted and observed plumes.  Subregional biases  of
 predictions versus observations may  result from spatial
 misalignments.

     Scatterplots  of predictions and observations - Scatterplots
 depict the extent of bias  and error in  the ensemble  of  hourly
                                                    Evaluation
                                51

-------
prediction-observation pairs.  Bias is indicated by the  systematic
'positioning  of  data  points  above  or below the  perfect correlation
line.   The dispersion of points is a measure of the error in the
simulation.   The scatterplot  also  reveals  outlier prediction-
observation  pairs.

     Ouantile-quantile plots - Quantile-quantile plots compare the
frequency  distributions  of  rank-ordered observed and rank-ordered
predicted  concentrations.   The observed and predicted  concentra-
tions  are  sorted from highest  to lowest  then  plotted  on an x-y
plot.  This  graphically  depicts any model  bias  over the frequency
distribution.

     "Paired" predictions of daily maxima  - In  attainment
demonstrations,  particular  interest  is focused on daily maximum
ozone concentrations.  One test that may provide insight  into model
performance  is to consider model predictions occurring within
±1 hour of  the observed daily maxima at each monitoring site in the
nine grid  squares surrounding  and including  the monitor.   The
"prediction," for purposes of this pairing, would be the one that
agrees most closely with the observed daily maximum for  each site.
This method may  be useful for sparse meteorological  and air quality
networks,  because it  recognizes  that  both the  inputs  and  air
quality observations have some  attendant  uncertainty.   Resulting
comparisons  can be superimposed on a map  depicting emissions and
monitors to help assess model performance.

     Recommendations
     At a minimum, the following graphical  displays  are recommended
     in the evaluation of each meteorological episode:
  Evaluation
                               52

-------
     Time-series plots  of predicted  and  observed hourly  ozone
     values should be constructed for each  simulation period for
     each monitoring station where data are available.

     Ground-level isopleths or tile maps of the spatial distribu-
     tion of  predicted  concentrations should be  constructed for
     selected hours.   Also, ground-level  isopleths or tile maps of
     the daily ozone maxima should be constructed.  The correspond-
     ing observed  concentrations should be superimposed on the
     predicted concentration  isopleths  to  analyze  spatial  plume
     patterns and ozone magnitudes.

     Scatterplots should be constructed for  all hourly prediction-
     observation pairs  for each simulation;   quantile-quantile
     plots are also recommended for each simulation.

     The development of  additional graphical displays, such as the
     paired  predictions of  daily  maxima,   is  encouraged.    The
     graphical  displays to  be  used  in the  model  performance
     evaluation should be described in the Protocol.
     5.1.2  Statistical performance measures

  i
  I         •     '               .               ,             • , ,
     Statistical  performance  measures  can  provide  meaningful

measures of model accuracy for dense monitoring networks, such as
those for special field studies.  However, statistical measures may

give a  distorted  view of model  performance in cases  of routine

monitoring networks, where coverage may be sparse.  The Technical

Cpmmittee should evaluate the adequacy of the existing monitoring
network   for   conducting   statistical   tests  for   performance

evaluation.


     Recommendations
     It is recommended that, at a minimum,  the following
     mathematical formulations be applied as measures  for model
      For this purpose,  the predicted value  is  the  weighted
average of the predictions from the four grid cells nearest to
the monitoring station.  The four-cell weighted average is
derived from bilinear interpolation as described in Reference 17.
                                53

-------
    performance evaluation.
    Appendix C.
These formulations are detailed  in
    Unpaired highest-prediction accuracy - This measure quantifies
    the difference between the highest observed value and highest
    predicted value over all hours and monitoring stations.

    Normalized bias test - This'test measures the model's ability
    to replicate observed patterns during the times  of  day  when
    available monitoring  and modeled  data  are  most likely  to
    represent similar spatial scales.
    Gross error of all pairs above 60 ppb -  In  conjunction  with
    bias  measurements, this metric provides an overall assessment
    of base case performance and  can be used as a reference  to
    other modeling applications.   Gross  error can be  interpreted
    as precision.

    Additional  measures  may  include the  following:

    Average station peak prediction accuracy  - This is a  measure
    of peak performance at all monitor sites, using pairings based
    on time and space.

    Bias  of all pairs above (50 ppb  - This bias metric measures the
    overall degree  to  which model  predictions  overestimate  or
    underestimate  observed values.   Note,  however,  that  a  zero
    bias  for several observation-prediction pairs can be caused  by
    a  canceling effect of overprediction and  underprediction  in
    different subregions.

    Bias  of all station peak£5 - For this metric, bias calculations
    are performed on observation-prediction pairs associated  with
    peak  ozone  values for each monitoring station.    This  metric
    provides information on the model's ability to replicate  peak
    ozone observations.

    Fractional  bias for peak concentration -  Fractional bias  is
    calculated  for both the mean  and the standard deviation  of
    peak  predicted and  observed  values.   This  metric provides
    additional  information on the  model's  ability  to  replicate
    peak  ozone  observations.
Evaluation
                              54

-------
5.2  Assessing Model Performance Results
     Both graphical and statistical performance measures should be
used  for  the  performance  evaluation.    However,  although  the
recommended  statistical  measures  should  be  applied  for  all
meteorological  episodes  and  monitoring  networks,  caution  is
suggested  for  interpreting  these  measures  in  cases of  sparse
  r
monitoring  network  coverage.    The  Technical Committee  should
consider the monitoring network design in interpreting statistical
measures.
  ,   In  assessing  model simulation  results for  the performance
evaluation, there  is  no rigid criterion  for model acceptance or
rejection  (i.e., no  pass/fail test).  Reference  17 states that,
ba|sed  on  past photochemical  model  evaluations,  this type  of
modeling "generally produces peak (unpaired) prediction accuracy,
ov.erall bias,  and gross error statistics in  the approximate ranges
of ±15-20 percent, ±5-15 percent, and 30-35 percent, respectively."
In general, performance results that fall within these ranges would
be acceptable.  However, caution is urged  in using these ranges as
th|e sole basis for  determining the acceptability of model perform-
ance.   These  ranges  were  derived  from  past iiodel performance
  1                                        •
evaluations with varying densities of air  quality  and meteorologi-
cal monitoring networks and corresponding variations .in the quality
arid quantity of aerometric  model input data.  In  some cases, they
reflect use of earlier versions of the UAM.   Thus,  these ranges
should be  used in conjunction  with the  graphical procedures to
assess overall model performance.
     If statistical  results  are worse than the  above ranges and
graphical  analyses  also indicate  poor model  performance,  users
should consider choosing an alternative meteorological episode for
modeling.    Performance  evaluations  should   be* done  on  other
                                55

-------
candidate  episodes  to identify those that might result in better
model performance.

     If statistical results are worse than the above ranges for any
of the three statistics, but graphical  analyses generally  indicate
acceptable model performance,  simulation results used for attain-
ment  demonstration  should  be applied with  caution.   Users may
consider  conducting  performance  evaluations on  other  candidate
episodes to identify  any that  might  yield improved model
performance.

     A minimum of 3  primary  episode days is required for  use in the
model simulations for attainment demonstration (Section 6.4).  If
fewer than 3 primary episode days  can  be  identified  that  have
acceptable model performance for the  attainment demonstration, the
responsible regulatory  agencies are strongly encouraged  to  take
steps that will  improve model performance for  any future attainment
demonstrations.  For  serious  and above nonattainment areas,  this
may require short, intensive field studies to supplement installa-
tion of the enhanced monitoring network required under the CAAA of
1990.
     Recommendations
     It is recommended that the model performance for each
     meteorological episode be assessed as follows:
     1.  The graphical performance procedures specified in Section
     5.1.1 should be conducted for each meteorological episode.  To
     assess  model  performance,  the  Technical  Committee  should
     analyze  the  time-series   plots r   ground-level   isopleths,
     quantile-guantile plots, and scatterplots.  Use  of "paired"
     predictions of daily maxima may also be considered.
     2.  The statistical performance measures specified in Section
     5.1.2 should also be derived and evaluated  for each meteoro-
     logical episode.    When  interpreting these  measures,  the
     monitoring network density  and design  should  be  considered.
 Evaluation ;!:
                               56

-------
Caution is urged when  interpreting  the statistical measures
for a sparse monitoring network.

It is recommended that the statistical performance measures be
compared with the following ranges:

     •  Unpaired highest prediction accuracy: ±15-20 percent
     •  Normalized bias: ±5-15 percent
     •  Gross error of all pairs >60 ppb: 30-35 percent

If all of  these statistical measures  are  within  the ranges
shown,  and the  graphical performance  procedures  also  are
interpreted to  yield acceptable results,  then  the model is
judged to be performing acceptably.

If any of  the statistical measures  are worse than the above
ranges, or the graphical procedures are interpreted to yield
unacceptable performance,  users  should consider choosing an
alternative  highly  ranked meteorological  episode  for  the
attainment demonstration.  Performance evaluations should be
conducted  on a  prospective alternative episode to determine
whether it yields improved model performance.
Additional model performance measures are encouraged.
should be described in the Modeling Protocol.
These
                                              * Evaluation.
                           57

-------
Evaluation
                              58

-------
 j                           CHAPTER 6
 ;                   ATTAINMENT DEMONSTRATION
 i          . •                    '             '  I
 i               -                •
 [    This chapter provides guidance on using modeling simulations
for attainment demonstrations.  The primary reason for conducting
 f                                          '
photochemical  modeling is  to  demonstrate  the effectiveness  of
alternative control strategies in attaining the  NAAQS  for ozone
throughout the modeling  domain.  This  demonstration  consists  of
four main parts:  (1) developing attainment-year modeling emission
inventories, (2)  developing alternative-control strategy emission
inventories, (3)  performing model  simulations  for the attainment
year with and without alternative control strategies,  and
($)  comparing  attainment  year and  control strategy  simulation
results with the ozone NAAQS.  Attainment year and control strategy
simulations are conducted for  each selected meteorological episode
(see Section 3.1).
6;. 1  Developing Attainment—Year Model Inputs

 ;    The attainment-year modeling  inventory  must  be derived from
the  1990  SIP  nonattainment  base  year  inventory,  adjusted  for
episode-specific meteorology, and then projected to the attainment
year.   Also,  to  the extent possible,  initial- and  boundary-
condition ozone and precursor concentrations must be projected to
the  attainment  year.   The  attainment year is determined  by  the
npnattainment area designation and the attainment dates specified
in the 1990  CAAA.   Projections of emission inventories reflect the
net effect of mandated controls  and growth projections for various
source categories.  Guidance for projecting inventories is
                                                "  Demonstration
                                59

-------
available  in Procedures for Preparing Emissions Projections.14  The
most direct method for projecting  initial- and boundary-condition
precursor  concentrations is by applying ROM simulation results for
which the  UAM  domain  is nested within the ROM  domain  (see
Chapter  3).  In the absence of available ROM data, the projection
of  ozone  precursor concentrations  used  for  initial conditions
typically  has been done by  linear scaling based on emission changes
projected  to take place from the 1990  base year to the  future year.
For initial ozone  concentrations,  there is little basis for doing
anything other than assuming  initial ozone remains constant.   In
the absence of regional modeling results or better information, the
guidance in Reference  7 for specifying future  boundary conditions
may be followed.

     Recommendations
     It  is recommended that  the  EPA guidance document entitled
     Procedures for Preparing Emissions Projections14  be consulted
     for developing attainment-year  inventories.    The guidance
     document  provides procedures for projecting  point-source,
     area-source,   mobile-source,  and biogenic  emissions,  and
     addresses  projections of  spatial,  temporal,  and  chemical
     composition changes  between the  1990 SIP inventory  and the
     attainment-year inventory.
     Also, if regional modeling predictions for the attainment year
     are available, it is recommended that these be used to derive
     the attainment-year initial  and boundary conditions  for the
     attainment-year model simulations (see Chapter 3).
6.2
;tion of Attainment
     Many possible attainment-year emission control strategies can
be set up and simulated.  Eventually, a modeling analysis must be
submitted for approval  as  the basis of a SIP demonstration.  The
effectiveness of a given set of control measures in reducing ozone
(and perhaps other pollutants)  is a major factor in selecting the
final emission control strategy.
  Demonstration
                                60

-------
 i    Prior studies have  typically used a  progression  of control
strategy  scenarios  in  the modeling  to  ascertain an  effective
 i                     - •           •     -      -.
strategy for attainment.  A suggested logical  progression is the
following:
          Simulate the CAAA and other mandated control measures for
          the attainment year to determine  if  these measures are
          sufficient to demonstrate attainment of the ozone NAAQS.

          If mandated controls are  insufficient to  demonstrate
          attainment, superimpose a series of additional, across-
          the-board  reductions  in VOCs-only,  VOCs-plus-NOv,  and
                                                           A
          NOx-only strategies,  relative  to the mandatory  CAAA
          controls,  to   identify  a  suitable  emission-reduction
          target range.                          I
     3.   Once an approximate target range is ascertained in
          steps 1 and 2, simulate control strategies that reflect
          source-specific  or   source-category-specific   control
          measures  and  that realize  the  approximate  emission
          reductions  identified as  sufficient  to  reduce  daily
          maximum ozone to 0.12  ppm or less.
     4.    Adjust  the strategy  chosen  in  step  3  until  it  is
          sufficient to demonstrate  attainment of the  NAAQS,  as
          described in Section 6.4.   Adjustments; may be needed in
          VOC controls,  or NOX controls, or  both.

     Recommendations
     The procedures for deriving control strategies for evaluation
     in  the  attainment demonstration  must be  specified in  the
     Modeling Protocol.
                               61

-------
     Emission  control  strategies for  linked urban-area modeling
     domains (e.g., northeastern U.S. Corridor) should be coordi-
     nated  among State  agencies  having lead  responsibility for
     respective domains to ensure consistency among the domains.


6.3  performing	Attainment-Year  Simulations  to  Assess  Various
     Control Strategies


     Many graphical display and numerical procedures are available

for  illustrating  the  effects  of  alternative emission  control

strategies on predicted concentrations of ozone and  other species.

For example,  the emission  levels  in the  control  strategies are
often compared with the  attainment-year base emissions.   Also of

interest are comparisons with the inventory derived for purposes of

model  performance evaluations  and  corresponding  base-case  UAM

results.  Difference  maps are  extremely useful  for illustrating

changes in daily maximum ozone predictions  throughout the modeling
domain.
     Recommendations

     The focus  of any ozone  attainment demonstration is  on the
     daily maximum 1-hour concentration predicted at each location
     in the  modeling domain.   However,  it is recommended  that
     responsible parties broaden the scope of an attainment demon-
     stration to  examine the impact on  other  important  metrics,
     such as different concentration averaging times,  population
     exposure,  subdomain  and temporal impacts, effects  on other
     pollutant  species,  and  other  important  measures  that are
     sensitive to emission control strategies.

     For deriving initial and boundary  conditions for a particular
     urban-area domain, using appropriate regional  model predic-
     tions that reflect  control measures applied  in other urban-
     area  domains   within   the  regional  modeling   domain  is
     recommended.
 Demonstration
                               62

-------
6.4  Using Modeling Results in the Attainment: Demonstration

  j   As described in Section 3.1, at least 3 primary episode days
shpuld  be modeled  for the  attainment demonstration.    Also,  a
minimum of 1  primary day should be modeled  from each identified
meteorological regime.  Therefore,  for example,  if there are three
meteorological regimes, at least 1  primary episode day from each
regime  should be modeled; if  there are  only  two meteorological
regimes, at least 2 primary episode days should be modeled  from one
of I the regimes and at least 1 primary  episode day modeled  from the
other regime.   Note that the episodes  simulated  would generally be
at least  48 hours long (i.e.,  the first  day would be an initial
modeling day and the second day would  be the  primary episode day).
This would count as simulation of 1 primary  episode day.

     To demonstrate attainment of the  ozone NAAQS, there should be
  f
no! predicted  daily maximum ozone concentrations  greater than
0.12 ppm anywhere in the modeling domain  for each primary episode
day modeled.   Alternative methods for demonstrating attainment must
be! approved by the  appropriate EPA Regional  Office on a  case-by-
case basis.

  i   The  attainment  test  described in the preceding paragraph is
consistent with  the  flexibility  allowed in the choice of episode
days (see Section 3.1)  and reflects concerns over the difficulty of
accurately estimating  emissions inputs  to the model.
     Recommendations
     To demonstrate attainment of the ozone NAAOS.  there  should be
     no predicted daily maximum ozone concentrations greater than
     0.12  t>t>m anvwhere in  the modeling domain  for each primary
     episode day modeled.
     be modeled.
At least 3 primary episode days should
                                63

-------
     States  may opt  to  conduct  more  comprehensive statistical
     testing of the modeling results for the attainment demonstra-
     tion.   Any alternative methods for attainment demonstration
     must be approved by the appropriate EPA Regional Office on a
     case-by-case basis.   Any optional  methods  should be agreed
     upon during the development of the Modeling Protocol.


6.5  libeceptions to Guidance Document


     It is not possible  in a general guidance document like this to

anticipate   all  contingencies  associated  with  developing  an
attainment demonstration study.  The Modeling Policy Oversight and

Technical Committees responsible for a specific modeling study may

propose an  alternative  photochemical  modeling  approach provided

that (1) the Modeling Protocol requires consensus on the proposed
alternative  approach within the  Technical  Committee,   and  (2)

justification for the proposed approach is documented.  Application

of  any alternative  photochemical  modeling  approach  must  first

receive concurrence in  writing  from the responsible EPA Regional
Office(s).
  Demons tr ertioti'; I.
                                64

-------
                           REFERENCES
i.;
2.
3.
4.
5.1
6.
 7.1
Morris, R.E., T.C. Myers and J.L. Haney, 1990.  User's Guide
for the Urban Airshed Model.  Volume  I:  User's Manual for UAM
(CB-IV) .  EPA-450/4-90-007A,  U.S.   Environmental  Protection
Agency, Research Triangle Park, NC  (NTIS No.: PB91-131227).

Morris,  R.E., T.C.  Myers,  E.L.  Carr,  M.C. Causley,  S.G.
Douglas  and  J.L.  Haney, 1990.   User's Guide for  the Urban
Airshed Model. Volume  II;  User's  Manual for the UAM fCB-IV)
Modeling   System.   EPA-450/4-90-007B,   U.S.   Environmental
Protection Agency, Research Triangle Park, NC (NTIS No.: PB91-
131235).

Douglas, S.G., R.C. Kessler and E.L. Carr, 1990.  User's Guide
for the Urban Airshed Model. Volume III; User's Manual for the
Diagnostic Wind Model. EPA-450/4-90-007C, U.,S. Environmental
Protection Agency, Research Triangle Park, NC (NTIS No.: PB91-
131243).                                    i

Causley, M.C., J.L. Fieber, M. Jimenez and L. Gardner, 1990.
User's Guide  for  the Urban Airshed Model. Volume IV; User's
Manual for  the Emissions Preprocessor System. EPA-450/4-90-
007D, U.S. Environmental Protection Agency, Research Triangle
Park, NC  (NTIS No.: PB91-131250).

Tang,  R.T.,  S.C.   Gerry,  J.S.  Newsom, A.R.  Van Meter, R.A.
Wayland, J.M.  Godowitch  and K.L. Schere,  1990.  User's Guide
for  the  Urban Airshed Model.  Volume  V;	Description—and
Operation of the RQM-UAM Interface Program System. EPA-450/4-
90-007E,  U.S.  Environmental  Protection Agency,  Research
Triangle Park, NC  (NTIS  No.: PB91-131268).

DaMassa, J. , 1990. Technical Guidance Document: Photochemical
Modeling.  California  Air  Resources Board, Technical Support
Division, Sacramento,  CA.

U.S.  Environmental Protection  Agency, 1991.   Criteria  for
Assessing  the Role of Transported  Ozone/Precursors in Ozone
Nonattainment Areas.  EPA-450/4-91-015,  Office of Air Quality
Planning and  Standards, Research Triangle Park,  NC  (NTIS No.:
PB91-195958).
                                65

-------
8.
9.
10,
11,
12,
13,
14,
15,
16,
17,
 U.S. Environmental Protection Agency,  1991.   Gridded Model
Information Support System (GMISS1  User's Guide. Volume II;
UAM  Subsystem,  EPA-450/4-91-009/  Office  of  Air  Quality
Planning and Standards, Research Triangle Park, NC  (NTIS No.:
PB91-206268).

U.S. Environmental Protection Agency,  1987.  On-Site Meteoro-
logical Program Guidance for Regulatory Modeling Applications.
EPA-450/4-87-013,  Office  of Air  Quality Planning  and Stan-
dards, Research Triangle Park, NC (NTIS  No.: PB87-227542).

U.S. Environmental Protection Agency,  1989.   Procedures for
Applying City-Specific  EKMAf EPA-450/4-89-012, Office of Air
Quality Planning  and  Standards,  Research  Triangle Park,  NC
(NTIS No.: PB90-256777)„

U.S. Environmental Protection Agency,  1991.   Enhanced Ozone
Monitoring Regulations: Proposed Rule (Drafts , 40 CFR Part 58,
Office  of  Air  Quality  Planning  and   Standards,  Research
Triangle Park, NC.
U.S. Environmental Protection Agency, 1991.  Emission Inven-
tory Requirements for Ozone State Implementation Plans, EPA-
450/4-91-010, Office of  Air Quality Planning and Standards,
Research Triangle Park, NC.                            ,

U.S. Environmental Protection Agency,  1991.   Procedures for
the Preparation of Emission Inventories for Volatile Organic
Compounds.  Volume  II;  Emission Inventory Requirements  for
Photochemical Air Quality Simulation Models, EPA-450/4-91-014,
Office  of  Air  Quality  Planning   and  Standards,  Research
Triangle Park, NC (NTIS No.:  PB91-216176).

U.S. Environmental Protection Agency,  1991.   Procedures for
Preparing  Emissions  Projections,  Office  of  Air  Quality
Planning and Standards, Research Triangle Park, NC (Draft).

U.S. Environmental  Protection  Agency,  1991.    Guidance  for
Initiating Ozone/CO SIP Emission Inventories Pursuant to the
1990 Clean Air Act Amendments, Office of Air Quality Planning
and Standards, Research Triangle Park, NC.

U.S.  Environmental   Protection   Agency,  1989.    Aerometric
Information Retrieval System (AIRS'). Volume I.  Office of Air
Quality Planning and Stemdards,  Research Triangle Park, NC.

Tesche, T.W.,  P. Georgopoulos,  F.L. Lurmann and  P.M.  Roth,
1990.  Improvement of Procedures for Evaluating Photochemical
  Referenced
                                66

-------
     Models. Draft Final Report.   California Air Resources Board,
     Sacramento, CA.                            :

18.  U.S.  Environmental  Protection  Agency,  1991.  Urban  Airshed
     Model  Postprocessing System  User's  Guide, Office  of  Air
     Quality Planning and Standards,  Research  Triangle Park,  NC
     (Draft).
                                                   Inferences
                               67

-------
References
                              68

-------
                           APPENDIX A
             RECOMMENDED MODELING PROTOCOL CONTENTS

  !   Table 1 of Chapter 2 lists recommended contents for a Modeling
Protocol.  This appendix gives a general description of each compo-
nent, to  aid in  the development of  the Protocol.  As  stated in
Chapter 2, the contents presented here are  adopted  from the GARB
Technical Guidance Document: Photochemical Modeling.6

UAM MODELING STUDY DESIGN

Background and Objectives
  t                                     '         '              -
     The Protocol Document should describe the policy and technical
objectives of the study and pertinent background, information such
as the legislative mandate under which the study is being done.

Schedule

  ;   Development  of a complete schedule for all phases  of the
project  is needed.   The  critical  paths  and  deadlines  should be
identified  and discussed,  as  should  a  schedule  for addressing
critical issues that require special attention, such as air quality
and meteorological data preparation and quality eissurance, episode
selection,   and   emission  inventory   preparation   and  quality
assurance.
                                69

-------
Deliverables

     A list of the interim and final deliverables for the modeling
study should be specified.
Modeling Policy
rersiaht/Technical Committees
     The composition and  responsibilities  of the Modeling Policy
Oversight  and Technical  Committees should  be specified  to  the
extent  possible.      Meeting  frequency   and  circumstances  for
convening  a  meeting  should  be identified.   Because  technical
conflicts may arise, a resolution process for handling them should
be included.

Participating Organizations

     The organizations that are sponsoring the modeling study and
those that may contribute to  it should be  identified.

^Relationship to Regional Modeling Protocols

     Procedures  for coordinating  development of  the  urban-area
Modeling Protocol  with the regional Modeling  Protocol  should be
described.  This would  include a description of  control strategies,
emission inventories,  projection years, modeling episodes, etc.
The  coordination  of   urban-area  Modeling  Policy  and  Technical
Committees with their  regional counterparts  should be described.

Relationship to Other  Urban Area Modeling  Protocols

     In  some cases,  such as the   Northeast U.S.,  nonattainment
MSA/CMSAs required to do attainment  demonstrations may be linked to
other nonattainment MSA/CMSAs.  It  is important that procedures be
  Appendix A
                                70

-------
established for  coordinating  the Modeling Protocols  among these
areas,  and  that these  procedures be  clearly specified  in  each
nonattainment area Modeling Protocol.  It is likely that Modeling
Policy Oversight and Technical Committees will include some joint
membership among the nonattainment areas.
  t                                        '
  i                      .
Relationship to Planning/Strategy Groups

  I   Key planning  agencies and  others responsible for emission
projections or  other  model inputs should be  identified,  and the
means by which  these  groups interact  to  obtain  realistic growth
prpjections and control strategies should be discussed.
 • i                          .      •
DOMAIN AND DATA BASE ISSUES
  i
Preprocessor Programs

     The preprocessor programs to be used in constructing any of
the model input fields should be identified and described.

Data Bases

  :   The proposed air  quality and meteorological data bases should
be described.   The  completeness  of  the data base, techniques for
filling in missing data, and quality assurance procedures should be
discussed.

Base Meteorological Episode Selection
  i

  i   The episode selection criteria  should be detailed, including
the methodology to group  candidate  episodes into meteorological
  I
regimes.   How  the  episodes will be used in the modeling study
should also be  described.
                                                    Appendix
                                71

-------
Modeling Domain

     The Protocol should  describe  the criteria for selecting the
size and  location  of the modeling domain.   This  would include a
description of the MSA/CMSA area size, locations of major sources
outside the MSA/CMSA that may affect  it, sensitivity analyses that
may be conducted to assess boundary effects on domain predictions,
relationship of domain size to the episodes selected for use in the
modeling study, etc.

Horizontal Grid Resolution

     The Protocol should describe the horizontal grid resolution to
be applied to the modeling domain.   If a resolution coarser than
5  x 5  km is  chosen,  justification for this choice  should  be
provided.

Number of Vertical Layers

     The Protocol should  specify the number  of  vertical layers to
be used in the UAM simulations.  If a layering scheme other than
the one recommended in Chapter 3 is chosen, justification for using
the alternative  layering  should be given.

Emission  Inventory

     The  assumptions,  methodologies,  and  appropriate guidance
references  to  be  used  in   constructing  the  modeling emission
inventory should be described. Quality assurance procedures should
also be described.                                              /
  Appendix A
                                72

-------
Specification of Initial and Boundary Conditions

  i   The techniques to be used to  specify the initial and boundary
conditions for the base meteorological episodes and the attainment
year should be described.  The assumptions to be used in forecast-
ing attainment-year conditions should be documented.  If a nested
grid approach is used  (e.g., using predictions from the ROM through
the  ROM/UAM Interface  System),   the  details for implementation
should be described (see Chapter  3).

Wind Field Specification

     The proposed techniques for.specifying  the wind fields should
be described.  The procedures  to be used to determine the represen-
tat:iveness  of the  simulated wind  fields  should  be  technically
justified and documented  (see Chapter 3).

Miking Heights

  !   The techniques to  be  used for deriving the mixing height  for
the modeling domain should be described.

Sources of Other  Input  Data

  !   The Protocol Document should describe the data "and techniques
to| be used to specify other input data,  such as  cloud cover,  water
vapor, UV radiation, surface temperature, terrain, and land use  and
surface characteristics.
                                73

-------
QUALITY ASSURANCE AND DIAGNOSTIC ANALYSES

Quality Assurance Tests of Input. Components

     The specific quality assurance tests  to  be  used on the data
input fields should be described (see Chapter 4).

Diagnostic Tests of Base Case Simulation

     The specific diagnostic  tests to be  used for the base-case
meteorological  episode  simulations  should  be  described.    As
discussed in Chapter 4, these should include, at a minimum, time-
series plots,  observed and predicted ozone maps, zero emissions and
zero boundary conditions tests, and  tests on the mixing height
variations  and wind  fields.    Additional  diagnostic  tests  are
encouraged and should be described in the Protocol.

MODEL PERFORMANCE EVALUATION

Performance Evaluation Tests

     The graphical, statistical, and other measures to be used in
the  model performance evaluation  should  be specified.   At  a
minimum, the  tests  recommended in Chapter  5  should be included.
Additional measures may also be considered and should be described
if they are to be used.
  Appendix A
                                74

-------
ATTAINMENT DEMONSTRATIONS

Identification of Attainment-Year Mandated Control Measures

  i   The Protocol Document should include a description of the 1990
CAAA control measures and other measures mandated to be implemented
by the attainment year.

Methodologies for Generating Control Strategy Emission Inventories

  1   The  procedures  for  deriving  alternative-control-strategy
emission  scenarios to  meet  the study objectives should  be de-
scribed.  A description of how the control scenarios would relate
to applicable control strategies for areas adjacent to the modeling
domain  (particularly  upwind  areas) should be included.

Procedures for Attainment  Demonstration
  i                 •   •
  i   Procedures   for  using   the   model  simulation  results   in
demonstrating attainment of  the ozone  NAAQS  should be  included.

SUBMITTAL PROCEDURES

  i   The  documentation and analyses that will be submitted for EPA
Regional   Office  review  should   be  described.     Also,  any
documentation  other  than the  Modeling  Protocol  requiring EPA
Regional  Office  approval  should be  described.
                                                    Appendix A.
                                75

-------
Appendix
                             76

-------
                           APPENDIX B
            IDENTIFICATION OF METEOROLOGICAL REGIMES
             CORRESPONDING WITH HIGH OBSERVED OZONE

     The following is  a  procedure that may be used to  assist in
selecting modeling episodes.  Other techniques may be considered on
a case-by-case  basis;  they should  be  described in the Modeling
Protocol and approved by the appropriate EPA Regional  Office.

  !   Identification  of meteorological  regimes  for a given area
under review begins with  constructing a climatological windrose of
high ozone days.  The  windrose  is constructed  by first selecting
all days from the period  1987 to present during which at least one
ozone monitor within the area recorded an exceedance of the ozone
NAAQS or some other cutoff level (e.g.,  100 ppb).  Additional years
of: data are encouraged in constructing the climatological windrose
(e.g., 1980-1991).   Next,  for  each exceedance  day, calculate the
morning  (i.e.,  7:00  a.m. - 10:00  a.m.)  resultant wind velocity.
Then group the resultant wind velocities  for all  of the exceedance
days  into eight  compass directions  plus  calm, to  establish a
climatic  windrose of high-ozone days for  the  area under review.
Calm winds are defined as those with speeds less than 1.5 m/s and
referred to as the null wind direction.   The windrose will include
nine  bins (0-8); place the wind  directions corresponding to the
eight  compass points  into  bins 1-8, and  the  calm or  null wind
direction into bin  0.   The bins  with frequencies significantly
higher than the average frequency  for all bins should be  defined as
the  "predominant wind  directions"  (PWD).
                                                    Appendix B
                                77

-------
     Next,  compare  the morning  (i.e.,  7:00  a.m.  -  10:00  a.m.)
resultant wind velocity for each  exceedance day during 1987-89 and
more recent years with  the climatic windrose of high-ozone days.
Categorize exceedance days with  wind directions corresponding to
previously  identified  climatic PWD's as  belonging to  that PWD.
Lump all other exceedance days occurring during 1987-89 and later
into a category called  "other."   Rank each exceedance day within
each PWD category and within the  "other" category according to its
areawide daily maximum  ozone observation.   Within each category,
the day with the  highest  areciwide daily maximum concentration is
ranked first.

     After the steps described in the two preceding paragraphs are
completed, meteorological  regimes can be defined for  use in the
attainment demonstration test described in Section 6.4.  This may
be done as follows:

     1.   Choose the two PWD's which contain the highest areawide
          daily maximum ozone values from 1987 to the most recent
          year with  data  available.  These  represent  two of the
          meteorological  regimes to  consider in the attainment
          test.

     2.   The third "meteorological regime" to be considered in the
          attainment  test  is  comprised of  all  exceedance days
          previously categorized as "other" plus those belonging to
          any PWD not chosen in step 1.
  f
     3.   Identify the  top 3-ranked exceedance  days from within
          each of the three meteorological regimes identified in
          steps 1 and 2.  These days are candidates for modeling in
          the  attainment   test.   Final  choice from  among these
  Appendix B
                                78

-------
candidates is based on criteria identified in
Section 3.1.
                                       !
It may  happen that one  or more of  the meteorological
regimes  identified in  step  1  contains  fewer than  3
exceedance  days.     If   this  occurs,  exceedance  days
included within PWD's  which have been lumped in the third
meteorological regime (see step 2) may be added to one or
both of the first two regimes.  If this  proves necessary,
selection of days to supplement those  in one or both of
the first two regimes needs to be decided on a case-by-
case basis keeping in  mind the goal of this exercise:  to
provide a choice of exceedance days reflecting high ozone
concentrations  with  meteorological   conditions  which
frequently  coincide with observed exceedances.
                      79

-------
;•-  Appendix
                                80

-------
   ;                         APPENDIX C

   i              PERFORMANCE MEASURE FORMULATIONS
   t                •                      •        ;
   |
RECOMMENDED PERFORMANCE MEASURES1
   i
   j
1. !  Unpaired Highest-Prediction Accuracy  (Au)
   i                                            '  !
   [
   [                                      ,        '
   [ .                                 •             :
   i                      r* (    \ — C* (   }         i
                    A  =  °°{-' ^	£p*'"; X100%
where
     cft(.,.)
unpaired highest-prediction accuracy
(quantifies the difference between the
magnitude of the highest2 1-hour observed
value and the highest 1-hour predicted value

maximum 1-hour observed concentration over
all hours and monitoring stations

maximum 1-hour predicted concentration over
all hours and surface grid squares
            on Reference 17.
   I   2"Highest" refers to the maximum 1-hour concentration across
 all  hours and monitoring stations;
                                                   .. Appendix C
                                81

-------
2.   Normalized Bias Test CD 'I
where
          N
normalized bias obtained from all hourly
prediction-observation pairs


number of monitoring stations


number of hourly prediction-observation pairs

for monitoring station i
          Nn,
total number of station-hours
                                N
                    observed value at monitoring station i  for

                    hour j


                    predicted value at monitoring station i for
                    hour j
     Predicted value derived from bilinear interpolation of the
predicted values at the four grid cells nearest to station
station ± for the given hour.
                                82

-------
3.   Gross Error of All Pairs >60 ppb
                                   ,J) - Cp(i,j)
                         iKL J=l
where
          N
     cp(i,j)4
normalized gross error for all hourly
prediction-observation pairs for hourly
observed values >60 ppb
                             i

total number of station hours (defined
previously)
                             i
number of monitoring stations?

number of hourly prediction-observation pairs
for monitoring station i

observed value >60 ppb at monitoring station
i for hour j

predicted value at monitoring station i for
hour j
      Predicted value derived from bilinear interpolation of the
 predicted values  at the four grid cells nearest to station i for
 the  given hour.
                                                      '
                                                    Appendix C
                                83

-------
OTHER SUGGESTED PERFORMANCE MEASURES
1.   Average Station Peak Prediction Accuracy A
1 = 1

                    N
                              .tj -cd.tj
                                            ,     .
                                            '
where
                    mean paired peak5 prediction accuracies

                    averaged over all monitoring stations
          N
    number of monitoring stations
                    peak observed value at monitoring station i

                    for hour tj


                    predicted value at monitoring station i for

                    hour t<
     5 "Peak" refers to the daily maximum 1-hour concentration at
a particular monitoring station.

     6For these "Other Suggested Performance Measures,"
"predicted" can be  interpreted  in one of several ways:   (1) as
the result of bilinear interpolation described  in footnote 4;
(2) using the procedures described for paired predictions of
daily maxima (described on page 52 of the text); (3) using the
prediction for the  grid square  containing the monitor  site only.
The Modeling Protocol should document the procedure used to
determine "predicted" values in these tests.
                                84

-------
          t:
hour of peak observed value at monitoring
station i
2. :  Bias of All Pairs >60 ppbfDgQ1
                          N
where
          D
           60
          N
non-normalized bias from all hourly
prediction-observation pairs for observed
values >60 ppb
                             I

total number of station-hours  (defined
previously)

number of monitoring stations;

number of hourly prediction-observation  pairs
for monitoring station i
                    observed value  >60 ppb at monitoring  station
                    i for hour  j
                                85
                                                    Appendix C

-------
     cp(i,j)7   =
predicted value at monitoring station i for

hour j
3.   Bias of All Station Peaks (D,
                           N
                                          ±, tj
where
     Dpeak
non-normalized bias from all prediction-

observation pairs for peak8 observed values

at all monitoring stations
          N
number of monitoring stations
                    peak observed value at monitoring station i

                    for hour t)
     'For these "Other Suggested Performance Measures,"
"predicted" can be interpreted in one of several ways:   (1) as
the result of bilinear interpolation described in footnote 4;
(2) using the procedures described for paired predictions of
daily maxima (described on page 52 of the text); (3) using the
prediction for the grid square containing the monitor  site only.
The Modeling Protocol should document the procedure used to
determine "predicted" values in these tests.

     8"Peak"  refers to the daily maximum l-hour concentration at
a particular monitoring station.
  Appendix C
                                86

-------
                    predicted value at monitoring station i for

                    hour t:
  I        tj •  . =    hour of peak observed value at monitoring

  1                  station i
  [

4. ; •  Fractional Bias for Peak Concentration      j

  f                               ' .
  •;   The fractional bias is calculated for both the mean and

standard deviation of peak ozone values, as follows:
                          ,  .
where
                    fractional bias of means
                    fractional bias of standard deviation
     9For these "Other Suggested Performance Measures,"
"predicted" can be interpreted in one of several ways:   (1) as
the result of bilinear interpolation described in footnote 4;
(2|) using the procedures described for paired predictions of
daily maxima (described on page 52 of the text); (3) using the
prediction for the grid square containing the monitor  site only,
Th<= Modeling Protocol should document the procedure used to
determine "predicted" values in these tests.
                                                    Appendix C
                                87

-------
          m,.
mean maximum observed concentration
          m    =    mean peak predicted concentration


          s0   =    standard deviation of peak observed

                    concentrations


          Sp   =    standard deviation of peak predicted

                    concentrations


     The means and standard deviations of predicted and observed

concentrations are determined by each of two methods:
Peak station values:
                    maximum observed concentration at monitoring

                    station i across all hours
     cp(i,O
            10  _
maximum predicted concentration at monitoring

station i across all hours
     where i
1,...,N monitoring stations
Peak hourly values:
          these "Other Suggested Performance Measures,"
"predicted" can be interpreted in one of several ways;  (1) as
the result of bilinear interpolation described in footnote 4;
(2) using the procedures described for paired predictions of
daily maxima (described on page 52 of the text); (3) using the
prediction for the grid square containing the monitor site only.
The Modeling Protocol should document the procedure used to
determine "predicted" values in these tests.
  Appendix C
                                88

-------
                    maximum observed concentration at hour j

                    across all monitoring stations
,11  -
                    maximum predicted concentration at hour j

                    across all monitoring stations
     where j
         1,...,H hours
   i  The fractional bias of the mean and standard deviation
   I s
varies from -2 to +2.  Negative values indicate c>verprediction

and positive values indicate underprediction.
     nFor these  "Other  Suggested  Performance  Meassures,"
"predicted" can  be interpreted in one of several ways:   (1) as
the result of bilinear  interpolation described in footnote 4;
(2) using the procedures described for paired predictions of
daily maxima (described on page 52 of the text); (3) using the
prediction for the grid square containing the monitor site only,
The Modeling Protocol should document the procedure used to
determine "predicted" values in these tests.
                                                    Appendix C
                                89

-------

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
 BEPOBT N.O,
 EPA-450/4-91-013
                                                            3. RECIPIENT'S ACCESSION NO.
  HTLE AND SUBTITLE
     Guideline for  Regulatory Application
     of the Urban Airshed Model
                                                            5. REPORT DATE
                                                                 July,  1991
             6. PERFORMING ORGANIZATION CODE
 AUTHOR(S)
                                                            8. PERFORMING ORGANIZATION REPORT NO.
 PERFORMING ORGANIZATION NAME AND ADDRESS
                                                             10. PROGRAM ELEMENT NO.
    U.S. Environmental  Protection  Agency
    Office of Air  Quality Planning and Standards
    Technical Support Division
    Research Triangle Park, NC   27711	_
             11. CONTRACT/GRANT NO.
 2. SPONSORING AGENCY NAME AND ADDRESS
                                                             13. TYPE OF REPORT AND PERIOD COVERED
                                                             14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT

       This document provides  guidance on application of the  U.S.  EPA Urban Airshed
  Model for ozone  nonattainment  area State Implementation Plan  development as
  required by  the  Clean Air Act   Amendments of  1990.   The Urban Airshed Model  is  an
  urban-Scale,  grid-based photochemical dispersion model.  The  model  provides  a means
  for studying  the relationship  of volatile organic compound  and nitrogen oxide emission
  to amb(ient levels of ozone  in  urban areas.  This document provides  information  on
  data requirements and model  application strategy.
                                 KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                               b. IDENTIFIERS/OPEN ENDED TERMS
                           c.  COSATI Field/Group
     Air Pollution
     Atmospheric Dispersion  Models
  Photochemical Model
  ozone
18. DISTRIBUTION STATEMENT

     Release Unlimited
19. SECURITY CLASS (TillsReport!
  Unclassified	
                                                                           21. NO. OF PAGES
20. SECURITY CLASS (This page/
  Unclassified
                            22. PRICE
EPA Form 2220-1 (Rev. 4-77)   PREVIOUS EDITION is OBSOLETE

-------

-------

-------
    U.S.EPAOAQPS
RESEARCH TRIANGLE PARK, NC


  AEPA

-------