Appw 23 2023-22876

72826 Federal Register / Vol. 88, No.
203 / Monday, October 23, 2023 / Proposed Rules
ENVIRONMENTAL PROTECTION • Fax: (202) 566–9744. VI. Statutory and Executive Order Reviews
AGENCY • Mail: Environmental Protection A. Executive Order 12866: Regulatory
Agency, EPA Docket Center, Office of Planning and Review and Executive
40 CFR Part 51 Air and Radiation Docket, Mail code Order 14094: Modernizing Regulatory
Review
[EPA–HQ–OAR–2022–0872; FRL–10391–01– 28221T, Attention Docket No. EPA–HQ– B. Paperwork Reduction Act (PRA)
OAR] OAR–2022–0872, 1200 Pennsylvania C. Regulatory Flexibility Act (RFA)
Ave. NW, Washington, DC 20460. D. Unfunded Mandates Reform Act
RIN 2060–AV92 • Hand/Courier Delivery: EPA Docket (UMRA)
Center, Room 3334, EPA WJC West E. Executive Order 13132: Federalism
Guideline on Air Quality Models; F. Executive Order 13175: Consultation
Building, 1301 Constitution Ave. NW,
Enhancements to the AERMOD and Coordination With Indian Tribal
Washington, DC. The Docket Center’s
Dispersion Modeling System Governments
hours of operations are 8:30 a.m.–4:30
G. Executive Order 13045: Protection of
AGENCY: Environmental Protection p.m., Monday–Friday (except Federal
Children From Environmental Health
Agency (EPA). Holidays). Risks and Safety Risks
ACTION: Proposed rule; notification of Instructions: All submissions received H. Executive Order 13211: Actions
public hearing and conference. must include the Docket ID No. for this Concerning Regulations That
rulemaking. Comments received may be Significantly Affect Energy Supply,
SUMMARY: In this action, the posted without change to https:// Distribution, or Use
Environmental Protection Agency (EPA) www.regulations.gov, including any I. National Technology Transfer and
proposes to revise the Guideline on Air personal information provided. For Advancement Act
Quality Models (‘‘Guideline’’). The detailed instructions on sending J. Executive Order 12898: Federal Actions
Guideline has been incorporated into To Address Environmental Justice in
comments and additional information Minority Populations and Low-Income
EPA’s regulations, satisfying a on the rulemaking process, see the Populations
requirement under the Clean Air Act ‘‘Public Participation’’ heading of the
(CAA) section 165(e)(3)(D) for the EPA SUPPLEMENTARY INFORMATION section of I. General Information
to specify, with reasonable particularity, this document.
models to be used in the Prevention of A. Does this action apply to me?
The public hearing will be held at 109
Significant Deterioration (PSD) program. T.W. Alexander Drive, Research This action applies to Federal, State,
It provides EPA-preferred models and Triangle Park, North Carolina 27711. territorial, and local air quality
other recommended techniques, as well The hearing will convene at 8:30 a.m. management programs that conduct air
as guidance for their use in predicting (local time) and will conclude at 5:00 quality modeling as part of State
ambient concentrations of air pollutants. p.m. (local time). Refer to the Implementation Plan (SIP) submittals
In this action, the EPA is proposing SUPPLEMENTARY INFORMATION section and revisions, New Source Review
revisions to the Guideline, including below for additional information. (NSR), including new or modifying
enhancements to the formulation and industrial sources under Prevention of
FOR FURTHER INFORMATION CONTACT: Mr.
application of the EPA’s near-field Significant Deterioration (PSD),
George M. Bridgers, Office of Air
dispersion modeling system, AERMOD, Conformity, and other air quality
Quality Planning and Standards, Air
and updates to the recommendations for assessments required under EPA
Quality Assessment Division, Air
the development of appropriate regulation. Categories and entities
background concentration for Quality Modeling Group, U.S.
potentially regulated by this action
cumulative impact analyses. Within this Environmental Protection Agency, Mail
include:
action, the EPA is also announcing the code C439–01, Research Triangle Park,
Thirteenth Conference on Air Quality NC 27711; telephone: (919) 541–5563; NAICS a
email: [email protected]. (and Category
Modeling and invites the public to code
participate in the conference. The include ‘‘2023 Revisions to the
Guideline on Air Quality Models’’ in the Federal/State/territorial/local/Trib-
conference will focus on the proposed al government ......................... 924110
revisions to the Guideline, and part of subject line of the message).
the conference will also serve as the SUPPLEMENTARY INFORMATION: a North American Industry Classification
public hearing for these revisions. The information in this preamble is System.
DATES: Comments must be received on organized as follows: B. Where can I get a copy of this
or before December 22, 2023. Table of Contents document?
Public hearing and conference: The
I. General Information In addition to being available in the
public hearing for this action and the
A. Does this action apply to me? docket, an electronic copy of this
Thirteenth Conference on Air Quality
B. Where can I get a copy of this proposed rule and relative supporting
Modeling will be held November 14–15, document? documentation will also be available on
2023, from 8:30 a.m. to 5:00 p.m. II. Background EPA’s Support Center for Regulatory
Eastern Standard Time (EST). A. The Guideline on Air Quality Models Atmospheric Modeling (SCRAM)
ADDRESSES: Submit your comments, and EPA Modeling Conferences
website. Following signature, these
identified by Docket ID No. EPA–HQ– B. The Twelfth Conference on Air Quality
Modeling materials will be posted on SCRAM at
OAR–2022–0872, by one of the the following address: https://
following methods: C. Alpha and Beta Categorization of Non-
Regulatory Options www.epa.gov/scram/13th-conference-
• Federal eRulemaking Portal:
lotter on DSK11XQN23PROD with PROPOSALS2
III. Public Participation air-quality-modeling.

https://www.regulations.gov. Follow the
A. Written Comments
online instructions for submitting B. Notice of Public Hearing and the II. Background
comments. Thirteenth Conference on Air Quality
• Email: [email protected]. A. The Guideline on Air Quality Models
Models and EPA Modeling Conferences
Include Docket ID No. EPA–HQ–OAR– IV. Proposed Revisions to the Guideline
2022–0872 in the subject line of the A. Proposed Revisions The Guideline is used by the EPA,
message. V. Ongoing Model Development other Federal, State, territorial, and local
VerDate Sep<11>2014 18:34 Oct 20, 2023 Jkt 262001 PO 00000 Frm 00002 Fmt 4701 Sfmt 4702 E:\FR\FM\23OCP2.SGM 23OCP2
Federal Register / Vol. 88, No. 203 / Monday, October 23, 2023 / Proposed Rules 72827
air quality agencies, and industry to the EPA revised the Guideline on April version of the Guideline. The conference
prepare and review preconstruction 15, 2003 (68 FR 18440), to adopt included presentations summarizing the
permit applications for new sources and CALPUFF as the preferred model for proposed updates to the AERMOD
modifications, SIP submittals and long-range transport of emissions from Modeling System, replacement of
revisions, determinations that actions by 50 to several hundred kilometers and to CALINE3 with AERMOD for modeling
Federal agencies are in conformity with make various editorial changes to of mobile sources, incorporation of
SIPs, and other air quality assessments update and reorganize information and prognostic meteorological data for use
required under EPA regulation. The remove obsolete models. in dispersion modeling, the proposed
Guideline serves as a means by which We held the Eighth Conference on Air screening approach for long-range
national consistency is maintained in Quality Modeling in September 2005. transport for NAAQS and PSD
air quality analyses for regulatory This conference provided details on increments assessments with use of
activities under CAA regulations, changes to the preferred air quality CALPUFF as a screening technique
including 40 CFR 51.112, 51.117, models, including available methods for rather than an EPA-preferred model, the
51.150, 51.160, 51.165, 51.166, 52.21, model performance evaluation and the proposed 2-tiered screening approach to
93.116, 93.123, and 93.150. notice of data availability that the EPA address ozone and PM2.5 in PSD
The EPA originally published the published in September 2003, related to compliance demonstrations, the status
Guideline in April 1978 (EPA–450/2– the incorporation of the PRIME and role of the Model Clearinghouse,
78–027), and it was incorporated by downwash algorithm in the AERMOD and updates to procedures for single-
reference in the regulations for the PSD dispersion model (in response to source and cumulative modeling
program in June 1978. The EPA revised comments received from the Seventh analyses (e.g., modeling domain, source
the Guideline in 1986 (51 FR 32176) and Conference). Additionally, at the Eighth input data, background data, and
updated it with supplement A in 1987 Conference, a panel of experts discussed compliance demonstration procedures).
(53 FR 32081), supplement B in July the use of state-of-the-science prognostic Additionally, the 2015 proposed
1993 (58 FR 38816), and supplement C meteorological data for informing the action included a reorganization of the
in August 1995 (60 FR 40465). The EPA dispersion models. The EPA further Guideline to make it easier to use and
published the Guideline as Appendix W revised the Guideline on November 9, to streamline the compliance
to 40 CFR part 51 when the EPA issued 2005 (70 FR 68218), to adopt AERMOD assessment process (80 FR 45340), and
supplement B. The EPA republished the as the preferred model for near-field also included additional clarity in
Guideline in August 1996 (61 FR 41838) dispersion of emissions for distances up distinguishing requirements from
to adopt the CFR system for labeling to 50 kilometers. recommendations while noting the
paragraphs. The publication and The Ninth Conference on Air Quality continued flexibilities provided within
incorporation of the Guideline by Modeling was held in October 2008 and the Guideline, including but not limited
reference into the EPA’s PSD regulations emphasized the following topics: to use and approval of alternative
satisfies the requirement under the CAA reinstituting the Model Clearinghouse, models (82 FR at 45344). These
section 165(e)(3)(D) for the EPA to review of non-guideline applications of proposed revisions were adopted and
promulgate regulations that specify with dispersion models, regulatory status reflected in the latest version of the
reasonable particularity models to be updates of AERMOD and CALPUFF, Guideline, promulgated on January 17,
used under specified sets of conditions continued discussions on the use of 2017 (82 FR 5182).
for purposes of the PSD program. prognostic meteorological data for
To support the process of developing informing dispersion models, and B. The Twelfth Conference on Air
and revising the Guideline during the presentations reviewing the available Quality Modeling
period of 1977 to 1988, we held the model evaluation methods. To further The most recent EPA modeling
First, Second, and Third Conferences on inform the development of additional conference was the Twelfth Conference
Air Quality Modeling as required by revisions to the Guideline, we held the on Air Quality Modeling, which was
CAA section 320 to help standardize Tenth Conference on Air Quality held in August 2019 in continuing
modeling procedures. These modeling Modeling in March 2012. The compliance with CAA section 320.
conferences provided a forum for conference addressed updates on: the While not associated with a regulatory
comments on the Guideline and regulatory status and future action, the Twelfth Conference was held
associated revisions, thereby helping us development of AERMOD and with the intent to inform the ongoing
introduce improved modeling CALPUFF, review of the Mesoscale development of EPA’s preferred air
techniques into the regulatory process. Model Interface (MMIF) prognostic quality models and potential revisions
Between 1988 and 1995, we conducted meteorological data processing tool for to the Guideline. The conference
the Fourth, Fifth, and Sixth Conferences dispersion models, draft modeling included expert panel discussions and
on Air Quality Modeling to solicit guidance for compliance invited presentations covering the
comments from the stakeholder demonstrations of the fine particulate following model/technique
community to guide our consideration matter (PM2.5) National Ambient Air enhancements: treatment of low wind
of further revisions to the Guideline, Quality Standards (NAAQS), modeling conditions, overwater modeling, mobile
update the available modeling tools for compliance demonstration of the 1- source modeling, building downwash,
based on the current state-of-the- hour nitrogen dioxide (NO2) and sulfur prognostic meteorological data, near-
science, and advise the public on new dioxide (SO2) NAAQS, and new and field and long-range model evaluation
modeling techniques. emerging models/techniques for future criteria, NO2 modeling techniques,
The Seventh Conference was held in plume rise, deposition, and single
consideration under the Guideline to

June 2000 and also served as a public address single-source modeling for source ozone and PM2.5 modeling
hearing for the proposed revisions to the ozone and secondary PM2.5, as well as techniques. At the conclusion of the
recommended air quality models in the long-range transport and chemistry. expert panels and invited presentations,
Guideline (65 FR 21506). These changes The Eleventh Conference on Air there were several presentations given
included the CALPUFF modeling Quality Modeling was held August 12– by the public, including industrial trade
system, AERMOD Modeling System, 13, 2015, and included the public groups, on recommended areas for
and ISC–PRIME model. Subsequently, hearing for the most recently proposed additional model development and
72828 Federal Register / Vol. 88, No. 203 / Monday, October 23, 2023 / Proposed Rules
future revision in the Guideline. The still in the research and development The written comment is considered the
proposed regulatory updates to the stage. Alpha options have not yet been official comment and should include
AERMOD Modeling System in this fully tested, evaluated, or vetted through discussion of all points you wish to
action address topics on which there peer review and should not be make. The EPA will generally not
was focused discussion and engagement considered for use as an alternative consider comments or comment
with the stakeholder community model for regulatory applications of the contents located outside of the primary
through these expert panels and invited model. submission (i.e., on the web, cloud, or
and public presentations during the Beta options, on the other hand, have other file sharing system). Please visit
Twelfth Conference. been demonstrated to be applicable on https://www.epa.gov/dockets/
All the presentations, along with the a theoretical basis, have undergone commenting-epa-dockets for additional
transcript of the conference scientific peer review, and are submission methods; the full EPA
proceedings, are available in the docket supported with performance evaluations public comment policy; information
for the Twelfth Conference on Air using available and adequate databases about CBI, PBI, or multimedia
Quality Models (Docket ID No. EPA– that demonstrate unbiased, improved submissions; and general guidance on
HQ–OAR–2019–0454). Additionally, all model performance. In general, beta making effective comments.
the materials associated with the options have met the necessary criteria
Twelfth Conference and the public to be formally proposed and adopted as B. Notice of Public Hearing and the
hearing are available on the EPA’s updates to the regulatory version of the Thirteenth Conference on Air Quality
SCRAM website at https://www.epa.gov/ model but have not yet been proposed Models
scram/12th-conference-air-quality- through the required rulemaking The public hearing for this action and
modeling. process, which includes a public the Thirteenth Conference on Air
hearing and formal comment period. Quality Modeling will be held on
C. Alpha and Beta Categorization of November 14–15, 2023, in the EPA
Beta options are mature enough in the
Non-Regulatory Options Nantahala Auditorium, Room C111, 109
development process to be considered
With the release of AERMOD version for use as an alternative model, T.W. Alexander Drive, Research
18181 in 2018, the EPA adopted a new provided an appropriate site-specific Triangle Park, NC 27711. The hearing
paradigm for engagement with the modeling demonstration is completed to and conference will convene each day at
scientific community to facilitate the show the alternative model is 8:30 a.m. EST and will conclude at 5:00
continued development of the AERMOD appropriate for the site and conditions p.m. EST.
Modeling System. Previously, updates where it will be applied and the The Thirteenth Conference on Air
to the scientific formulation of the requirements of the Guideline, section Quality Modeling will be open to the
model were not made available to the 3.2, are fully satisfied, including formal public. No registration fee is charged.
public for review, testing, evaluation, concurrence by the EPA’s Model The conference will be formally
and comment prior to the proposal stage Clearinghouse. conducted and chaired by an EPA
of the formal rulemaking process when official. As required under CAA section
an update was made to the Guideline. III. Public Participation 320, a verbatim transcript of the
This limited the public’s engagement Interested persons may provide the conference proceedings will be
and feedback to a short, predefined EPA with their views on the proposed produced and placed in the docket for
comment period, typically only one to revisions to the Guideline in several this proposed action. The conference
two months. The new approach enables ways. This includes submitting written will begin with introductory remarks by
the EPA to release potential formulation comments to the EPA, participating in the presiding EPA official. The EPA staff
updates as non-regulatory ‘‘alpha’’ and the Thirteenth Conference on Air and EPA invited speakers will then
‘‘beta’’ options as they are being Quality Modeling, and speaking at the provide a structured overview of the
developed. As non-regulatory options, public hearing that will be conducted as revisions to the Guideline as proposed
they can be made available during any part of the conference. Additional in this document and present on the
release cycle, thereby enabling feedback information on where to submit written research that supports those revisions
as they are being developed. This comments on the proposed revisions to and supports formulation updates to the
approach allows for more robust testing the Guideline is provided in the preferred models. The following topics
and evaluation during development, ADDRESSES section above. will be presented:
benefitting from the experience of a I. Overview of the Thirteenth Conference
broad expert community. In addition, A. Written Comments
on Air Quality Modeling;
the EPA developed a protocol to enable Submit your comments, identified by II. Review of the proposed revisions to the
the external community to submit Docket ID No. EPA–HQ–OAR–2022– preferred air quality models; and
model updates to the EPA for review 0872, at https://www.regulations.gov III. Review of the proposed revisions to the
and consideration for inclusion as new (our preferred method), or the other Guideline.
alpha or beta options. A pathway such methods identified in the ADDRESSES At the conclusion of these
as this that facilitates more frequent and section. Once submitted, comments presentations, the EPA will convene the
active engagement with the external cannot be edited or removed from the public hearing on the proposed
community allows for a more informed docket. The EPA may publish any revisions to the Guideline. The public
and timely regulatory update process comment received to its public docket. hearing will span a portion of the
when the EPA has determined an Do not submit to EPA’s docket at afternoon of the first day and
https://www.regulations.gov any throughout the second day of the
update has met the criteria required for

consideration as a science formulation information you consider to be conference. The EPA will make every
update to the regulatory version of the Confidential Business Information (CBI), effort to follow the schedule as closely
model. Proprietary Business Information (PBI), as possible on the days of the
In this alpha/beta construct, alpha or other information whose disclosure is conference; however, please plan for the
options are updates to the scientific restricted by statute. Multimedia public hearing to run either ahead of
formulation that are thought to have submissions (audio, video, etc.) must be schedule or behind schedule. The EPA
merit but are considered experimental, accompanied by a written comment. may close the hearing 15 minutes after
the last pre-registered speaker has REAL ID Act, passed by Congress in 3, 2019,1 and additional relevant
testified on November 15, if there are no 2005, established new requirements for research since 2017, the EPA and other
additional speakers. entering Federal facilities. For purposes researchers have conducted additional
Those wishing to reserve time to of the REAL ID Act, EPA will accept model evaluations and developed
speak at the public hearing, whether to government-issued IDs, including changes to the model formulation of the
offer specific comments on the proposed drivers’ licenses, from the District of AERMOD Modeling System to improve
rule, volunteer a presentation on a Columbia and all States and territories model performance in its regulatory
special topic, or to offer except from American Samoa. If your applications. One update is to the
recommendations on any regulatory identification is issued by American AERMET meteorological preprocessor
modeling techniques, should contact us Samoa, you must present an additional for AERMOD. This update provides the
at the address given in the FOR FURTHER form of identification to enter the capability to process measured and
INFORMATION CONTACT section by no later prognostic marine-based meteorology
Federal building where the public
than November 10, 2023. Such persons hearing will be held. Acceptable for offshore applications. Separate
should identify the organization (if any) updates are related to the AERMOD
alternative forms of identification
on whose behalf they are speaking and dispersion model and include (1) a new
include Federal employee badges,
the length of the presentation. If a Tier 3 screening method for the
passports, enhanced driver’s licenses,
scheduled presentation is projected to conversion of nitrogen oxides (NOX)
be longer than 10 minutes, the presenter and military identification cards. For
additional information for the status of emissions to NO2 and (2) a new source
should also state why a longer period is type for modeling vehicle roadway
needed. Scheduled speakers should your State regarding REAL ID, go to:
https://www.dhs.gov/real-id- emissions.
bring extra copies of their presentation Each of the proposed formulation
for inclusion in the docket and for the enforcement-brieffrequently-asked-
updates to the AERMOD Modeling
convenience of the recorder. Scheduled questions. Any objects brought into the
System is provided as a non-regulatory
speakers will also be permitted to enter building need to fit through the security
beta option in the release of the relevant
additional written comments into the screening system, such as a purse,
modeling system components that is
record. laptop bag, or small backpack.
occurring concurrent with this proposed
Any person in attendance wishing to Demonstrations will not be allowed on
rule. If EPA adopts these formulation
speak at the public hearing who has not Federal property for security reasons.
updates in a subsequent final rule, the
reserved time in advance may provide Attendees are encouraged to arrive at beta categorization would be removed
oral comments on the proposed least 15 minutes prior to the start of the and the respective model option(s)
revisions to the Guideline during time meeting to allow enough time for could be considered regulatory model
allotted on the last day. These parties security screening. options.
will need to sign up to speak on the The EPA proposes the following
second day of the hearing, and the EPA IV. Proposed Revisions to the Guideline
updates to the AERMOD Modeling
may need to limit the duration of In this action, the EPA is proposing System to address several technical
presentations to allow all participants to updates to the Guideline corresponding concerns expressed by stakeholders:
be heard. to updates to the scientific formulation
The EPA may ask clarifying questions a. Incorporation of COARE Algorithms
of the AERMOD Modeling System and
during the oral presentations but will Into AERMET for Use in Overwater
not respond to the presentations at that updates to the recommendations for the
Marine Boundary Layer Environments
time. Information submitted to the EPA development of appropriate background
concentration for cumulative impact As the number of overwater
during the public hearing will be placed applications has increased in recent
in the docket for this proposed action. analyses. When and where appropriate,
the EPA has engaged with our Federal years, the EPA is proposing to add the
Written statements and supporting Coupled Ocean-Atmosphere Response
information submitted during the partners, including the Bureau of Ocean
Energy Management (BOEM) and the Experiment (COARE) 2 3 algorithms to
comment period will be considered AERMET for meteorological data
with the same weight as oral testimony Federal Highway Administration
(FHWA), to collaborate on these processing in applications using either
and supporting information presented at observed or prognostic meteorological
the public hearing. proposed updates to the Guideline.
There are additional editorial changes data in overwater marine boundary
Conference background information.
proposed to the Guideline to correct layer environments. One of the first
Preregistration details, additional
minor typographical errors found in the notable uses of AERMOD for an
background information, and a more
2017 Guideline and update website overwater application was an alternative
detailed agenda for the Thirteenth
links. model application—AERMOD–COARE
Conference on Air Quality Modeling are
was used in 2011 in an ice-free arctic
electronically available at https://
A. Proposed Revisions environment of Alaska.4 5 In this
www.epa.gov/scram/13th-conference-
air-quality-modeling. Preregistration for This section provides a detailed 1 https://www.epa.gov/scram/12th-conference-air-
the conference, while not required, is overview of the substantive proposed quality-modeling.
strongly recommended due to changes to the Guideline that are 2 Fairall, C.W., E.F. Bradley, J.E. Hare, A.A.
heightened security protocols at the intended to improve the science of the Grachev, and J.B. Edson, 2003: ‘‘Bulk
EPA–RTP facility. Parameterization of Air-Sea Fluxes: Updates and
models and approaches used in Verification for the COARE Algorithm.’’ Journal of
Access to U.S. government facility.
regulatory assessments. Climate, 16, 571–591.

Because this hearing is being held at a 3 Evaluation of the Implementation of the
U.S. government facility, individuals 1. Proposed Updates to EPA’s AERMOD Coupled Ocean-Atmosphere Response Experiment
planning to attend the conference and/ Modeling System (COARE) algorithms into AERMET for Boundary
or public hearing should be prepared to Layer Environments. EPA–2023/R–23–008, Office
Based on studies presented and of Air Quality Planning and Standards, RTP, NC.
show valid picture identification to the 4 U.S. EPA, 2011: COARE Bulk Flux Algorithm to
security staff in order to gain access to discussed at the Twelfth Conference on Generate Hourly Meteorological Data for Use with
the meeting room. Please note that the Air Quality Models held on October 2– Continued
application, the incorporation of the elimination is contingent upon primary motivation behind the
COARE bulk flux algorithm was used as consultation with the EPA Regional formulation and development of the
an alternative to the AERMET Office and appropriate reviewing GRSM NO2 screening option was to
meteorological processor to AERMOD. authority. This consultation will ensure address photolytic conversion of NO2 to
This led to the development of the that platform downwash and shoreline NO and to address the time-of-travel
AERCOARE 6 processor that can be used fumigation are adequately considered in necessary for NOX plumes to convert the
with either measured or prognostic data the modeling demonstration. NO portion of the plume to NO2 via
for overwater applications in lieu of titration and entrainment of ambient
b. Proposed Addition of a New Tier 3
AERMET. AERCOARE has been ozone. The existing regulatory non-
Detailed Screening Technique for NO2
approved as an alternative model for default Tier 3 NO2 screening options,
several overwater applications since Section 4.2.3.4 of the 2017 Guideline PVMRM and OLM, do not address or
2011.7 details a 3-tiered approach for provide for treatment of these
For overwater applications, the evaluating the modeled impacts of NOX mechanisms, and have been shown to
algorithms in COARE are better suited sources, which was recommended to over-predict for some source
for overwater boundary layer assess hourly and annual average NO2 characterizations and model
calculations than the existing algorithms impacts from point, volume, and area configurations at project source ambient
in AERMET that are better suited for sources for the purposes of the PSD air boundaries and within the first 1–3
land-based data. These calculations program, SIP planning, and km.13
include calculation of surface transportation general conformity. This The functionality of the GRSM
roughness, stability classification, 3-tiered approach addresses the co- implementation in AERMOD is similar
effects of moisture on Monin-Obukhov emissions of NO and NO2 and the to that of the PVMRM and OLM
length, and the use of Bowen ratio by subsequent conversion of NO to NO2 in schemes, with exception to some
AERMET for heat flux calculations.5 the atmosphere. The tiered levels additional input requirements necessary
The EPA proposes to add COARE to include: for treatment of the reverse NO2
AERMET in order to ensure that the Tier 1—assuming that all emitted NO photolysis reaction during daytime
COARE algorithms are updated is converted to NO2 (full conversion). hours. Modeled source inputs for GRSM
Tier 2—using the Ambient Ratio require NO2/NOX in-stack ratios, with
regularly as part of routine AERMET
Method 2 (ARM2), which applies an similar assumptions as applied to
updates, to provide consistent data
assumed equilibrium ratio of NO2 to PVMRM and OLM according to section
handling among land based and
NOX, based on analysis of and 4.2.3.4 of the Guideline. Ambient inputs
overwater based meteorological data
correlation with nationwide hourly for GRSM require hourly ozone
(e.g., treatment of missing data and
observed ambient conditions. concentrations taken from an
treatment of calms), and to have all Tier 3—applying the Ozone Limiting
meteorological processing for AERMOD appropriately representative monitoring
Method (OLM) and Plume Volume station or selection of monitoring
applications in one program. Molar Ratio (PVMRM) screening options
The addition of the COARE stations for varying upwind sector
based on site-specific hourly ozone data concentrations. GRSM also requires
algorithms to AERMET would replace and source-specific NO2 to NOX in-stack
the standalone AERCOARE program and hourly NOX concentration inputs to
ratios.8 9 10 11 resolve the daytime photolysis of NO2
the AERCOARE output option in MMIF As further discussed in section
for prognostic data overwater. This reaction in equilibrium with ozone
4.2.3.4(e) of the Guideline, regulatory titration conversion of the NO portion of
proposed option is selected by the user application of Tier 3 screening options
with the METHOD COARE RUN– the NOX plume. GRSM hourly NOX
shall occur in consultation with the EPA concentration inputs can also vary by
COARE record in the AERMET Stage 2 Regional Office and appropriate
input file. For prognostic applications upwind sector concentration, as
reviewing authority. appropriate. Background NO2
processed through the MMIF, the user The EPA proposes to include the
can run MMIF for AERMET input for concentrations are accounted for in the
Generic Reaction Set Method (GRSM) as
overwater applications. GRSM daytime equilibrium NO2
a regulatory non-default Tier 3 NO2
The addition of COARE to AERMET concentration estimates based on the
screening option. Following a peer-
would eliminate the previous chemical reaction balance between
reviewed publication in 2017, GRSM
alternative model demonstration ozone entrainment and NO titration,
was added to AERMOD as an alpha
requirements for use of AERMOD in photolysis of NO2 to NO, and ambient
option in version 21112 and updated as
marine environments, and this background NO2 participation in
a beta option in version 22112.12 The titration and photolysis reactions.
the AERMOD dispersion program; Section 3.2.2.e 8 Podrez, M. 2015. An Update to the Ambient
Nighttime GRSM NO2 estimates are
Alternative Refined Model Demonstration Herman Ratio Method for 1-h NO2 Air Quality Standards based on ozone entrainment and
Wong Memorandum dated April 1, 2011, Office of Dispersion Modeling. Atmospheric Environment, titration of available NO in the NOX
Environmental Assessment, Region 10, Seattle, 103: 163–170. plume. Note that all hourly ozone and
Washington 98101. 9 Cole, H.S. and J.E. Summerhays, 1979. A Review
5 U.S. EPA, 2011: Model Clearinghouse Review
NOX ambient inputs to GRSM must
of Techniques Available for Estimation of Short-
AERMOD–COARE as an Alternative Model in an Term NO2 Concentrations. Journal of the Air coincide with the hourly meteorological
Arctic Ice Free Environment. George Bridgers Pollution Control Association, 29(8): 812–817.
Memorandum dated May 6, 2011, Office of Air 10 Hanrahan, P.L., 1999. The Polar Volume Polar Evaluation of an explicit NOX chemistry method in
Quality Planning and Standards, Research Triangle Ratio Method for Determining NO2/NOX Ratios in AERMOD, Journal of the Air & Waste Management
Park, North Carolina 27711. Modeling—Part I: Methodology. Journal of the Air Association, 67:6, 702–712, DOI: 10.1080/
6 U.S. EPA, 2012: User’s Manual AERCOARE & Waste Management Association, 49: 1324–1331. 10962247.2017.1280096.
Version 1.0. EPA–910–R–12–008. U.S. EPA, Region 11 Chu, S.H. and E.L. Meyer, 1991. Use of 13 Jenny Stocker, Martin Seaton, Stephen Smith,
10, Seattle, WA. Ambient Ratios to Estimate Impact of NOX Sources James O’Neill, Kate Johnson, Rose Jackson, David
7 Please reference the EPA Model Clearinghouse on Annual NO2 Concentrations. Proceedings, 84th Carruthers (CERC). Evaluation of the Generic
Information Storage and Retrieval System Annual Meeting & Exhibition of the Air & Waste Reaction Set Method for NO2 conversion in
(MCHISRS) database for more information regarding Management Association, Vancouver, B.C.; 16–21 AERMOD. The modification of AERMOD to include
AERCOARE alternative model approvals (https:// June 1991. (16pp.) (Docket No. A–92–65, II–A–9). ADMS chemistry. August 8, 2023. Cambridge
cfpub.epa.gov/oarweb/MCHISRS, text Search term 12 David J. Carruthers, Jenny R. Stocker, Andrew Environmental Research Consultants (CERC)
‘‘AERCOARE’’). Ellis, Martin D. Seaton & Stephen E. Smith (2017) Technical Report.
data records for the period of the The RLINE source type for this concentration, for use in NAAQS
modeling analysis (i.e., minimum of 1 proposed action has undergone implementation modeling
year for on-site data, 3 years of significant evaluation by the EPA and demonstrations (e.g., PSD compliance
prognostic data, and 5 years of airport the FHWA as part of the Interagency demonstrations, SIP demonstrations for
data). Agreement and has shown improved inert pollutants, and SO2 designations).
Updates to the GRSM formulation in performance.17 18 This proposed option The Guideline recommends that a
AERMOD version 22112 were is selected by the model user with the representative background
developed in late 2022 to address more SOURCE type ‘‘RLINE’’. In addition to concentration should include
realistic building effects on proposing RLINE as a new source type, contributions from all sources,
instantaneous plume spread, accounting the EPA is also proposing the use of the including both nearby and other
of multiple plume effects on AERMOD urban option (accounting for sources. When identifying nearby
entrainment of ozone, and the tendency urban heat island effect in stable sources that may not be adequately
of GRSM to over-predict in the far-field conditions) and terrain with the RLINE represented by ambient monitoring data,
(e.g., beyond approximately 3 km for source type. However, the inclusion of the Guideline recommends selecting
typical point source releases). terrain with RLINE does not supersede sources ‘‘that cause a significant
Sensitivity testing and model the EPA’s PM Hot-spot guidance where concentration gradient in the vicinity of
performance evaluations of these FLAT terrain is recommended for the source(s) under consideration.’’ The
updates to GRSM in AERMOD version modeling applications.19 The EPA also EPA recognizes that the recommended
23132 have shown consistent or emphasizes that the inclusion of RLINE method for identifying nearby sources
improved model behavior and as a source type for mobile source lacks specificity, is used and referenced
performance.14 modeling does not preclude the use of inconsistently, and may lead to overly
the existing AREA, LINE, and VOLUME conservative modeling exercises. The
c. Proposed Addition of RLINE as source types thereby extending the proposed revisions to section 8 are
Mobile Source Type flexibility of users in best characterizing intended to provide a more robust
mobile source for regulatory modeling. framework for characterizing
As a culmination of an Interagency
Agreement between EPA and FHWA, d. Support Information, Documentation, background concentrations for
the EPA proposes to add the RLINE and Model Code cumulative modeling with particular
source type as a new source type attention to identifying and modeling
Model performance evaluation and nearby sources in multi-source areas.
applicable for regulatory modeling of peer-reviewed scientific references for The EPA proposes to revise
mobile sources. This is in addition to each of these three proposed updates to recommendations for the determination
the AREA, LINE, and VOLUME source the AERMOD Modeling System are of background concentrations in
types already available for mobile cited and placed in the docket, as constructing the design concentration,
source modeling. The proposed addition appropriate. An updated user’s guide or total air quality concentration in
of RLINE as a mobile source type is an and model formulation documents for multi-source areas (see section 8.3), as
extension of the 2017 update to the version 23132 have also been placed in part of a cumulative impact analysis for
Guideline in which AERMOD replaced the docket. We have updated the NAAQS implementation modeling
CALINE3 as the addendum A 15 model summary description of the AERMOD demonstrations. The EPA’s proposed
for mobile source modeling. At that Modeling System to addendum A of the framework includes a stepwise set of
time, AERMOD’s AREA, LINE, and Guideline to reflect these proposed considerations to replace the narrow
VOLUME sources were available for updates. The essential codes, recommendation of modeling nearby
mobile source modeling. The basis of preprocessors, and test cases have been sources that cause a significant
the RLINE source type is the EPA’s updated and posted to the EPA’s concentration gradient. This framework
Office of Research and Development SCRAM website, https://www.epa.gov/ focuses the inherent discretion in
(ORD) Research LINE (RLINE) model 16 scram. defining representative background
released in 2013. The RLINE model was
2. Proposed Updates to concentrations through qualitative and
designed for near-surface releases to
Recommendations on the Development semi-quantitative considerations within
simulate mobile source dispersion with
of Background Concentration a transparent process using the variety
an emphasis on the near-road
Based on permit modeling of emissions and air quality data
environment. The RLINE model was
experiences since the 2017 revisions to available to the permit applicant. To
first incorporated into AERMOD as a
the Guideline, the EPA proposes construct a background concentration
beta source type in AERMOD version
revisions to section 8 of the Guideline for model input under the framework,
19191 in 2019.
to refine the recommendations regarding permit applicants should consider the
14 Environmental Protection Agency, 2023. the determination of appropriate model representativeness of relevant
Technical Support Document (TSD) for Adoption of input data, specifically background emissions, air quality monitoring, and
the Generic Reaction Set Method (GRSM) as a pre-exiting air quality modeling to
Regulatory Non-Default Tier-3 NO2 Screening 17 Incorporation and Evaluation of the RLINE appropriately represent background
Option, Publication No. EPA–454/R–23–009. Office source type in AERMOD for Mobile Source concentrations for the cumulative
of Air Quality Planning & Standards, Research Applications. EPA–2023/R–23–011, Office of Air
Triangle Park, NC.
impact analysis.
Quality Planning and Standards, RTP, NC. In conjunction with the proposed
15 Under the codification requirements of the 18 Heist, D., et al., 2023. Integration of RLINE
Administrative Committee of the Federal Register dispersion model into EPA’s AERMOD: updated
revisions to section 8 of the Guideline,
(ACFR), only subparts, parts, subchapters, and formulation and evaluations. Journal of the Air & the EPA developed the Draft Guidance
chapters may have appendices. Therefore, we have Waste Management Association, Manuscript on Developing Background
changed the naming convention from ‘‘appendix A’’ submitted for publication.
to ‘‘addendum A’’. 19 U.S. EPA, 2021: PM Hot-spot Guidance;
Concentrations for Use in Modeling
16 Snyder, M.G., Venkatram, A., Heist, D.K.,
Transportation Conformity Guidance for
Demonstrations.20 This draft guidance
Perry, S.G., Petersen, W.B. and Isakov, V., 2013. Quantitative Hot-spot Analyses in PM2.5 and PM10
RLINE: A line source dispersion model for near- Nonattainment and Maintenance Areas. EPA–42–B– 20 U.S. Environmental Protection Agency, 2023.
surface releases. Atmospheric environment, 77, 21–037. U.S. EPA, Office of Transportation and Air Draft Guidance on Developing Background
pp.748–756. Quality, Ann Arbor, MI. Continued
document details the EPA- 4. Proposed Revisions by Section d. Section 5.0—Models for Ozone and
recommended framework with a. Section 1.0—Introduction Secondarily Formed Particulate Matter
illustrative examples to assist permit The EPA proposes to update reference
applicants in characterizing a credible The EPA proposes to correct numbers where necessary due to added
and appropriately representative paragraph (i) by combining the references. In section 5.2, the EPA
background concentration for inadvertently created paragraph (A), proposes to revise paragraph (c) to
cumulative impact analyses including which is actually part of the phrase include a reference for guidance on the
the contributions from nearby sources in ‘‘addendum A’’ in the first sentence. use of models to assess the impacts of
multi-source areas. emissions from single sources on
b. Section 3.0—Preferred and
3. Transition Period for Applicability of Alternative Air Quality Models secondarily formed ozone and PM2.5.
Revisions to the Guideline e. Section 6.0—Modeling for Air Quality
The EPA proposes to revise an
In previous rulemakings to revise the outdated website link in section 3.0(b). Related Values and Other Governmental
Guideline, we have traditionally Programs
In sections 3.1.1(c) and 3.1.2(a), the
communicated that it would be EPA proposes to correct the sections by The EPA proposes to update reference
appropriate to provide 1-year to combining the inadvertently created numbers where necessary due to added
transition to the use of new models, paragraph (A), which is actually part of references and revise an outdated
techniques and procedures in the the phrase ‘‘addendum A’’ in the first website link in section 6.3(a).
context of PSD permit applications and sentence.
other regulatory modeling applications. f. Section 7.0—General Modeling
We invite comments whether it would c. Section 4.0—Models for Carbon Considerations
be appropriate to apply a 1-year Monoxide, Lead, Sulfur Dioxide, The EPA proposes to update reference
transition after promulgation of the Nitrogen Dioxide and Primary numbers where necessary due to added
revised Guideline (i.e., from its effective Particulate Matter references.
date) such that applications conducted In section 7.2.3, the EPA proposes to
The EPA proposes to update reference revise paragraph (b) to include the
under the existing Guideline with numbers where necessary due to added
approved protocols would be acceptable addition of RLINE as a source type for
references. use in regulatory applications of
during that period, but new
requirements and recommendations In sections 4.1(b) and 4.2.2(a), the AERMOD and remove references to
should be used for applications EPA proposes to correct the sections by specific distances that receptors can be
submitted after that period or protocols combining the inadvertently created placed from the roadway.
approved after that period. paragraph (A), which is actually part of Also in section 7.2.3, the EPA
Such a transition period is consistent the phrase ‘‘addendum A’’ in the first proposes to revise paragraph (c) to
with previous revisions to the Guideline sentence. include RLINE as a source type that can
and appropriate to avoid the time and In section 4.2.2.1, the EPA proposes to be used to model mobile sources and
expense of revisiting modeling that is add a new paragraph (f) regarding the clarify that an area source can be
substantially complete, which would use of AERMOD in certain overwater categorized in AERMOD using the
cause undue delays to permit situations. A typographical correction is AREA, LINE, or RLINE source type.
applications that are pending when the proposed in section 4.2.2.1(b). g. Section 8.0—Model Input Data
proposed revisions to the Guideline are The EPA proposes amendments to
section 4.2.2.3 to account for The EPA proposes to update reference
finalized. The proposed revisions to the
circumstances where OCD is available numbers where necessary due to added
Guideline are intended as incremental
to evaluate situations where shoreline references.
improvements to the Guideline, and
The EPA proposes to revise Table 8–
such improvements do not necessarily fumigation and/or platform downwash
1 and Table 8–2 to correct typographical
invalidate past practices under the are important.
errors and update the footnotes in each
previous editions of the Guideline. The In section 4.2.3.4, the EPA proposes to of the tables.
requirements and recommendations in revise paragraph (e) to adopt the Generic The EPA proposes to revise section
the existing (2017) version of the Reaction Set Method (GRSM) as a 8.3.1 to address current EPA practices
Guideline were previously identified as regulatory Tier 3 detailed screening and recommendations for determining
acceptable by the EPA, and they will technique for NO2 modeling the appropriate background
continue to be acceptable for air quality demonstrations. Sentences in this concentration as model input data for a
assessments during the period of section would be updated to incorporate new or modifying source(s) or sources
transition to the revised version of the GRSM with the existing regulatory Tier under consideration for a revised permit
Guideline, if finalized. 3 screening techniques OLM and limit. This revision would provide a
Where a proposed revision to the PVMRM. An additional statement is stepwise framework for modeling
Guideline does raise questions about the proposed indicating GRSM model isolated single sources and multi-source
acceptability of a requirement or performance may be better than OLM areas as part of a cumulative impact
recommendation that it replaces, model and PVMRM under certain source analysis. The EPA also proposes to
users and applicants are encouraged to characterization situations. The EPA remove the term ‘‘significant
consult with the appropriate reviewing also proposes to add two references to concentration gradient’’ and its related
authority as soon as possible to assure the section including one for the peer-
content in section 8.3.1(a)(i) due to the

the acceptability of modeling used to reviewed paper on development and ambiguity and lack of definition of this
support permit applications during this evaluation of GRSM, and a second term in the context of modeling multi-
period. reference to the EPA Technical Support source areas.
Document (TSD) on GRSM. The EPA proposes to remove
Concentrations for Use in Modeling
Demonstrations. Publication No. EPA–454/P–23–
The EPA proposes to revise Table 4– paragraph (d) in section 8.3.2 and
001. Office of Air Quality Planning and Standards, 1 in section 4.2.3.4(f) to include GRSM renumber paragraphs (e) and (f) to (d)
Research Triangle Park, NC. as a Tier 3 detailed screening option. and (e), respectively. The content of
paragraph (d) is proposed to be included 8.3.1(a)(ii) and the proposed revisions to our proposed updates to the AERMOD
in the proposed revision of paragraph Tables 8–1 and 8–2. Modeling System.
(a) in section 8.3.2.
i. Section 10.0—References c. Section A.3
In section 8.3.3, the EPA proposes
revisions to the content in section The EPA proposes updates to In section A.3, the EPA proposes to
8.3.3(b) on the recommendations for references in section 10.0 to remove remove the reference to the Bureau of
determining nearby sources to explicitly outdated website links and reflect Ocean Energy Management’s (BOEM)
model as part of a cumulative impact current versions of guidance documents, outdated guidance.
analysis. The EPA proposes to remove user’s guides, and other supporting
documentation where applicable. The V. Ongoing Model Development
the content related to the term
‘‘significant concentration gradient’’ in EPA also proposes to add references to In addition to the proposed beta
section 8.3.3(b)(i), section 8.3.3(b)(ii), support proposed updates to the options above, AERMOD version 23132
and section 8.3.3(b)(iii) due to the lack AERMOD Modeling System described also includes alpha options that are
of definition of this term in the context in this proposed update to the thought to have scientific merit and that
of modeling multi-source areas. The Guideline. are still being developed or evaluated
EPA also proposes to revise the example and peer reviewed. These alpha options
5. Proposed Revisions to Addendum
given in section 8.3.3(d) to be consistent are not being proposed as updates to the
A 21 to Appendix W to Part 51
with the discussion of other sources in regulatory formulation of the AERMOD
a. Section A.0 Modeling System, and the EPA is not
section 8.3.1(a)(ii) and the proposed
revisions to Tables 8–1 and 8–2. The EPA proposes to revise section taking comment on the alpha options
In section 8.4.1, the EPA proposes to A.0 to remove references that indicate during this rulemaking. A list of alpha
include buoy data as an example of site- there are ‘‘many’’ preferred models options on which the EPA has placed a
specific data as a result of the inclusion while the number is currently only high priority for continued research and
of the Coupled-Ocean Atmosphere three. development for model improvement is
Response Experiment (COARE) included below. Refer to the AERMOD
b. Section A.1 User’s Guide for details and usage of
algorithms to AERMET for marine
boundary layer processing. The EPA The EPA proposes to revise the each option.
References section to include additional The AERMOD Modeling System,
proposes to revise paragraph (a) of
references that support our proposed version 23131, includes but is not
section 8.4.2 to note that MMIF should
updates to the AERMOD Modeling limited to the following alpha options:
be used to process prognostic
System. • Low Wind Default Overrides
meteorological data for both land-based
In the Abstract section, the EPA (LOW_WIND).
and overwater applications, and to LOW_WIND was first implemented as
proposes to add line type sources as one
revise paragraph (b) to clarify that a collection of non-regulatory beta test
of the source types AERMOD can
AERSURFACE should be used to options in AERMOD version 12345
simulate.
calculate surface characteristics for The EPA proposes to revise section (LOWWIND1 and LOWWIND2) and
land-based data and AERMET calculates A.1(a) to include overwater applications expanded in version
surface characteristics for overwater for regulatory modeling where shoreline 15481(LOWWIND3). Each of these
applications. Also, the EPA proposes to fumigation and/or platform downwash options altered the default model values
revise paragraph (e) of this section to are not important to facilitate the use of for minimum sigma-v, minimum wind
clarify that at least 1-year of site-specific AERMOD with COARE processing. This speed, and the minimum meander factor
data applies to both land-based and revision would remove the need to with different combinations of
overwater-based data. request an alternative model hardcoded values. Though the original
The EPA proposes to revise paragraph demonstration for such applications. LOW_WIND beta test options are no
(a) of section 8.4.3.2 to remove The EPA also proposes to clarify longer implemented in AERMOD, the
references to specific weblinks and to elevation data that can be used in LOW_WIND option was recategorized as
state that users should refer to the latest AERMOD, specifically the change in the an alpha option in AERMOD version
guidance documents for weblinks. name of the U.S. Geological Survey 18181. The LOW_WIND option in
The EPA proposes to add a new (USGS) National Dataset (NED) to 3D version 23132 enables the user to
section 8.4.6 to discuss the Elevation Program (3DEP). For override AERMOD default values with
implementation of COARE for marine consistency, references to NED would user-defined values for one or more of
boundary layer processing and to be updated to 3DEP throughout section the following parameters:
renumber the existing section 8.4.6 (in A.1. Æ Minimum standard deviation of the
the 2017 Guideline) to a new section The EPA proposes to revise section lateral velocity to the average wind
8.4.7. References to specific wind speed A.1(b) to include prognostic data as direction;
thresholds are proposed to be replaced meteorological input to the AERMOD Æ Minimum mean wind speed;
with guidance to consult the Modeling System, as applicable. Æ Minimum and maximum meander
appropriate guidance documents for the The EPA proposes to revise section factor;
latest thresholds. A.1(l) to include the proposed Generic Æ Minimum standard deviation of the
h. Section 9.0—Regulatory Application Reaction Set Method in the discussion vertical wind speed; and
of Models on chemical transformation in Æ Time scale for random dispersion.
• Modifications to PRIME Building
AERMOD. We also propose to clarify

The EPA proposes to update reference the status of the different deposition Downwash (AWMADWNW and ORD_
numbers where necessary due to added options in A.1(l). DWNW).
references. The EPA proposes to revise section Beginning with AERMOD version
In section 9.2.3, the EPA proposes to A.1(n) to include references to 19191, two distinct sets of alpha options
revise the example given in section additional evaluation studies to support were added that modify the building
9.2.3(a)(ii) to be consistent with the downwash algorithm, PRIME. The two
discussion of other sources in section 21 Formerly designated as appendix A. sets of options were developed
independently by EPA’s ORD (ORD_ Beginning with AERMOD version enforceable duty on any State, local or
DWNW) and the Air & Waste 23132, the characterization of AREA Tribal governments or the private sector.
Management Association (A&WMA) and VOLUME sources was extended to
(AWMADWNW). With a couple of account for the buoyancy and horizontal E. Executive Order 13132: Federalism
exceptions, the options within each set momentum of aircraft emissions. The This action does not have federalism
can be employed individually or aircraft plume rise formulation and code implications. It will not have substantial
combined with other options from each for AREA and VOLUME sources was direct effects on the States, on the
set. independently developed and provided relationship between the national
• Downwash from Offshore Drilling by the Federal Aviation Administration
Platforms (PLATFORM). government and the States, or on the
(FAA). EPA continues to collaborate distribution of power and
To enhance AERMOD’s offshore with FAA on model evaluation and peer
modeling capabilities, the platform responsibilities among the various
review of the aircraft plume rise
downwash algorithm, adapted from the levels of government.
formulations.
Offshore Coastal Dispersion (OCD) F. Executive Order 13175: Consultation
dispersion model, was incorporated in VI. Statutory and Executive Order
Reviews and Coordination With Indian Tribal
AERMOD version 22112 and requires Governments
further development, testing, and Additional information about these
evaluation. The PLATFORM option statutes and Executive Orders can be This action does not have Tribal
simulates the building downwash effect found at https://www.epa.gov/laws- implications, as specified in Executive
from platforms commonly used for regulations/laws-and-executive-orders. Order 13175. This action provides
offshore drilling, made up of both proposed revisions to the Guideline
porous and solid structures and which A. Executive Order 12866: Regulatory which is used by the EPA, other
are elevated with airflow beneath them. Planning and Review and Executive Federal, State, territorial, local, and
• Extended RLINE Source Type Order 14094: Modernizing Regulatory Tribal air quality agencies, and industry
Including Barriers and Depressed Review to prepare and review preconstruction
Roadways (RLINEXT). This action is not a significant permit applications, SIP submittals and
The RLINEXT source type was revisions, determinations of conformity,
regulatory action as defined in
implemented in AERMOD version and other air quality assessments
Executive Order 12866, as amended by
18181 and is an extended version of the required under EPA regulation. Separate
Executive Order 14094, and was
RLINE source type that allows for a from this action, the Tribal Air Rule
therefore not subject to a requirement
more refined characterization of a road implements the provisions of section
for Executive Order 12866 review.
segment. It accepts separate inputs for 301(d) of the CAA authorizing eligible
the elevations of each end of the road B. Paperwork Reduction Act (PRA) Tribes to implement their own Tribal air
segment and extended options for program. Thus, Executive Order 13175
This action does not impose an
modeling with roadway barriers does not apply to this action.
information collection burden under the
(RBARRIER) and depressed roadways
PRA. This action does not contain any The EPA provided information
(RDEPRESS).
• TTRM and TTRM2 for Conversion information collection activities, nor regarding this action to the Tribes
of NOX to NO2. does it add any information collection during a monthly National Tribal Air
The Travel Time Reaction Method requirements beyond those imposed by Association (NTAA) call and will
(TTRM) was implemented in AERMOD existing New Source Review continue to provide any new or
version 21112 as a stand-alone NOX-to- requirements. subsequent updates to EPA modeling
NO2 conversion option that accounts for C. Regulatory Flexibility Act (RFA) guidance and other regulatory
plume travel time, applicable only in compliance demonstration related
the near field. TTRM was further I certify that this action will not have topics upon request of the NTAA.
integrated in AERMOD version 22112 as a significant economic impact on a Additionally, the EPA specifically
TTRM2 which can be paired with any substantial number of small entities solicits any comments on this proposed
one of the Ambient Ration Method under the RFA. This action will not action from Tribal officials.
(ARM2), OLM, or PVMRM. When paired impose any requirements on small
entities. This action proposes revisions G. Executive Order 13045: Protection of
with one of these, TTRM is applied in
the near field and the other specified to the Guideline, including Children From Environmental Health
option is applied in the far field where enhancements to the formulation and Risks and Safety Risks
travel time is not as relevant. application of the EPA’s near-field
EPA interprets Executive Order 13045
• Highly Buoyant Plume (HBP). dispersion modeling system, AERMOD,
as applying only to those regulatory
A Highly Buoyant Plume (HBP) and updates to the recommendations for
the development of appropriate actions that concern environmental
option was implemented as an alpha
background concentration for health or safety risks that EPA has
option that can be applied to POINT
cumulative impact analyses. Use of the reason to believe may
source types beginning with AERMOD
models and/or techniques described in disproportionately affect children, per
version 23132 to further explore
this action is not expected to pose any the definition of ‘‘covered regulatory
AERMOD’s treatment of the penetrated
additional burden on small entities. action’’ in section 2–202 of the
plume. A penetrated plume occurs
Executive Order.
when a plume released into the mixed D. Unfunded Mandates Reform Act
layer, and a portion of the plume Therefore, this action is not subject to
(UMRA)
eventually penetrates the top of the Executive Order 13045 because it does
mixed layer during convective hours as This action does not contain any not concern an environmental health
it continues to rise due to either unfunded mandate as described in risk or safety risk. Since this action does
buoyancy or momentum. UMRA, 2 U.S.C. 1531–1538, and does not concern human health, EPA’s Policy
• Aircraft Plume Rise (AREA/ not significantly or uniquely affect small on Children’s Health also does not
VOLUME Source Types). governments. This action imposes no apply.
H. Executive Order 13211: Actions PART 51—REQUIREMENTS FOR 2.1 Suitability of Models
Concerning Regulations That PREPARATION, ADOPTION, AND 2.1.1 Model Accuracy and Uncertainty
2.2 Levels of Sophistication of Air Quality
Significantly Affect Energy Supply, SUBMITTAL OF IMPLEMENTATION Analyses and Models
Distribution, or Use PLANS 2.3 Availability of Models
3.0 Preferred and Alternative Air Quality
This action is not subject to Executive ■ 1. The authority citation for part 51 Models
Order 13211, because it is not a continues to read as follows: 3.1 Preferred Models
significant regulatory action under Authority: 23 U.S.C. 101; 42 U.S.C. 7401– 3.1.1 Discussion
Executive Order 12866. 7671q. 3.1.2 Requirements
3.2 Alternative Models
I. National Technology Transfer and ■ 2. Appendix W to part 51 is revised 3.2.1 Discussion
Advancement Act to read as follows: 3.2.2 Requirements
3.3 EPA’s Model Clearinghouse
Appendix W to Part 51—Guideline on 4.0 Models for Carbon Monoxide, Lead,
This rulemaking does not involve
Air Quality Models Preface Sulfur Dioxide, Nitrogen Dioxide and
technical standards.
a. Industry and control agencies have long Primary Particulate Matter
J. Executive Order 12898: Federal expressed a need for consistency in the 4.1 Discussion
Actions To Address Environmental application of air quality models for 4.2 Requirements
regulatory purposes. In the 1977 Clean Air 4.2.1 Screening Models and Techniques
Justice in Minority Populations and
Act (CAA), Congress mandated such 4.2.1.1 AERSCREEN
Low-Income Populations and Executive 4.2.1.2 CTSCREEN
consistency and encouraged the
Order 14096: Revitalizing Our Nation’s 4.2.1.3 Screening in Complex Terrain
standardization of model applications. The
Commitment to Environmental Justice Guideline on Air Quality Models (hereafter, 4.2.2 Refined Models
for All Guideline) was first published in April 1978 4.2.2.1 AERMOD
to satisfy these requirements by specifying 4.2.2.2 CTDMPLUS
The EPA believes that this action does models and providing guidance for their use. 4.2.2.3 OCD
not have disproportionate and adverse The Guideline provides a common basis for 4.2.3 Pollutant Specific Modeling
estimating the air quality concentrations of Requirements
human health or environmental effects 4.2.3.1 Models for Carbon Monoxide
on communities with environmental criteria pollutants used in assessing control
strategies and developing emissions limits. 4.2.3.2 Models for Lead
justice concerns because it does not 4.2.3.3 Models for Sulfur Dioxide
b. The continuing development of new air
establish an environmental health or quality models in response to regulatory 4.2.3.4 Models for Nitrogen Dioxide
safety standard. This action proposes 4.2.3.5 Models for PM2.5
requirements and the expanded requirements
revisions to the Guideline, including for models to cover even more complex 4.2.3.6 Models for PM10
problems have emphasized the need for 5.0 Models for Ozone and Secondarily
enhancements to the formulations and Formed Particulate Matter
application of EPA’s near-field periodic review and update of guidance on
these techniques. Historically, three primary 5.1 Discussion
dispersion modeling system, AERMOD, 5.2 Recommendations
activities have provided direct input to
that would assist and expand revisions of the Guideline. The first is a series 5.3 Recommended Models and Approaches
assessment options in Environmental of periodic EPA workshops and modeling for Ozone
Justice determinations. While the EPA conferences conducted for the purpose of 5.3.1 Models for NAAQS Attainment
ensuring consistency and providing Demonstrations and Multi-Source Air
does not expect this action to directly Quality Assessments
impact air quality, the proposed clarification in the application of models.
The second activity was the solicitation and 5.3.2 Models for Single-Source Air
revisions are important because the Quality Assessments
review of new models from the technical and
Guideline is used by air permitting user community. In the March 27, 1980, 5.4 Recommended Models and Approaches
authorities and industry to prepare and Federal Register, a procedure was outlined for Secondarily Formed PM2.5
review NSR permits and serves as a for the submittal to the EPA of privately 5.4.1 Models for NAAQS Attainment
developed models. After extensive evaluation Demonstrations and Multi-Source Air
benchmark of consistency across the Quality Assessments
nation. This consistency has value to all and scientific review, these models, as well
as those made available by the EPA, have 5.4.2 Models for Single-Source Air
communities including communities Quality Assessments
been considered for recognition in the
with environmental justice concerns. Guideline. The third activity is the extensive 6.0 Modeling for Air Quality Related Values
on-going research efforts by the EPA and and Other Governmental Programs
List of Subjects in 40 CFR Part 51 6.1 Discussion
others in air quality and meteorological
modeling. 6.2 Air Quality Related Values
Environmental protection, 6.2.1 Visibility
c. Based primarily on these three activities,
Administrative practice and procedure, new sections and topics have been included 6.2.1.1 Models for Estimating Near-Field
Air pollution control, Carbon monoxide, as needed. The EPA does not make changes Visibility Impairment
Criteria pollutants, Intergovernmental to the guidance on a predetermined schedule, 6.2.1.2 Models for Estimating Visibility
Impairment for Long-Range Transport
relations, Lead, Mobile sources, but rather on an as-needed basis. The EPA
believes that revisions of the Guideline 6.2.2 Models for Estimating Deposition
Nitrogen oxides, Ozone, Particulate Impacts
matter, Reporting and recordkeeping should be timely and responsive to user
needs and should involve public 6.3 Modeling Guidance for Other
requirements, Stationary sources, Sulfur participation to the greatest possible extent. Governmental Programs
oxides. All future changes to the guidance will be 7.0 General Modeling Considerations
proposed and finalized in the Federal 7.1 Discussion
Michael S. Regan, 7.2 Recommendations
Register. Information on the current status of

Administrator. 7.2.1 All sources
modeling guidance can always be obtained
7.2.1.1 Dispersion Coefficients
from the EPA’s Regional Offices.
For the reasons stated in the 7.2.1.2 Complex Winds
preamble, the Environmental Protection Table of Contents 7.2.1.3 Gravitational Settling and
Deposition
Agency is amending title 40, chapter I List of Tables 7.2.2 Stationary Sources
of the Code of Federal Regulations as 1.0 Introduction 7.2.2.1 Good Engineering Practice Stack
follows: 2.0 Overview of Model Use Height
7.2.2.2 Plume Rise by the EPA Regional Offices in judging the a common basis for estimating pollutant
7.2.3 Mobile Sources adequacy of modeling analyses performed by concentrations, assessing control strategies,
8.0 Model Input Data the EPA, by State, local, and Tribal and specifying emissions limits. Such
8.1 Modeling Domain permitting authorities, and by industry. It is consistency is not, however, promoted at the
8.1.1 Discussion appropriate for use by other Federal expense of model and database accuracy. The
8.1.2 Requirements government agencies and by State, local, and Guideline provides a consistent basis for
8.2 Source Data Tribal agencies with air quality and land selection of the most accurate models and
8.2.1 Discussion management responsibilities. The Guideline databases for use in air quality assessments.
8.2.2 Requirements serves to identify, for all interested parties, e. Recommendations are made in the
8.3 Background Concentrations those modeling techniques and databases Guideline concerning air quality models and
8.3.1 Discussion that the EPA considers acceptable. The techniques, model evaluation procedures,
8.3.2 Recommendations for Isolated Guideline is not intended to be a and model input databases and related
Single Sources compendium of modeling techniques. Rather, requirements. The guidance provided here
8.3.3 Recommendations for Multi-Source it should serve as a common measure of should be followed in air quality analyses
Areas acceptable technical analysis when relative to SIPs, NSR, and in supporting
8.4 Meteorological Input Data supported by sound scientific judgment. analyses required by the EPA and by State,
8.4.1 Discussion b. Air quality measurements 5 are routinely local, and Tribal permitting authorities.
8.4.2 Recommendations and used to characterize ambient concentrations Specific models are identified for particular
Requirements of criteria pollutants throughout the nation applications. The EPA may approve the use
8.4.3 National Weather Service Data but are rarely sufficient for characterizing the of an alternative model or technique that can
8.4.3.1 Discussion ambient impacts of individual sources or be demonstrated to be more appropriate than
8.4.3.2 Recommendations demonstrating adequacy of emissions limits those recommended in the Guideline. In all
8.4.4 Site-specific data for an existing source due to limitations in cases, the model or technique applied to a
8.4.4.1 Discussion spatial and temporal coverage of ambient given situation should be the one that
8.4.4.2 Recommendations monitoring networks. The impacts of new provides the most accurate representation of
8.4.5 Prognostic meteorological data sources that do not yet exist, and atmospheric transport, dispersion, and
8.4.5.1 Discussion modifications to existing sources that have chemical transformations in the area of
8.4.5.2 Recommendations yet to be implemented, can only be interest. However, to ensure consistency,
8.4.6 Marine Boundary Layer determined through modeling. Thus, models deviations from the Guideline should be
Environments have become a primary analytical tool in carefully documented as part of the public
8.4.6.1 Discussion most air quality assessments. Air quality record and fully supported by the
8.4.6.2 Recommendations appropriate reviewing authority, as discussed
measurements can be used in a
later.
8.4.7 Treatment of Near-Calms and Calms complementary manner to air quality models,
f. From time to time, situations arise
8.4.7.1 Discussion with due regard for the strengths and
requiring clarification of the intent of the
8.4.7.2 Recommendations weaknesses of both analysis techniques, and
guidance on a specific topic. Periodic
9.0 Regulatory Application of Models are particularly useful in assessing the workshops are held with EPA headquarters,
9.1 Discussion accuracy of model estimates. EPA Regional Offices, and State, local, and
9.2 Recommendations c. It would be advantageous to categorize Tribal agency modeling representatives to
9.2.1 Modeling Protocol the various regulatory programs and to apply ensure consistency in modeling guidance and
9.2.2 Design Concentration and Receptor a designated model to each proposed source to promote the use of more accurate air
Sites needing analysis under a given program. quality models, techniques, and databases.
9.2.3 NAAQS and PSD Increments However, the diversity of the nation’s The workshops serve to provide further
Compliance Demonstrations for New or topography and climate, and variations in explanations of Guideline requirements to
Modified Sources source configurations and operating the EPA Regional Offices and workshop
9.2.3.1 Considerations in Developing characteristics dictate against a strict materials are issued with this clarifying
Emissions Limits modeling ‘‘cookbook.’’ There is no one model information. In addition, findings from
9.2.4 Use of Measured Data in Lieu of capable of properly addressing all ongoing research programs, new model
Model Estimates conceivable situations even within a broad development, or results from model
10.0 References category such as point sources. evaluations and applications are
Meteorological phenomena associated with continuously evaluated. Based on this
Addendum A to Appendix W of Part threats to air quality standards are rarely information, changes in the applicable
51—Summaries of Preferred Air amenable to a single mathematical treatment; guidance may be indicated and appropriate
Quality Models thus, case-by-case analysis and judgment are revisions to the Guideline may be considered.
List of Tables frequently required. As modeling efforts g. All changes to the Guideline must follow
become more complex, it is increasingly rulemaking requirements since the Guideline
important that they be directed by highly is codified in Appendix W to 40 Code of
Table No. Title
competent individuals with a broad range of Federal Regulations (CFR) part 51. The EPA
experience and knowledge in air quality will promulgate proposed and final rules in
8–1 ......... Point Source Model Emission In-
meteorology. Further, they should be the Federal Register to amend this appendix.
puts for SIP Revisions of Inert
coordinated closely with specialists in The EPA utilizes the existing procedures
Pollutants.
emissions characteristics, air monitoring and under CAA section 320 that requires the EPA
8–2 ......... Point Source Model Emission In-
data processing. The judgment of to conduct a Conference on Air Quality
puts for NAAQS Compliance in
experienced meteorologists, atmospheric Modeling at least every 3 years (CAA 320, 42
PSD Demonstrations.
scientists, and analysts is essential. U.S.C. 7620). These modeling conferences are
d. The model that most accurately intended to develop standardized air quality
1.0 Introduction estimates concentrations in the area of modeling procedures and form the basis for
a. The Guideline provides air quality interest is always sought. However, it is clear associated revisions to this Guideline in
modeling techniques that should be applied from the needs expressed by the EPA support of the EPA’s continuing effort to
to State Implementation Plan (SIP) submittals Regional Offices, by State, local, and Tribal prescribe with ‘‘reasonable particularity’’ air
and revisions, to New Source Review (NSR), agencies, by many industries and trade quality models and meteorological and
including new or modifying sources under associations, and also by the deliberations of emission databases suitable for modeling
Prevention of Significant Deterioration Congress, that consistency in the selection National Ambient Air Quality Standards
(PSD),1 2 3 conformity analyses,4 and other air and application of models and databases (NAAQS) 6 and PSD increments. Ample
quality assessments required under EPA should also be sought, even in case-by-case opportunity for public comment will be
regulation. Applicable only to criteria air analyses. Consistency ensures that air quality provided for each proposed change and
pollutants, the Guideline is intended for use control agencies and the general public have public hearings scheduled.
h. A wide range of topics on modeling and meteorological complexities that are often iii. Photochemical grid models are three-
databases are discussed in the Guideline. difficult to measure and simulate. Models dimensional Eulerian grid-based models that
Section 2 gives an overview of models and with adequate performance are available for treat chemical and physical processes in each
their suitability for use in regulatory increasingly complex environments. grid cell and use diffusion and transport
applications. Section 3 provides specific However, they are resource intensive and processes to move chemical species between
guidance on the determination of preferred frequently require site-specific observations grid cells.9 Eulerian models assume that
air quality models and on the selection of and formulations. Such complexities and the emissions are spread evenly throughout each
alternative models or techniques. Sections 4 related challenges for the air quality model grid cell. At coarse grid resolutions,
through 6 provide recommendations on simulation should be considered when Eulerian models have difficulty with fine
modeling techniques for assessing criteria selecting the most appropriate air quality scale resolution of individual plumes.
pollutant impacts from single and multiple model for an application. However, these types of models can be
sources with specific modeling requirements c. Appropriate model input data should be appropriately applied for assessment of near-
for selected regulatory applications. Section available before an attempt is made to field and regional scale reactive pollutant
7 discusses general considerations common evaluate or apply an air quality model. impacts from specific sources 7 10 11 12 or all
to many modeling analyses for stationary and Assuming the data are adequate, the greater sources.13 14 15 Photochemical grid models
mobile sources. Section 8 makes the detail with which a model considers the simulate a more realistic environment for
recommendations for data inputs to models spatial and temporal variations in chemical transformation,7 12 but simulations
including source, background air quality, and meteorological conditions and permit- can be more resource intensive than
meteorological data. Section 9 summarizes enforceable emissions, the greater the ability Lagrangian or Gaussian plume models.
how estimates and measurements of air to evaluate the source impact and to e. Competent and experienced
quality are used in assessing source impact distinguish the effects of various control meteorologists, atmospheric scientists, and
and in evaluating control strategies. strategies. analysts are an essential prerequisite to the
i. Appendix W to 40 CFR part 51 contains d. There are three types of models that successful application of air quality models.
an addendum: addendum A. Thus, when have historically been used in the regulatory The need for such specialists is critical when
reference is made to ‘‘addendum A’’ in this demonstrations applicable in the Guideline, sophisticated models are used or the area has
document, it refers to addendum A to each having strengths and weaknesses that complicated meteorological or topographic
appendix W to 40 CFR part 51. Addendum lend themselves to particular regulatory features. It is important to note that a model
A contains summaries of refined air quality applications. applied improperly or with inappropriate
models that are ‘‘preferred’’ for particular i. Gaussian plume models use a ‘‘steady- data can lead to serious misjudgments
applications; both EPA models and models state’’ approximation, which assumes that regarding the source impact or the
developed by others are included. over the model time step, the emissions, effectiveness of a control strategy.
meteorology and other model inputs, are f. The resource demands generated by use
2.0 Overview of Model Use constant throughout the model domain, of air quality models vary widely depending
a. Increasing reliance has been placed on resulting in a resolved plume with the on the specific application. The resources
concentration estimates from air quality emissions distributed throughout the plume required may be important factors in the
models as the primary basis for regulatory according to a Gaussian distribution. This selection and use of a model or technique for
decisions concerning source permits and formulation allows Gaussian models to a specific analysis. These resources depend
emission control requirements. In many estimate near-field impacts of a limited on the nature of the model and its
situations, such as review of a proposed new number of sources at a relatively high complexity, the detail of the databases, the
source, no practical alternative exists. Before resolution, with temporal scales of an hour difficulty of the application, the amount and
attempting to implement the guidance and spatial scales of meters. However, this level of expertise required, and the costs of
contained in this document, the reader formulation allows for only relatively inert manpower and computational facilities.
should be aware of certain general pollutants, with very limited considerations 2.1.1 Model Accuracy and Uncertainty
information concerning air quality models of transformation and removal (e.g., a. The formulation and application of air
and their evaluation and use. Such deposition), and further limits the domain for quality models are accompanied by several
information is provided in this section. which the model may be used. Thus, sources of uncertainty. ‘‘Irreducible’’
Gaussian models may not be appropriate if uncertainty stems from the ‘‘unknown’’
2.1 Suitability of Models model inputs are changing sharply over the conditions, which may not be explicitly
a. The extent to which a specific air quality model time step or within the desired model accounted for in the model (e.g., the
model is suitable for the assessment of source domain, or if more advanced considerations turbulent velocity field). Thus, there are
impacts depends upon several factors. These of chemistry are needed. likely to be deviations from the observed
include: (1) the topographic and ii. Lagrangian puff models, on the other concentrations in individual events due to
meteorological complexities of the area; (2) hand, are non-steady-state, and assume that variations in the unknown conditions.
the detail and accuracy of the input model input conditions are changing over the ‘‘Reducible’’ uncertainties 16 are caused by:
databases, i.e., emissions inventory, model domain and model time step. (1) uncertainties in the ‘‘known’’ input
meteorological data, and air quality data; (3) Lagrangian models can also be used to conditions (e.g., emission characteristics and
the manner in which complexities of determine near- and far-field impacts from a meteorological data); (2) errors in the
atmospheric processes are handled in the limited number of sources. Traditionally, measured concentrations; and (3) inadequate
model; (4) the technical competence of those Lagrangian models have been used for model physics and formulation.
undertaking such simulation modeling; and relatively inert pollutants, with slightly more b. Evaluations of model accuracy should
(5) the resources available to apply the complex considerations of removal than focus on the reducible uncertainty associated
model. Any of these factors can have a Gaussian models. Some Lagrangian models with physics and the formulation of the
significant influence on the overall model treat in-plume gas and particulate chemistry. model. The accuracy of the model is
performance, which must be thoroughly However, these models require time and normally determined by an evaluation
evaluated to determine the suitability of an space varying concentration fields of procedure which involves the comparison of
air quality model to a particular application oxidants and, in the case of fine particulate model concentration estimates with
or range of applications. matter (PM2.5), neutralizing agents, such as measured air quality data.17 The statement of
b. Air quality models are most accurate and ammonia. Reliable background fields are model accuracy is based on statistical tests or
reliable in areas that have gradual transitions critical for applications involving secondary performance measures such as bias, error,
of land use and topography. Meteorological pollutant formation because secondary correlation, etc.18 19
conditions in these areas are spatially impacts generally occur when in-plume c. Since the 1980’s, the EPA has worked
uniform such that observations are broadly precursors mix and react with species in the with the modeling community to encourage
representative and air quality model background atmosphere.7 8 These oxidant and development of standardized model
projections are not further complicated by a neutralizing agents are not routinely evaluation methods and the development of
heterogeneous environment. Areas subject to measured, but can be generated with a three- continually improved methods for the
major topographic influences experience dimensional photochemical grid model. characterization of model
performance.16 18 20 21 22 There is general operational restrictions are implemented should become familiar and regularly visit
consensus on what should be considered in based on the screening modeling. for important model updates and additional
the evaluation of air quality models; namely, c. The second level consists of refined clarifications and revisions to modeling
quality assurance planning, documentation models that provide more detailed treatment guidance documents that are applicable to
and scrutiny should be consistent with the of physical and chemical atmospheric EPA programs and regulations. Codes and
intended use and should include: processes, require more detailed and precise documentation may also be available from
• Scientific peer review; input data, and provide spatially and the National Technical Information Service
• Supportive analyses (diagnostic temporally resolved concentration estimates. (NTIS), https://www.ntis.gov, and, when
evaluations, code verification, sensitivity As a result, they provide a more available, is referenced with the appropriate
analyses); sophisticated and, at least theoretically, a NTIS accession number.
• Diagnostic and performance evaluations more accurate estimate of source impact and
3.0 Preferred and Alternative Air Quality
with data obtained in trial locations; and the effectiveness of control strategies.
Models
• Statistical performance evaluations in d. There are situations where a screening
the circumstances of the intended model or a refined model is not available a. This section specifies the approach to be
applications. such that screening and refined modeling are taken in determining preferred models for
Performance evaluations and diagnostic not viable options to determine source- use in regulatory air quality programs. The
evaluations assess different qualities of how specific air quality impacts. In such status of models developed by the EPA, as
well a model is performing, and both are situations, a screening technique or reduced- well as those submitted to the EPA for review
needed to establish credibility within the form model may be viable options for and possible inclusion in this Guideline, is
client and scientific community. estimating source impacts. discussed in this section. The section also
d. Performance evaluations allow the EPA i. Screening techniques are differentiated provides the criteria and process for
and model users to determine the relative from a screening model in that screening obtaining EPA approval for use of alternative
performance of a model in comparison with techniques are approaches that make models for individual cases in situations
alternative modeling systems. Diagnostic simplified and conservative assumptions where the preferred models are not
evaluations allow determination of a model about the physical and chemical atmospheric applicable or available. Additional sources of
capability to simulate individual processes processes important to determining source relevant modeling information are: the EPA’s
that affect the results, and usually employ impacts, while screening models make Model Clearinghouse 23 (section 3.3); EPA
smaller spatial/temporal scale data sets (e.g., assumptions about conservative inputs to a modeling conferences; periodic Regional,
field studies). Diagnostic evaluations enable specific model. The complexity of screening State, and Local Modelers’ Workshops; and
the EPA and model users to build confidence techniques ranges from simplified the EPA’s SCRAM website (section 2.3).
that model predictions are accurate for the assumptions of chemistry applied to refined b. When approval is required for a specific
right reasons. However, the objective or screening model output to sophisticated modeling technique or analytical procedure
comparison of modeled concentrations with approximations of the chemistry applied in this Guideline, we refer to the
observed field data provides only a partial within a refined model. ‘‘appropriate reviewing authority.’’ Many
means for assessing model performance. Due ii. Reduced-form models are States and some local agencies administer
to the limited supply of evaluation datasets, computationally efficient simulation tools for NSR permitting under programs approved
there are practical limits in assessing model characterizing the pollutant response to into SIPs. In some EPA regions, Federal
performance. For this reason, the conclusions specific types of emission reductions for a authority to administer NSR permitting and
reached in the science peer reviews and the particular geographic area or background related activities has been delegated to State
supportive analyses have particular relevance environmental conditions that reflect or local agencies. In these cases, such
in deciding whether a model will be useful underlying atmospheric science of a refined agencies ‘‘stand in the shoes’’ of the
for its intended purposes. model but reduce the computational respective EPA Region. Therefore, depending
resources of running a complex, numerical on the circumstances, the appropriate
2.2 Levels of Sophistication of Air Quality air quality model such as a photochemical reviewing authority may be an EPA Regional
Analyses and Models grid model. Office, a State, local, or Tribal agency, or
a. It is desirable to begin an air quality In such situations, an attempt should be perhaps the Federal Land Manager (FLM). In
analysis by using simplified and conservative made to acquire or improve the necessary some cases, the Guideline requires review
methods followed, as appropriate, by more databases and to develop appropriate and approval of the use of an alternative
complex and refined methods. The purpose analytical techniques, but the screening model by the EPA Regional Office
of this approach is to streamline the process technique or reduced-form model may be (sometimes stated as ‘‘Regional
and sufficiently address regulatory sufficient in conducting regulatory modeling Administrator’’). For all approvals of
requirements by eliminating the need of more applications when applied in consultation alternative models or techniques, the EPA
detailed modeling when it is not necessary in with the EPA Regional Office. Regional Office will coordinate and shall
a specific regulatory application. For e. Consistent with the general principle seek concurrence with the EPA’s Model
example, in the context of a PSD permit described in paragraph 2.2(a), the EPA may Clearinghouse. If there is any question as to
application, a simplified and conservative establish a demonstration tool or method as the appropriate reviewing authority, you
analysis may be sufficient where it shows the a sufficient means for a user or applicant to should contact the EPA Regional Office
proposed construction clearly will not cause make a demonstration required by regulation, modeling contact (https://www.epa.gov/
or contribute to ambient concentrations in either by itself or as part of a modeling scram/air-modeling-regional-contacts),
excess of either the NAAQS or the PSD demonstration. To be used for such whose jurisdiction generally includes the
increments.2 3 regulatory purposes, such a tool or method physical location of the source in question
b. There are two general levels of must be reflected in a codified regulation or and its expected impacts.
sophistication of air quality models. The first have a well-documented technical basis and c. In all regulatory analyses, early
level consists of screening models that reasoning that is contained or incorporated in discussions among the EPA Regional Office
provide conservative modeled estimates of the record of the regulatory decision in staff, State, local, and Tribal agency staff,
the air quality impact of a specific source or which it is applied. industry representatives, and where
source category based on simplified appropriate, the FLM, are invaluable and are
assumptions of the model inputs (e.g., preset, 2.3 Availability of Models strongly encouraged. Prior to the actual
worst-case meteorological conditions). In the a. For most of the screening and refined analyses, agreement on the databases to be
case of a PSD assessment, if a screening models discussed in the Guideline, codes, used, modeling techniques to be applied, and
model indicates that the increase in associated documentation and other useful the overall technical approach helps avoid
concentration attributable to the source could information are publicly available for misunderstandings concerning the final
cause or contribute to a violation of any download from the EPA’s Support Center for results and may reduce the later need for
NAAQS or PSD increment, then the second Regulatory Atmospheric Modeling (SCRAM) additional analyses. The preparation of a
level of more sophisticated models should be website at https://www.epa.gov/scram. This written modeling protocol that is vetted with
applied unless appropriate controls or is a website with which air quality modelers the appropriate reviewing authority helps to
keep misunderstandings and resource (and/or tracer measurements) or with other the appropriate reviewing authority and
expenditures at a minimum. well-established analytical techniques. approved by the Regional Administrator.
d. The identification of preferred models in vi. The developer must be willing to make d. Where the EPA has not identified a
this Guideline should not be construed as a the model and source code available to users preferred model for a particular pollutant or
determination that the preferred models at reasonable cost or make them available for situation, the EPA may establish a multi-
identified here are to be permanently used to public access through the internet or tiered approach for making a demonstration
the exclusion of all others or that they are the National Technical Information Service. The required under PSD or another CAA program.
only models available for relating emissions model and its code cannot be proprietary. The initial tier or tiers may involve use of
to air quality. The model that most accurately d. The EPA’s process of establishing a demonstration tools, screening models,
estimates concentrations in the area of preferred model includes a determination of screening techniques, or reduced-form
interest is always sought. However, technical merit, in accordance with the above models; while the last tier may involve the
designation of specific preferred models is six items, including the practicality of the use of demonstration tools, refined models or
needed to promote consistency in model model for use in ongoing regulatory techniques, or alternative models approved
selection and application. programs. Each model will also be subjected under section 3.2.
3.1 Preferred Models to a performance evaluation for an 3.2 Alternative Models
appropriate database and to a peer scientific
3.1.1 Discussion review. Models for wide use (not just an 3.2.1 Discussion
a. The EPA has developed some models isolated case) that are found to perform better a. Selection of the best model or techniques
suitable for regulatory application, while will be proposed for inclusion as preferred for each individual air quality analysis is
other models have been submitted by private models in future Guideline revisions. always encouraged, but the selection should
developers for possible inclusion in the e. No further evaluation of a preferred be done in a consistent manner. A simple
Guideline. Refined models that are preferred model is required for a particular application listing of models in this Guideline cannot
and required by the EPA for particular if the EPA requirements for regulatory use alone achieve that consistency nor can it
applications have undergone the necessary specified for the model in the Guideline are necessarily provide the best model for all
peer scientific reviews 24 25 and model followed. Alternative models to those listed possible situations. As discussed in section
performance evaluation exercises 26 27 that in addendum A should generally be 3.1.1, the EPA has determined and applied a
include statistical measures of model compared with measured air quality data specific evaluation protocol that provides a
performance in comparison with measured when they are used for regulatory statistical technique for evaluating model
air quality data as described in section 2.1.1. applications consistent with performance for predicting peak
b. An American Society for Testing and recommendations in section 3.2. concentration values, as might be observed at
Materials (ASTM) reference 28 provides a individual monitoring locations.29 This
3.1.2 Requirements
general philosophy for developing and protocol is available to assist in developing
implementing advanced statistical a. Addendum A identifies refined models a consistent approach when justifying the use
evaluations of atmospheric dispersion that are preferred for use in regulatory of other-than-preferred models recommended
models, and provides an example statistical applications. If a model is required for a in the Guideline (i.e., alternative models).
technique to illustrate the application of this particular application, the user must select a The procedures in this protocol provide a
philosophy. Consistent with this approach, model from addendum A or follow general framework for objective decision-
the EPA has determined and applied a procedures in section 3.2.2 for use of an making on the acceptability of an alternative
specific evaluation protocol that provides a alternative model or technique. Preferred model for a given regulatory application.
statistical technique for evaluating model models may be used without a formal These objective procedures may be used for
performance for predicting peak demonstration of applicability as long as they conducting both the technical evaluation of
concentration values, as might be observed at are used as indicated in each model summary the model and the field test or performance
individual monitoring locations.29 in addendum A. Further recommendations evaluation.
c. When a single model is found to perform for the application of preferred models to b. This subsection discusses the use of
better than others, it is recommended for specific source applications are found in alternate models and defines three situations
application as a preferred model and listed subsequent sections of the Guideline. when alternative models may be used. This
in addendum A. If no one model is found to b. If changes are made to a preferred model subsection also provides a procedure for
clearly perform better through the evaluation without affecting the modeled implementing 40 CFR 51.166(l)(2) in PSD
exercise, then the preferred model listed in concentrations, the preferred status of the permitting. This provision requires written
addendum A may be selected on the basis of model is unchanged. Examples of approval of the Administrator for any
other factors such as past use, public modifications that do not affect modification or substitution of an applicable
familiarity, resource requirements, and concentrations are those made to enable use model. An applicable model for purposes of
availability. Accordingly, the models listed of a different computer platform or those that 40 CFR 51.166(l) is a preferred model in
in addendum A meet these conditions: only affect the format or averaging time of the addendum A to the Guideline. Approval to
i. The model must be written in a common model results. The integration of a graphical use an alternative model under section 3.2 of
programming language, and the executable(s) user interface (GUI) to facilitate setting up the the Guideline qualifies as approval for the
must run on a common computer platform. model inputs and/or analyzing the model modification or substitution of a model under
ii. The model must be documented in a results without otherwise altering the 40 CFR 51.166(l)(2). The Regional
user’s guide or model formulation report preferred model code is another example of Administrators have delegated authority to
which identifies the mathematics of the a modification that does not affect issue such approvals under section 3.2 of the
model, data requirements and program concentrations. However, when any changes Guideline, provided that such approval is
operating characteristics at a level of detail are made, the Regional Administrator must issued after consultation with the EPA’s
comparable to that available for other require a test case example to demonstrate Model Clearinghouse and formally
recommended models in addendum A. that the modeled concentrations are not documented in a concurrence memorandum
iii. The model must be accompanied by a affected. from the EPA’s Model Clearinghouse which
complete test dataset including input c. A preferred model must be operated demonstrates that the requirements within
parameters and output results. The test data with the options listed in addendum A for its section 3.2 for use of an alternative model
must be packaged with the model in intended regulatory application. If the
have been met.

computer-readable form. regulatory options are not applied, the model
iv. The model must be useful to typical is no longer ‘‘preferred.’’ Any other 3.2.2 Requirements
users, e.g., State air agencies, for specific air modification to a preferred model that would a. Determination of acceptability of an
quality control problems. Such users should result in a change in the concentration alternative model is an EPA Regional Office
be able to operate the computer program(s) estimates likewise alters its status so that it responsibility in consultation with the EPA’s
from available documentation. is no longer a preferred model. Use of the Model Clearinghouse as discussed in
v. The model documentation must include modified model must then be justified as an paragraphs 3.0(b) and 3.2.1(b). Where the
a robust comparison with air quality data alternative model on a case-by-case basis to Regional Administrator finds that an
alternative model is more appropriate than a model or technique is not inappropriately approach results in a plume that has an
preferred model, that model may be used biased for regulatory application; a and hourly-averaged distribution of emission
subject to the approval of the EPA Regional v. A protocol on methods and procedures mass according to a Gaussian curve through
Office based on the requirements of this to be followed has been established. the plume. Though Gaussian steady-state
subsection. This finding will normally result f. To formally document that the models conserve the mass of the primary
from a determination that: (1) a preferred air requirements of section 3.2 for use of an pollutant throughout the plume, they can
quality model is not appropriate for the alternative model are satisfied for a particular still take into account a limited consideration
particular application; or (2) a more application or range of applications, a of first-order removal processes (e.g., wet and
appropriate model or technique is available memorandum will be prepared by the EPA’s dry deposition) and limited chemical
and applicable. Model Clearinghouse through a consultative conversion (e.g., OH oxidation).
b. An alternative model shall be evaluated process with the EPA Regional Office. c. Due to the steady-state assumption,
from both a theoretical and a performance Gaussian plume models are generally
perspective before it is selected for use. There 3.3 EPA’s Model Clearinghouse considered applicable to distances less than
are three separate conditions under which a. The Regional Administrator has the 50 km, beyond which, modeled predictions
such a model may be approved for use: authority to select models that are of plume impact are likely conservative. The
1. If a demonstration can be made that the appropriate for use in a given situation. locations of these impacts are expected to be
model produces concentration estimates However, there is a need for assistance and unreliable due to changes in meteorology that
equivalent to the estimates obtained using a guidance in the selection process so that are likely to occur during the travel time.
preferred model; fairness, consistency, and transparency in d. The applicability of Gaussian plume
2. If a statistical performance evaluation modeling decisions are fostered among the models may vary depending on the
has been conducted using measured air EPA Regional Offices and the State, local, topography of the modeling domain, i.e.,
quality data and the results of that evaluation and Tribal agencies. To satisfy that need, the simple or complex. Simple terrain is
indicate the alternative model performs EPA established the Model Clearinghouse 23 considered to be an area where terrain
better for the given application than a to serve a central role of coordination and features are all lower in elevation than the
comparable model in addendum A; or collaboration between EPA headquarters and top of the stack(s) of the source(s) in
3. If there is no preferred model. the EPA Regional Offices. Additionally, the question. Complex terrain is defined as
Any one of these three separate conditions EPA holds periodic workshops with EPA terrain exceeding the height of the stack(s)
may justify use of an alternative model. Some Headquarters, EPA Regional Offices, and being modeled.
known alternative models that are applicable State, local, and Tribal agency modeling e. Gaussian models determine source
for selected situations are listed on the EPA’s representatives. impacts at discrete locations (receptors) for
SCRAM website (section 2.3). However, b. The appropriate EPA Regional Office each meteorological and emission scenario,
inclusion there does not confer any unique should always be consulted for information and generally attempt to estimate
status relative to other alternative models and guidance concerning modeling methods concentrations at specific sites that represent
that are being or will be developed in the and interpretations of modeling guidance, an ensemble average of numerous repetitions
future. and to ensure that the air quality model user of the same ‘‘event.’’ Uncertainties in model
c. Equivalency, condition (1) in paragraph has available the latest most up-to-date estimates are driven by this formulation, and
(b) of this subsection, is established by policy and procedures. As appropriate, the as noted in section 2.1.1, evaluations of
demonstrating that the appropriate regulatory EPA Regional Office may also request model accuracy should focus on the
metric(s) are within +/¥2 percent of the assistance from the EPA’s Model reducible uncertainty associated with
estimates obtained from the preferred model. Clearinghouse on other applications of physics and the formulation of the model.
The option to show equivalency is intended models, analytical techniques, or databases The ‘‘irreducible’’ uncertainty associated
as a simple demonstration of acceptability for or to clarify interpretation of the Guideline or with Gaussian plume models may be
an alternative model that is nearly identical related modeling guidance. responsible for variation in concentrations of
(or contains options that can make it c. The EPA Regional Office will coordinate as much as +/¥50 percent.30 ‘‘Reducible’’
identical) to a preferred model that it can be with the EPA’s Model Clearinghouse after an uncertainties 16 can be on a similar scale. For
treated for practical purposes as the preferred initial evaluation and decision has been example, Pasquill 31 estimates that, apart
model. However, notwithstanding this developed concerning the application of an from data input errors, maximum ground-
demonstration, models that are not alternative model. The acceptability and level concentrations at a given hour for a
equivalent may be used when one of the two formal approval process for an alternative point source in flat terrain could be in error
other conditions described in paragraphs (d) model is described in section 3.2. by 50 percent due to these uncertainties.
and (e) of this subsection are satisfied. Errors of 5 to 10 degrees in the measured
d. For condition (2) in paragraph (b) of this 4.0 Models for Carbon Monoxide, Lead, wind direction can result in concentration
subsection, established statistical Sulfur Dioxide, Nitrogen Dioxide and errors of 20 to 70 percent for a particular time
performance evaluation procedures and Primary Particulate Matter and location, depending on stability and
techniques 28 29 for determining the 4.1 Discussion station location. Such uncertainties do not
acceptability of a model for an individual indicate that an estimated concentration does
case based on superior performance should a. This section identifies modeling not occur, only that the precise time and
be followed, as appropriate. Preparation and approaches generally used in the air quality locations are in doubt. Composite errors in
implementation of an evaluation protocol impact analysis of sources that emit the highest estimated concentrations of 10 to 40
that is acceptable to both control agencies criteria pollutants carbon monoxide (CO), percent are found to be typical.32 33 However,
and regulated industry is an important lead, sulfur dioxide (SO2), nitrogen dioxide estimates of concentrations paired in time
element in such an evaluation. (NO2), and primary particulates (PM2.5 and and space with observed concentrations are
e. Finally, for condition (3) in paragraph (b) PM10). less certain.
of this subsection, an alternative model or b. The guidance in this section is specific f. Model evaluations and inter-comparisons
technique may be approved for use provided to the application of the Gaussian plume should take these aspects of uncertainty into
that: models identified in addendum A. Gaussian account. For a regulatory application of a
i. The model or technique has received a plume models assume that emissions and model, the emphasis of model evaluations is
scientific peer review; meteorology are in a steady-state, which is generally placed on the highest modeled
ii. The model or technique can be typically based on an hourly time step. This impacts. Thus, the Cox-Tikvart model
demonstrated to be applicable to the problem evaluation approach, which compares the
a For PSD and other applications that use the
on a theoretical basis; highest modeled impacts on several
model results in an absolute sense, the model
iii. The databases which are necessary to should not be biased toward underestimates.
timescales, is recommended for comparisons
perform the analysis are available and Alternatively, for ozone and PM2.5 SIP attainment of models and measurements and model
adequate; demonstrations and other applications that use the inter-comparisons. The approach includes
iv. Appropriate performance evaluations of model results in a relative sense, the model should bootstrap techniques to determine the
the model or technique have shown that the not be biased toward overestimates. significance of various modeled predictions
and increases the robustness of such increments analysis beyond 50 km is classified urban or rural in accordance with
comparisons when the number of available necessary, the selection and use of an section 7.2.1.1. The climatology of the area
measurements are limited.34 35 Because of the alternative model shall occur in agreement must be studied to help define the worst-case
uncertainty in paired modeled and observed with the appropriate reviewing authority meteorological conditions. Agreement shall
concentrations, any attempts at calibration of (paragraph 3.0(b)) and approval by the EPA be reached between the model user and the
models based on these comparisons is of Regional Office based on the requirements of appropriate reviewing authority (paragraph
questionable benefit and shall not be done. paragraph 3.2.2(e). 3.0(b)) on the choice of the screening model
4.2.1 Screening Models and Techniques or technique for each analysis, on the input
4.2 Requirements data and model settings, and the appropriate
a. For NAAQS compliance demonstrations a. Where a preliminary or conservative metric for satisfying regulatory requirements.
under PSD, use of the screening and estimate is desired, point source screening
techniques are an acceptable approach to air 4.2.1.1 AERSCREEN
preferred models for the pollutants listed in
this subsection shall be limited to the near- quality analyses. a. Released in 2011, AERSCREEN is the
field at a nominal distance of 50 km or less. b. As discussed in paragraph 2.2(a), EPA’s recommended screening model for
Near-field application is consistent with screening models or techniques are designed simple and complex terrain for single sources
capabilities of Gaussian plume models and, to provide a conservative estimate of including point sources, area sources,
based on the EPA’s assessment, is sufficient concentrations. The screening models used horizontal stacks, capped stacks, and flares.
to address whether a source will cause or in most applications are the screening AERSCREEN runs AERMOD in a screening
contribute to ambient concentrations in versions of the preferred models for refined mode and consists of two main components:
excess of a NAAQS. In most cases, maximum applications. The two screening models, (1) the MAKEMET program which generates
source impacts of inert pollutants will occur AERSCREEN 37 38 and CTSCREEN, are a site-specific matrix of meteorological
within the first 10 to 20 km from the source. screening versions of AERMOD (American conditions for input to the AERMOD model;
Therefore, the EPA does not consider a long- Meteorological Society (AMS)/EPA and (2) the AERSCREEN command-prompt
range transport assessment beyond 50 km Regulatory Model) and CTDMPLUS interface.
necessary for these pollutants if a near-field (Complex Terrain Dispersion Model Plus b. The MAKEMET program generates a
NAAQS compliance demonstration is Algorithms for Unstable Situations), matrix of meteorological conditions, in the
required.36 respectively. AERSCREEN is the form of AERMOD-ready surface and profile
recommended screening model for most files, based on user-specified surface
b. For assessment of PSD increments
applications in all types of terrain and for characteristics, ambient temperatures,
within the near-field distance of 50 km or
applications involving building downwash. minimum wind speed, and anemometer
less, use of the screening and preferred
For those applications in complex terrain height. The meteorological matrix is
models for the pollutants listed in this
where the application involves a well- generated based on looping through a range
subsection shall be limited to the same
defined hill or ridge, CTSCREEN 39 can be of wind speeds, cloud covers, ambient
screening and preferred models approved for
used. temperatures, solar elevation angles, and
NAAQS compliance demonstrations.
c. Although AERSCREEN and CTSCREEN convective velocity scales (w*, for convective
c. To determine if a compliance
are designed to address a single-source conditions only) based on user-specified
demonstration for NAAQS and/or PSD scenario, there are approaches that can be surface characteristics for surface roughness
increments may be necessary beyond 50 km used on a case-by-case basis to address multi- (Zo), Bowen ratio (Bo), and albedo (r). For
(i.e., long-range transport assessment), the source situations using screening unstable cases, the convective mixing height
following screening approach shall be used meteorology or other conservative model (Zic) is calculated based on w*, and the
to determine if a significant ambient impact assumptions. However, the appropriate mechanical mixing height (Zim) is calculated
will occur with particular focus on Class I reviewing authority (paragraph 3.0(b)) shall for unstable and stable conditions based on
areas and/or the applicable receptors that be consulted, and concurrence obtained, on the friction velocity, u*.
may be threatened at such distances. the protocol for modeling multiple sources c. For applications involving simple or
i. Based on application in the near-field of with AERSCREEN or CTSCREEN to ensure complex terrain, AERSCREEN interfaces with
the appropriate screening and/or preferred that the worst case is identified and assessed. AERMAP. AERSCREEN also interfaces with
model, determine the significance of the d. As discussed in section 4.2.3.4, there are BPIPPRM to provide the necessary building
ambient impacts at or about 50 km from the also screening techniques built into parameters for applications involving
new or modifying source. If a near-field AERMOD that use simplified or limited building downwash using the Plume Rise
assessment is not available or this initial chemistry assumptions for determining the Model Enhancements (PRIME) downwash
analysis indicates there may be significant partitioning of NO and NO2 for NO2 algorithm. AERSCREEN generates inputs to
ambient impacts at that distance, then further modeling. These screening techniques are AERMOD via MAKEMET, AERMAP, and
assessment is necessary. part of the EPA’s preferred modeling BPIPPRM and invokes AERMOD in a
ii. For assessment of the significance of approach for NO2 and do not need to be screening mode. The screening mode of
ambient impacts for NAAQS and/or PSD approved as an alternative model. However, AERMOD forces the AERMOD model
increments, there is not a preferred model or as with other screening models and calculations to represent values for the plume
screening approach for distances beyond 50 techniques, their usage shall occur in centerline, regardless of the source-receptor-
km. Thus, the appropriate reviewing agreement with the appropriate reviewing wind direction orientation. The maximum
authority (paragraph 3.0(b)) and the EPA authority (paragraph 3.0(b)). concentration output from AERSCREEN
Regional Office shall be consulted in e. As discussed in section 4.2(c)(ii), there represents a worst-case 1-hour concentration.
determining the appropriate and agreed upon are screening techniques needed for long- Averaging-time scaling factors of 1.0 for 3-
screening technique to conduct the second range transport assessments that will hour, 0.9 for 8-hour, 0.60 for 24-hour, and
level assessment. Typically, a Lagrangian typically involve the use of a Lagrangian 0.10 for annual concentration averages are
model is most appropriate to use for these model. Based on the long-standing practice applied internally by AERSCREEN to the
second level assessments, but applicants and documented capabilities of these models highest 1-hour concentration calculated by
shall reach agreement on the specific model for long-range transport assessments, the use the model for non-area type sources. For area
and modeling parameters on a case-by-case of a Lagrangian model as a screening type source concentrations for averaging
basis in consultation with the appropriate technique for this purpose does not need to times greater than one hour, the
reviewing authority (paragraph 3.0(b)) and be approved as an alternative model.
concentrations are equal to the 1-hour

EPA Regional Office. When Lagrangian However, their usage shall occur in estimates.37 40
models are used in this manner, they shall consultation with the appropriate reviewing
not include plume-depleting processes, such authority (paragraph 3.0(b)) and EPA 4.2.1.2 CTSCREEN
that model estimates are considered Regional Office. a. CTSCREEN 39 41 can be used to obtain
conservative, as is generally appropriate for f. All screening models and techniques conservative, yet realistic, worst-case
screening assessments. shall be configured to appropriately address estimates for receptors located on terrain
d. In those situations where a cumulative the site and problem at hand. Close attention above stack height. CTSCREEN accounts for
impact analysis for NAAQS and/or PSD must be paid to whether the area should be the three-dimensional nature of plume and
terrain interaction and requires detailed 4.2.2 Refined Models shoreline fumigation are adequately
terrain data representative of the modeling a. A brief description of each preferred considered in consultation with the Regional
domain. The terrain data must be digitized in model for refined applications is found in Office and appropriate reviewing authority.
the same manner as for CTDMPLUS and a addendum A. Also listed in that addendum Where the effects of shoreline fumigation and
terrain processor is available.42 CTSCREEN is Are availability, the model input platform downwash need to be assessed, the
designed to execute a fixed matrix of requirements, the standard options that shall Offshore and Coastal Dispersion (OCD)
meteorological values for wind speed (u), be selected when running the program, and model is the applicable model (paragraph
standard deviation of horizontal and vertical output options. 4.2.2.3).
wind speeds (sv, sw), vertical potential 4.2.2.1 AERMOD 4.2.2.2 CTDMPLUS
temperature gradient (dq/dz), friction a. If the modeling application involves an
velocity (u*), Monin-Obukhov length (L), a. For a wide range of regulatory
applications in all types of terrain, and for elevated point source with a well-defined hill
mixing height (zi) as a function of terrain or ridge and a detailed dispersion analysis of
height, and wind directions for both neutral/ aerodynamic building downwash, the
required model is AERMOD.44 45 The the spatial pattern of plume impacts is of
stable conditions and unstable convective interest, CTDMPLUS is available.
AERMOD regulatory modeling system
conditions. The maximum concentration CTDMPLUS provides greater resolution of
consists of the AERMOD dispersion model,
output from CTSCREEN represents a worst- concentrations about the contour of the hill
the AERMET meteorological processor, and
case 1-hour concentration. Time-scaling feature than does AERMOD through a
the AERMAP terrain processor. AERMOD is
factors of 0.7 for 3-hour, 0.15 for 24-hour and a steady-state Gaussian plume model different plume-terrain interaction algorithm.
0.03 for annual concentration averages are applicable to directly emitted air pollutants 4.2.2.3 OCD
applied internally by CTSCREEN to the that employs best state-of-practice
highest 1-hour concentration calculated by a. The OCD (Offshore and Coastal
parameterizations for characterizing the Dispersion) model is a straight-line Gaussian
the model. meteorological influences and dispersion. model that incorporates overwater plume
4.2.1.3 Screening in Complex Terrain Differentiation of simple versus complex transport and dispersion as well as changes
a. For applications utilizing AERSCREEN, terrain is unnecessary with AERMOD. In that occur as the plume crosses the shoreline.
AERSCREEN automatically generates a polar- complex terrain, AERMOD employs the well- OCD can determine the impact of offshore
known dividing-streamline concept in a emissions from point, area, or line sources on
grid receptor network with spacing
simplified simulation of the effects of plume- the air quality of coastal regions. OCD is also
determined by the maximum distance to
terrain interactions. applicable for situations that involve
model. If the application warrants a different
b. The AERMOD Modeling System has platform building downwash.
receptor network than that generated by
been extensively evaluated across a wide
AERSCREEN, it may be necessary to run 4.2.3 Pollutant Specific Modeling
range of scenarios based on numerous field
AERMOD in screening mode with a user- Requirements
studies, including tall stacks in flat and
defined network. For CTSCREEN
complex terrain settings, sources subject to 4.2.3.1 Models for Carbon Monoxide
applications or AERMOD in screening mode
building downwash influences, and low- a. Models for assessing the impact of CO
outside of AERSCREEN, placement of level non-buoyant sources.27 These
receptors requires very careful attention emissions are needed to meet NSR
evaluations included several long-term field requirements to address compliance with the
when modeling in complex terrain. Often the studies associated with operating plants as
highest concentrations are predicted to occur CO NAAQS and to determine localized
well as several intensive tracer studies. Based impacts from transportations projects.
under very stable conditions, when the on these evaluations, AERMOD has shown
plume is near or impinges on the terrain. Examples include evaluating effects of point
consistently good performance, with ‘‘errors’’ sources, congested roadway intersections and
Under such conditions, the plume may be in predicted versus observed peak
quite narrow in the vertical, so that even highways, as well as the cumulative effect of
concentrations, based on the Robust Highest numerous sources of CO in an urban area.
relatively small changes in a receptor’s Concentration (RHC) metric, consistently
location may substantially affect the b. The general modeling recommendations
within the range of 10 to 40 percent (cited and requirements for screening models in
predicted concentration. Receptors within in paragraph 4.1(e)).
about a kilometer of the source may be even section 4.2.1 and refined models in section
c. AERMOD incorporates the PRIME 4.2.2 shall be applied for CO modeling. Given
more sensitive to location. Thus, a dense algorithm to account for enhanced plume
array of receptors may be required in some the relatively low CO background
growth and restricted plume rise for plumes concentrations, screening techniques are
cases. affected by building wake effects.46 The
b. For applications involving AERSCREEN, likely to be adequate in most cases. In
PRIME algorithm accounts for entrainment of applying these recommendations and
AERSCREEN interfaces with AERMAP to plume mass into the cavity recirculation
generate the receptor elevations. For requirements, the existing 1992 EPA
region, including re-entrainment of plume guidance for screening CO impacts from
applications involving CTSCREEN, digitized mass into the wake region beyond the cavity.
contour data must be preprocessed 42 to highways may be consulted.49
d. AERMOD incorporates the Buoyant Line
provide hill shape parameters in suitable and Point Source (BLP) Dispersion model to 4.2.3.2 Models for Lead
input format. The user then supplies receptor account for buoyant plume rise from line a. In January 1999 (40 CFR part 58,
locations either through an interactive sources. The BLP option utilizes the standard appendix D), the EPA gave notice that
program that is part of the model or directly, meteorological inputs provided by the concern about ambient lead impacts was
by using a text editor; using both methods to AERMET meteorological processor. being shifted away from roadways and
select receptor locations will generally be e. The state-of-the-science for modeling toward a focus on stationary point sources.
necessary to assure that the maximum atmospheric deposition is evolving, new Thus, models for assessing the impact of lead
concentrations are estimated by either model. modeling techniques are continually being emissions are needed to meet NSR
In cases where a terrain feature may ‘‘appear assessed, and their results are being requirements to address compliance with the
to the plume’’ as smaller, multiple hills, it compared with observations. Consequently, lead NAAQS and for SIP attainment
may be necessary to model the terrain both while deposition treatment is available in demonstrations. The EPA has also issued
as a single feature and as multiple hills to AERMOD, the approach taken for any guidance on siting ambient monitors in the
determine design concentrations. purpose shall be coordinated with the vicinity of stationary point sources.50 For
c. Other screening techniques may be appropriate reviewing authority (paragraph lead, the SIP should contain an air quality
acceptable for complex terrain cases where 3.0(b)). analysis to determine the maximum rolling 3-
established procedures 43 are used. The user f. The AERMET meteorological processor month average lead concentration resulting
is encouraged to confer with the appropriate incorporates the COARE algorithms to derive from major lead point sources, such as
reviewing authority (paragraph 3.0(b)) if any marine boundary layer parameters for smelters, gasoline additive plants, etc. The
unforeseen problems are encountered, e.g., overwater applications of AERMOD.47 48 EPA has developed a post-processor to
applicability, meteorological data, receptor AERMOD is applicable for some overwater calculate rolling 3-month average
siting, or terrain contour processing issues. applications when platform downwash and concentrations from model output.51 General
guidance for lead SIP development is also ozone and subsequent reactions between based on source-specific data which satisfies
available.52 ozone and NO, and the photolysis of NO2 to all quality assurance procedures that ensure
b. For major lead point sources, such as NO. data accuracy for both NO2 and NOX within
smelters, which contribute fugitive emissions b. Due to the complexity of NO2 modeling, the typical range of measured values.
and for which deposition is important, a multi-tiered screening approach is required However, alternate information may be used
professional judgment should be used, and to obtain hourly and annual average to justify a source’s anticipated NO2/NOX in-
there shall be coordination with the estimates of NO2.56 Since these methods are stack ratios, such as manufacturer test data,
appropriate reviewing authority (paragraph considered screening techniques, their usage State or local agency guidance, peer-reviewed
3.0(b)). For most applications, the general shall occur in agreement with the appropriate literature, and/or the EPA’s NO2/NOX ratio
requirements for screening and refined reviewing authority (paragraph 3.0(b)). database.
models of section 4.2.1 and 4.2.2 are Additionally, since screening techniques are
applicable to lead modeling. e. For Tier 3, a detailed screening
conservative by their nature, there are
technique shall be applied on a case-by-case
4.2.3.3 Models for Sulfur Dioxide limitations to how these options can be used.
basis. Because of the additional input data
Specifically, modeling of negative emissions
a. Models for SO2 are needed to meet NSR requirements and complexities associated
rates should only be done after consultation
requirements to address compliance with the with the Tier 3 options, their usage shall
with the EPA Regional Office to ensure that
SO2 NAAQS and PSD increments, for SIP occur in consultation with the EPA Regional
decreases in concentrations would not be
attainment demonstrations,53 and for Office in addition to the appropriate
overestimated. Each tiered approach (see
characterizing current air quality via reviewing authority. The Ozone Limiting
Figure 4–1) accounts for increasingly
modeling.54 SO2 is one of a group of highly Method (OLM),58 the Plume Volume Molar
complex considerations of NO2 chemistry
reactive gases known as ‘‘oxides of sulfur’’ Ratio Method (PVMRM),59 and the Generic
and is described in paragraphs c through e
with largest emissions sources being fossil
of this subsection. The tiers of NO2 modeling Set Reaction Method (GRSM) 60 61 are three
fuel combustion at power plants and other
include: detailed screening techniques that may be
industrial facilities.
b. Given the relatively inert nature of SO2 i. A first-tier (most conservative) ‘‘full’’ used for most sources. These three
on the short-term time scales of interest (i.e., conversion approach; techniques use an appropriate section 4.2.2
1-hour) and the sources of SO2 (i.e., ii. A second-tier approach that assumes model to estimate NOX concentrations and
stationary point sources), the general ambient equilibrium between NO and NO2; then estimate the conversion of primary NO
modeling requirements for screening models and emissions to NO2 based on the ambient levels
in section 4.2.1 and refined models in section iii. A third-tier consisting of several of ozone and the plume characteristics. OLM
4.2.2 are applicable for SO2 modeling detailed screening techniques that account only accounts for NO2 formation based on the
applications. For urban areas, AERMOD for ambient ozone and the relative amount of ambient levels of ozone while PVMRM and
automatically invokes a half-life of 4 hours 55 NO and NO2 emitted from a source. GRSM also accommodate distance-dependent
to SO2. Therefore, care must be taken when c. For Tier 1, use an appropriate refined conversion ratios based on ambient ozone.
determining whether a source is urban or model (section 4.2.2) to estimate nitrogen GRSM, PVMRM and OLM require explicit
rural (see section 7.2.1.1 for urban/rural oxides (NOX) concentrations and assume a
specification of the NO2/NOX in-stack ratios
determination methodology). total conversion of NO to NO2.
and that ambient ozone concentrations be
d. For Tier 2, multiply the Tier 1 result(s)
4.2.3.4 Models for Nitrogen Dioxide by the Ambient Ratio Method 2 (ARM2), provided on an hourly basis. GRSM requires
a. Models for assessing the impact of which provides estimates of representative hourly ambient NOX concentrations in
sources on ambient NO2 concentrations are equilibrium ratios of NO2/NOX value based addition to hourly ozone.
needed to meet NSR requirements to address ambient levels of NO2 and NOX derived from f. Alternative models or techniques may be
compliance with the NO2 NAAQS and PSD national data from the EPA’s Air Quality considered on a case-by-case basis and their
increments. Impact of an individual source System (AQS).57 The national default for usage shall be approved by the EPA Regional
on ambient NO2 depends, in part, on the ARM2 includes a minimum ambient NO2/ Office (section 3.2). Such models or
chemical environment into which the NOX ratio of 0.5 and a maximum ambient techniques should consider individual
source’s plume is to be emitted. This is due ratio of 0.9. The reviewing agency may quantities of NO and NO2 emissions,
to the fact that NO2 sources co-emit NO along establish alternative minimum ambient NO2/ atmospheric transport and dispersion, and
with NO2 and any emitted NO may react with NOX values based on the source’s in-stack atmospheric transformation of NO to NO2.
ambient ozone to convert to additional NO2 emissions ratios, with alternative minimum Dispersion models that account for more
downwind. Thus, comprehensive modeling ambient ratios reflecting the source’s in-stack explicit photochemistry may also be
of NO2 would need to consider the ratio of NO2/NOX ratios. Preferably, alternative considered as an alternative model to
emitted NO and NO2, the ambient levels of minimum ambient NO2/NOX ratios should be estimate ambient impacts of NOX sources.
Tier 1: Full
Assume total conversion of NO to NO2
Conversion
Tier 2:
Ambient
Multiply Tier 1 results by an appropriate ambient ratio
Ratio
Method
l
Tier 3:
Detailed analysis with OLM, PVMRM, GRSM or other
Detailed
alternative screening model or technique
Screening
Figure 4–1: Multi-Tiered Approach for ambient concentrations need to be secondary pollutants. For example, ground-
Estimating NO2 Concentrations considered, particle deposition may be used level ozone and a portion of PM2.5 are
4.2.3.5 Models for PM2.5 on a case-by-case basis and their usage shall secondary pollutants formed through
be coordinated with the appropriate photochemical reactions. Ozone and
a. PM2.5 is a mixture consisting of several reviewing authority. A SIP development secondarily formed particulate matter are
diverse components.62 Ambient PM2.5 guide 66 is also available to assist in PM10 closely related to each other in that they
generally consists of two components: (1) the analyses and control strategy development. share common sources of emissions and are
primary component, emitted directly from a c. Fugitive dust usually refers to dust put formed in the atmosphere from chemical
source; and (2) the secondary component, into the atmosphere by the wind blowing reactions with similar precursors.
formed in the atmosphere from other over plowed fields, dirt roads, or desert or b. Ozone formation is driven by emissions
pollutants emitted from the source. Models sandy areas with little or no vegetation. of NOX and volatile organic compounds
for PM2.5 are needed to meet NSR Fugitive emissions include the emissions (VOCs). Ozone formation is a complicated
requirements to address compliance with the resulting from the industrial process that are nonlinear process that requires favorable
PM2.5 NAAQS and PSD increments and for not captured and vented through a stack, but meteorological conditions in addition to VOC
SIP attainment demonstrations. may be released from various locations and NOX emissions. Sometimes complex
b. For NSR modeling assessments, the within the complex. In some unique cases, a terrain features also contribute to the build-
general modeling requirements for screening model developed specifically for the up of precursors and subsequent ozone
models in section 4.2.1 and refined models situation may be needed. Due to the difficult formation or destruction.
in section 4.2.2 are applicable for the primary nature of characterizing and modeling c. PM2.5 can be either primary (i.e., emitted
component of PM2.5, while the methods in fugitive dust and fugitive emissions, the directly from sources) or secondary in nature.
section 5.4 are applicable for addressing the proposed procedure shall be determined in The fraction of PM2.5 which is primary versus
secondary component of PM2.5. Guidance for consultation with the appropriate reviewing secondary varies by location and season. In
PSD assessments is available for determining authority (paragraph 3.0(b)) for each specific the United States, PM2.5 is dominated by a
the best approach to handling sources of situation before the modeling exercise is variety of chemical species or components of
primary and secondary PM2.5.63 begun. Re-entrained dust is created by atmospheric particles, such as ammonium
c. For SIP attainment demonstrations and vehicles driving over dirt roads (e.g., haul sulfate, ammonium nitrate, organic carbon
regional haze reasonable progress goal roads) and dust-covered roads typically mass, elemental carbon, and other soil
analyses, effects of a control strategy on PM2.5 found in arid areas. Such sources can be compounds and oxidized metals. PM2.5
are estimated from the sum of the effects on characterized as line, area or volume sulfate, nitrate, and ammonium ions are
the primary and secondary components sources.65 67 Emission rates may be based on predominantly the result of chemical
composing PM2.5. Model users should refer to site-specific data or values from the general reactions of the oxidized products of SO2 and
section 5.4.1 and associated SIP modeling literature. NOX emissions with direct ammonia
guidance 64 for further details concerning d. Under certain conditions, recommended emissions.68
appropriate modeling approaches. dispersion models may not be suitable to d. Control measures reducing ozone and
d. The general modeling requirements for appropriately address the nature of ambient PM2.5 precursor emissions may not lead to
the refined models discussed in section 4.2.2 PM10. In these circumstances, the alternative proportional reductions in ozone and PM2.5.
shall be applied for PM2.5 hot-spot modeling modeling approach shall be approved by the Modeled strategies designed to reduce ozone
for mobile sources. Specific guidance is EPA Regional Office (section 3.2). or PM2.5 levels typically need to consider the
available for analyzing direct PM2.5 impacts e. The general modeling requirements for chemical coupling between these pollutants.
from highways, terminals, and other the refined models discussed in section 4.2.2 This coupling is important in understanding
transportation projects.65 shall be applied for PM10 hot-spot modeling processes that control the levels of both
pollutants. Thus, when feasible, it is
4.2.3.6 Models for PM10 for mobile sources. Specific guidance is
important to use models that take into
available for analyzing direct PM10 impacts
a. Models for PM10 are needed to meet NSR account the chemical coupling between
from highways, terminals, and other

requirements to address compliance with the ozone and PM2.5. In addition, using such a
transportation projects.65
PM10 NAAQS and PSD increments and for multi-pollutant modeling system can reduce
SIP attainment demonstrations. 5.0 Models for Ozone and Secondarily the resource burden associated with applying
b. For most sources, the general modeling Formed Particulate Matter and evaluating separate models for each
requirements for screening models in section pollutant and promotes consistency among
4.2.1 and refined models in section 4.2.2 5.1 Discussion the strategies themselves.
shall be applied for PM10 modeling. In cases a. Air pollutants formed through chemical e. PM2.5 is a mixture consisting of several
EP23OC23.000</GPH>
where the particle size and its effect on reactions in the atmosphere are referred to as diverse chemical species or components of
atmospheric particles. Because chemical and contemporary applications.64 69 Because 5.3.2 Models for Single-Source Air Quality
physical properties and origins of each model application purpose and scope vary, Assessments
component differ, it may be appropriate to model users should consult with the a. Depending on the magnitude of
use either a single model capable of appropriate reviewing authority (paragraph emissions, estimating the impact of an
addressing several of the important 3.0(b)) to determine what model performance individual source’s emissions of NOX and
components or to model primary and elements should be emphasized and VOC on ambient ozone is necessary for
secondary components using different presented to provide confidence in the obtaining a permit. The simulation of ozone
models. Effects of a control strategy on PM2.5 regulatory model application. formation and transport requires realistic
is estimated from the sum of the effects on e. There is no preferred modeling system treatment of atmospheric chemistry and
the specific components comprising PM2.5. or technique for estimating ozone or deposition. Models (e.g., Lagrangian and
5.2 Recommendations secondary PM2.5 for specific source impacts photochemical grid models) that integrate
or to assess impacts from multiple sources. chemical and physical processes important
a. Chemical transformations can play an in the formation, decay, and transport of
important role in defining the concentrations For assessing secondary pollutant impacts
from single sources, the degree of complexity ozone and important precursor species
and properties of certain air pollutants. should be applied. Photochemical grid
Models that take into account chemical required to assess potential impacts varies
depending on the nature of the source, its models are primarily designed to characterize
reactions and physical processes of various precursor emissions and impacts from a wide
pollutants (including precursors) are needed emissions, and the background environment.
The EPA recommends a two-tiered approach variety of sources over a large geographic
for determining the current state of air area but can also be used to assess the
quality, as well as predicting and projecting where the first tier consists of using existing
impacts from specific sources.7 11 12
the future evolution of these pollutants. It is technically credible and appropriate
b. The first tier of assessment for ozone
important that a modeling system provide a relationships between emissions and impacts impacts involves those situations where
realistic representation of chemical and developed from previous modeling that is existing technical information is available
physical processes leading to secondary deemed sufficient for evaluating a source’s (e.g., results from existing photochemical
pollutant formation and removal from the impacts. The second tier consists of more grid modeling, published empirical estimates
atmosphere. sophisticated case-specific modeling of source specific impacts, or reduced-form
b. Chemical transport models treat analyses. The appropriate tier for a given models) in combination with other
atmospheric chemical and physical processes application should be selected in supportive information and analysis for the
such as deposition and motion. There are two consultation with the appropriate reviewing purposes of estimating secondary impacts
types of chemical transport models, Eulerian authority (paragraph 3.0(b)) and be consistent from a particular source. The existing
(grid based) and Lagrangian. These types of with EPA guidance.70 technical information should provide a
models are differentiated from each other by credible and representative estimate of the
their frame of reference. Eulerian models are 5.3 Recommended Models and Approaches secondary impacts from the project source.
based on a fixed frame of reference and for Ozone The appropriate reviewing authority
Lagrangian models use a frame of reference a. Models that estimate ozone (paragraph 3.0(b)) and appropriate EPA
that moves with parcels of air between the concentrations are needed to guide the guidance 70 71 should be consulted to
source and receptor point.9 Photochemical choice of strategies for the purposes of a determine what types of assessments may be
grid models are three-dimensional Eulerian nonattainment area demonstrating future appropriate on a case-by-case basis.
grid-based models that treat chemical and year attainment of the ozone NAAQS. c. The second tier of assessment for ozone
physical processes in each grid cell and use Additionally, models that estimate ozone impacts involves those situations where
diffusion and transport processes to move concentrations are needed to assess impacts existing technical information is not
chemical species between grid cells.9 These from specific sources or source complexes to available or a first tier demonstration
types of models are appropriate for satisfy requirements for NSR and other indicates a more refined assessment is
assessment of near-field and regional scale needed. For these situations, chemical
regulatory programs. Other purposes for
reactive pollutant impacts from specific transport models should be used to address
ozone modeling include estimating the
sources 7 10 11 12 or all sources.13 14 15 In some single-source impacts. Special considerations
impacts of specific events on air quality,
limited cases, the secondary processes can be are needed when using these models to
ozone deposition impacts, and planning for
treated with a box model, ideally in evaluate the ozone impact from an individual
combination with a number of other areas that may be attaining the ozone
source. Guidance on the use of models and
modeling techniques and/or analyses to treat NAAQS.
other analyses for demonstrating the impacts
individual source sectors. 5.3.1 Models for NAAQS Attainment of single sources for ozone is available.70
c. Regardless of the modeling system used Demonstrations and Multi-Source Air This guidance document provides a more
to estimate secondary impacts of ozone and/ Quality Assessments detailed discussion of the appropriate
or PM2.5, model results should be compared a. Simulation of ozone formation and approaches to obtaining estimates of ozone
to observation data to generate confidence transport is a complex exercise. Control impacts from a single source. Model users
that the modeling system is representative of agencies with jurisdiction over areas with should use the latest version of the guidance
the local and regional air quality. For ozone in consultation with the appropriate
ozone problems should use photochemical
related projects, model estimates of ozone reviewing authority (paragraph 3.0(b)) to
grid models to evaluate the relationship
should be compared with observations in determine the most suitable refined approach
between precursor species and ozone. Use of
both time and space. For PM2.5, model for single-source ozone modeling on a case-
photochemical grid models is the
estimates of speciated PM2.5 components by-case basis.
(such as sulfate ion, nitrate ion, etc.) should recommended means for identifying control
be compared with observations in both time strategies needed to address high ozone 5.4 Recommended Models and Approaches
and space.69 concentrations in such areas. Judgment on for Secondarily Formed PM2.5
d. Model performance metrics comparing the suitability of a model for a given a. Models that estimate PM2.5
observations and predictions are often used application should consider factors that concentrations are needed to guide the
to summarize model performance. These include use of the model in an attainment choice of strategies for the purposes of a
metrics include mean bias, mean error, test, development of emissions and nonattainment area demonstrating future
fractional bias, fractional error, and meteorological inputs to the model, and year attainment of the PM2.5 NAAQS.
correlation coefficient.69 There are no choice of episodes to model. Guidance on the Additionally, models that estimate PM2.5
specific levels of any model performance use of models and other analyses for concentrations are needed to assess impacts
metric that indicate ‘‘acceptable’’ model demonstrating attainment of the air quality from specific sources or source complexes to
performance. The EPA’s preferred approach goals for ozone is available.63 64 Users should satisfy requirements for NSR and other
for providing context about model consult with the appropriate reviewing regulatory programs. Other purposes for
performance is to compare model authority (paragraph 3.0(b)) to ensure the PM2.5 modeling include estimating the
performance metrics with similar most current modeling guidance is applied. impacts of specific events on air quality,
visibility, deposition impacts, and planning particular source. The existing technical analysis techniques. The CAA identifies the
for areas that may be attaining the PM2.5 information should provide a credible and FLM as the Secretary of the department, or
NAAQS. representative estimate of the secondary their designee, with authority over these
5.4.1 Models for NAAQS Attainment impacts from the project source. The lands. Mandatory Federal Class I areas are
Demonstrations and Multi-Source Air appropriate reviewing authority (paragraph defined in the CAA as international parks,
Quality Assessments 3.0(b)) and appropriate EPA guidance 70 71 national parks over 6,000 acres, and
should be consulted to determine what types wilderness areas and memorial parks over
a. Models for PM2.5 are needed to assess the of assessments may be appropriate on a case- 5,000 acres, established as of 1977. The FLMs
adequacy of a proposed strategy for meeting by-case basis. are also concerned with the protection of
the annual and 24-hour PM2.5 NAAQS. c. The second tier of assessment for resources in federally managed Class II areas
Modeling primary and secondary PM2.5 can secondary PM2.5 impacts involves those because of other statutory mandates to
be a multi-faceted and complex problem, situations where existing technical protect these areas. Where State or Tribal
especially for secondary components of PM2.5 information is not available or a first tier agencies have successfully petitioned the
such as sulfates and nitrates. Control demonstration indicates a more refined EPA and lands have been redesignated to
agencies with jurisdiction over areas with assessment is needed. For these situations, Class I status, these agencies may have
secondary PM2.5 problems should use models chemical transport models should be used for equivalent responsibilities to that of the
that integrate chemical and physical assessments of single-source impacts. Special FLMs for these non-Federal Class I areas as
processes important in the formation, decay, considerations are needed when using these described throughout the remainder of
and transport of these species (e.g., models to evaluate the secondary particulate section 6.2.
photochemical grid models). Suitability of a matter impact from an individual source. b. The FLM agency responsibilities include
modeling approach or mix of modeling Guidance on the use of models and other the review of air quality permit applications
approaches for a given application requires analyses for demonstrating the impacts of from proposed new or modified major
technical judgment as well as professional single sources for secondary PM2.5 is pollution sources that may affect these Class
experience in choice of models, use of the available.70 This guidance document I areas to determine if emissions from a
model(s) in an attainment test, development provides a more detailed discussion of the proposed or modified source will cause or
of emissions and meteorological inputs to the appropriate approaches to obtaining contribute to adverse impacts on air quality
model, and selection of days to model. estimates of secondary particulate matter related values (AQRVs) of a Class I area and
Guidance on the use of models and other concentrations from a single source. Model making recommendations to the FLM.
analyses for demonstrating attainment of the users should use the latest version of this AQRVs are resources, identified by the FLM
air quality goals for PM2.5 is available.63 64 guidance in consultation with the agencies, that have the potential to be
Users should consult with the appropriate appropriate reviewing authority (paragraph affected by air pollution. These resources
reviewing authority (paragraph 3.0(b)) to 3.0(b)) to determine the most suitable single- may include visibility, scenic, cultural,
ensure the most current modeling guidance source modeling approach for secondary physical, or ecological resources for a
is applied. PM2.5 on a case-by-case basis. particular area. The FLM agencies take into
5.4.2 Models for Single-Source Air Quality account the particular resources and AQRVs
6.0 Modeling for Air Quality Related that would be affected; the frequency and
Assessments
Values and Other Governmental Programs magnitude of any potential impacts; and the
a. Depending on the magnitude of
6.1 Discussion direct, indirect, and cumulative effects of any
emissions, estimating the impact of an
potential impacts in making their
individual source’s emissions on secondary a. Other Federal government agencies and
recommendations.
particulate matter concentrations may be State, local, and Tribal agencies with air
c. While the AQRV notification and impact
necessary for obtaining a permit. Primary quality and land management responsibilities
analysis requirements are outlined in the
PM2.5 components shall be simulated using have also developed specific modeling
PSD regulations at 40 CFR 51.166(p) and 40
the general modeling requirements in section approaches for their own regulatory or other
CFR 52.21(p), determination of appropriate
4.2.3.5. The simulation of secondary requirements. Although such regulatory
analytical methods and metrics for AQRV’s
particulate matter formation and transport is requirements and guidance have come about
are determined by the FLM agencies and are
a complex exercise requiring realistic because of EPA rules or standards, the
published in guidance external to the general
treatment of atmospheric chemistry and implementation of such regulations and the
recommendations of this paragraph.
deposition. Models should be applied that use of the modeling techniques is under the
d. To develop greater consistency in the
integrate chemical and physical processes jurisdiction of the agency issuing the
application of air quality models to assess
important in the formation, decay, and guidance or directive. This section covers
potential AQRV impacts in both Class I areas
transport of these species (e.g., Lagrangian such situations with reference to those
and protected Class II areas, the FLM
and photochemical grid models). guidance documents, when they are
agencies have developed the Federal Land
Photochemical grid models are primarily available.
Managers’ Air Quality Related Values Work
designed to characterize precursor emissions b. When using the model recommended or Group Phase I Report (FLAG).72 FLAG
and impacts from a wide variety of sources discussed in the Guideline in support of
over a large geographic area and can also be focuses upon specific technical and policy
programmatic requirements not specifically issues associated with visibility impairment,
used to assess the impacts from specific covered by EPA regulations, the model user effects of pollutant deposition on soils and
sources.7 10 For situations where a project should consult the appropriate Federal, State, surface waters, and ozone effects on
source emits both primary PM2.5 and PM2.5 local, or Tribal agency to ensure the proper vegetation. Model users should consult the
precursors, the contribution from both application and use of the models and/or latest version of the FLAG report for current
should be combined for use in determining techniques. These agencies have developed modeling guidance and with affected FLM
the source’s ambient impact. Approaches for specific modeling approaches for their own agency representatives for any application
combining primary and secondary impacts regulatory or other requirements. Most of the specific guidance which is beyond the scope
are provided in appropriate guidance for programs have, or will have when fully of the Guideline.
single source permit related developed, separate guidance documents that
demonstrations.70 cover the program and a discussion of the 6.2.1 Visibility
b. The first tier of assessment for secondary tools that are needed. The following a. Visibility in important natural areas (e.g.,
PM2.5 impacts involves those situations paragraphs reference those guidance Federal Class I areas) is protected under a
where existing technical information is documents, when they are available. number of provisions of the CAA, including
available (e.g., results from existing sections 169A and 169B (addressing impacts
photochemical grid modeling, published 6.2 Air Quality Related Values primarily from existing sources) and section
empirical estimates of source specific a. The 1990 CAA Amendments give FLMs 165 (new source review). Visibility
impacts, or reduced-form models) in an ‘‘affirmative responsibility’’ to protect the impairment is caused by light scattering and
combination with other supportive natural and cultural resources of Class I areas light absorption associated with particles and
information and analysis for the purposes of from the adverse impacts of air pollution and gases in the atmosphere. In most areas of the
estimating secondary impacts from a to provide the appropriate procedures and country, light scattering by PM2.5 is the most
significant component of visibility physical processes leading to secondary be applied for an assessment of deposition
impairment. The key components of PM2.5 pollutant formation and removal from the impacts due to one or a small group of
contributing to visibility impairment include atmosphere. sources. Over these distances, chemical and
sulfates, nitrates, organic carbon, elemental b. Chemical transport models treat physical transformations can change
carbon, and crustal material.72 atmospheric chemical and physical processes atmospheric residence time due to different
b. Visibility regulations (40 CFR 51.300 such as deposition and motion. There are two propensity for deposition to the surface of
through 51.309) require State, local, and types of chemical transport models, Eulerian different forms of nitrate and sulfate. Users
Tribal agencies to mitigate current and (grid based) and Lagrangian. These types of should consult the latest version of the FLAG
prevent future visibility impairment in any of models are differentiated from each other by report 72 and relevant FLM representatives
the 156 mandatory Federal Class I areas their frame of reference. Eulerian models are for guidance on the use of models for
where visibility is considered an important based on a fixed frame of reference and deposition. Where source and receptors are
attribute. In 1999, the EPA issued revisions Lagrangian models use a frame of reference in close proximity, users should contact the
to the regulations to address visibility that moves with parcels of air between the appropriate FLM for application-specific
impairment in the form of regional haze, source and receptor point.9 Photochemical guidance.
which is caused by numerous, diverse grid models are three-dimensional Eulerian
sources (e.g., stationary, mobile, and area grid-based models that treat chemical and 6.3 Modeling Guidance for Other
sources) located across a broad region (40 physical processes in each grid cell and use Governmental Programs
CFR 51.308 through 51.309). The state of diffusion and transport processes to move a. Dispersion and photochemical grid
relevant scientific knowledge has expanded chemical species between grid cells.9 These modeling may need to be conducted to
significantly since that time. A number of types of models are appropriate for ensure that individual and cumulative
studies and reports 73 74 have concluded that assessment of near-field and regional scale offshore oil and gas exploration,
long-range transport (e.g., up to hundreds of reactive pollutant impacts from specific development, and production plans and
kilometers) of fine particulate matter plays a sources 7 10 11 12 or all sources.13 14 15 activities do not significantly affect the air
significant role in visibility impairment c. Development of the requisite quality of any State as required under the
across the country. Section 169A of the CAA meteorological and emissions databases Outer Continental Shelf Lands Act (OCSLA).
requires States to develop SIPs containing necessary for use of photochemical grid Air quality modeling requires various input
long-term strategies for remedying existing models to estimate AQRVs should conform to datasets, including emissions sources,
and preventing future visibility impairment recommendations in section 8 and those meteorology, and pre-existing pollutant
in the 156 mandatory Class I Federal areas, outlined in the EPA’s Modeling Guidance for concentrations. For sources under the
where visibility is considered an important Demonstrating Attainment of Air Quality reviewing authority of the Department of
attribute. In order to develop long-term Goals for Ozone, PM2.5, and Regional Haze.64 Interior, Bureau of Ocean Energy
strategies to address regional haze, many Demonstration of the adequacy of prognostic Management (BOEM), guidance for the
State, local, and Tribal agencies will need to meteorological fields can be established development of all necessary Outer
conduct regional-scale modeling of fine through appropriate diagnostic and statistical Continental Shelf (OCS) air quality modeling
particulate concentrations and associated performance evaluations consistent with inputs and appropriate model selection and
visibility impairment. recommendations provided in the application is available from the BOEM’s
c. The FLAG visibility modeling appropriate guidance.64 Model users should website: https://www.boem.gov/about-boem/
recommendations are divided into two consult the latest version of this guidance regulations-guidance/guidance-portal.
distinct sections to address different and with the appropriate reviewing authority b. The Federal Aviation Administration
requirements for: (1) near field modeling (paragraph 3.0(b)) for any application- (FAA) is the appropriate reviewing authority
where plumes or layers are compared against specific guidance that is beyond the scope of for air quality assessments of primary
a viewing background, and (2) distant/multi- this subsection. pollutant impacts at airports and air bases.
source modeling for plumes and aggregations 6.2.2 Models for Estimating Deposition The Aviation Environmental Design Tool
of plumes that affect the general appearance Impacts (AEDT) is developed and supported by the
of a scene.72 The recommendations FAA, and is appropriate for air quality
separately address visibility assessments for a. For many Class I areas, AQRVs have assessment of primary pollutant impacts at
sources proposing to locate relatively near been identified that are sensitive to airports or air bases. AEDT has adopted
and at farther distances from these areas.72 atmospheric deposition of air pollutants.
AERMOD for treating dispersion. Application
Emissions of NOX, sulfur oxides, NH3,
6.2.1.1 Models for Estimating Near-Field of AEDT is intended for estimating the
mercury, and secondary pollutants such as
Visibility Impairment change in emissions for aircraft operations,
ozone and particulate matter affect
point source, and mobile source emissions on
a. To calculate the potential impact of a components of ecosystems. In sensitive
airport property and quantify the associated
plume of specified emissions for specific ecosystems, these compounds can acidify
pollutant level- concentrations. AEDT is not
transport and dispersion conditions (‘‘plume soils and surface waters, add nutrients that
intended for PSD, SIP, or other regulatory air
blight’’) for source-receptor distances less change biodiversity, and affect the ecosystem
quality analyses of point or mobile sources at
than 50 km, a screening model and guidance services provided by forests and natural
or peripheral to airport property that are
are available.72 75 If a more comprehensive areas.72 To address the relationship between
deposition and ecosystem effects, the FLM unrelated to airport operations. The latest
analysis is necessary, a refined model should version of AEDT may be obtained from the
be selected. The model selection, procedures, agencies have developed estimates of critical
loads. A critical load is defined as, ‘‘A FAA at: https://aedt.faa.gov.
and analyses should be determined in
consultation with the appropriate reviewing quantitative estimate of an exposure to one 7.0 General Modeling Considerations
authority (paragraph 3.0(b)) and the affected or more pollutants below which significant
harmful effects on specified sensitive 7.1 Discussion
FLM(s).
elements of the environment do not occur a. This section contains recommendations
6.2.1.2 Models for Estimating Visibility
according to present knowledge.’’ 76 concerning a number of different issues not
Impairment for Long-Range Transport
b. The FLM deposition modeling explicitly covered in other sections of the
a. Chemical transformations can play an recommendations are divided into two Guideline. The topics covered here are not
important role in defining the concentrations distinct sections to address different specific to any one program or modeling area,
and properties of certain air pollutants. requirements for: (1) near field modeling, and
but are common to dispersion modeling

Models that take into account chemical (2) distant/multi-source modeling for analyses for criteria pollutants.
reactions and physical processes of various cumulative effects. The recommendations
pollutants (including precursors) are needed separately address deposition assessments for 7.2 Recommendations
for determining the current state of air sources proposing to locate relatively near 7.2.1 All sources
quality, as well as predicting and projecting and at farther distances from these areas.72
the future evolution of these pollutants. It is Where the source and receptors are not in 7.2.1.1 Dispersion Coefficients
important that a modeling system provide a close proximity, chemical transport (e.g., a. For any dispersion modeling exercise,
realistic representation of chemical and photochemical grid) models generally should the urban or rural determination of a source
is critical in determining the boundary layer urban complexes, the entire area should be shoreline of bodies of water. This can affect
characteristics that affect the model’s modeled as an urban region if most of the both individual plumes and area-wide
prediction of downwind concentrations. sources are located in areas classified as emissions. When fumigation conditions are
Historically, steady-state Gaussian plume urban. For tall stacks located within or expected to occur from a source or sources
models used in most applications have adjacent to small or moderate sized urban with tall stacks located on or just inland of
employed dispersion coefficients based on areas, the stack height or effective plume a shoreline, this should be addressed in the
Pasquill-Gifford 77 in rural areas and height may extend above the urban boundary air quality modeling analysis. The EPA has
McElroy-Pooler 78 in urban areas. These layer and, therefore, may be more evaluated several coastal fumigation models,
coefficients are still incorporated in the BLP appropriately modeled using rural and the evaluation results of these models are
and OCD models. However, the AERMOD coefficients. Model users should consult with available for their possible application on a
model incorporates a more up-to-date the appropriate reviewing authority case-by-case basis when air quality estimates
characterization of the atmospheric boundary (paragraph 3.0(b)) and the latest version of under shoreline fumigation conditions are
layer using continuous functions of the AERMOD Implementation Guide 81 when needed.84 Selection of the appropriate model
parameterized horizontal and vertical evaluating this situation. for applications where shoreline fumigation
turbulence based on Monin-Obukhov f. Buoyancy-induced dispersion (BID), as is of concern should be determined in
similarity (scaling) relationships.44 Another identified by Pasquill,82 is included in the consultation with the appropriate reviewing
key feature of AERMOD’s formulation is the preferred models and should be used where authority (paragraph 3.0(b)).
option to use directly observed variables of buoyant sources (e.g., those involving fuel iii. Stagnation. Stagnation conditions are
the boundary layer to parameterize combustion) are involved. characterized by calm or very low wind
dispersion.44 45 7.2.1.2 Complex Winds speeds, and variable wind directions. These
b. The selection of rural or urban stagnant meteorological conditions may
dispersion coefficients in a specific a. Inhomogeneous local winds. In many
persist for several hours to several days.
application should follow one of the parts of the United States, the ground is
During stagnation conditions, the dispersion
procedures suggested by Irwin 79 to neither flat nor is the ground cover (or land
of air pollutants, especially those from low-
determine whether the character of an area is use) uniform. These geographical variations
level emissions sources, tends to be
primarily urban or rural (of the two methods, can generate local winds and circulations,
and modify the prevailing ambient winds minimized, potentially leading to relatively
the land use procedure is considered more high ground-level concentrations. If point
and circulations. Typically, geographic
definitive.): sources are of interest, users should note the
effects are more apparent when the ambient
i. Land Use Procedure: (1) Classify the land guidance provided in paragraph (a) of this
winds are light or calm, as stronger synoptic
use within the total area, Ao, circumscribed subsection. Selection of the appropriate
or mesoscale winds can modify, or even
by a 3 km radius circle about the source model for applications where stagnation is of
eliminate the weak geographic circulations.83
using the meteorological land use typing concern should be determined in
In general, these geographically induced
scheme proposed by Auer; 80 (2) if land use consultation with the appropriate reviewing
wind circulation effects are named after the
types I1, I2, C1, R2, and R3 account for 50 authority (paragraph 3.0(b)).
source location of the winds, e.g., lake and
percent or more of Ao, use urban dispersion
sea breezes, and mountain and valley winds. 7.2.1.3 Gravitational Settling and
coefficients; otherwise, use appropriate rural In very rugged hilly or mountainous terrain, Deposition
dispersion coefficients. along coastlines, or near large land use
ii. Population Density Procedure: (1) a. Gravitational settling and deposition
variations, the characteristics of the winds may be directly included in a model if either
Compute the average population density, p are a balance of various forces, such that the
per square kilometer with Ao as defined is a significant factor. When particulate
assumptions of steady-state straight-line matter sources can be quantified and settling
above; (2) If p is greater than 750 people per transport both in time and space are
square kilometer, use urban dispersion and dry deposition are problems, use
inappropriate. In such cases, a model should professional judgment along with
coefficients; otherwise use appropriate rural be chosen to fully treat the time and space
dispersion coefficients. coordination with the appropriate reviewing
variations of meteorology effects on transport authority (paragraph 3.0(b)). AERMOD
c. Population density should be used with and dispersion. The setup and application of
caution and generally not be applied to contains algorithms for dry and wet
such a model should be determined in
highly industrialized areas where the deposition of gases and particles.85 For other
consultation with the appropriate reviewing
population density may be low and, thus, a Gaussian plume models, an ‘‘infinite half-
authority (paragraph 3.0(b)) consistent with
rural classification would be indicated. life’’ may be used for estimates of particle
limitations of paragraph 3.2.2(e). The
However, the area is likely to be sufficiently concentrations when only exponential decay
meteorological input data requirements for
built-up so that the urban land use criteria terms are used for treating settling and
developing the time and space varying three-
would be satisfied. Therefore, in this case, deposition. Lagrangian models have varying
dimensional winds and dispersion
the classification should be ‘‘urban’’ and degrees of complexity for dealing with
meteorology for these situations are
urban dispersion parameters should be used. settling and deposition and the selection of
discussed in paragraph 8.4.1.2(c). Examples
d. For applications of AERMOD in urban a parameterization for such should be
of inhomogeneous winds include, but are not
areas, under either the Land Use Procedure limited to, situations described in the included in the approval process for selecting
or the Population Density Procedure, the user following paragraphs: a Lagrangian model. Eulerian grid models
needs to estimate the population of the urban i. Inversion breakup fumigation. Inversion tend to have explicit parameterizations for
area affecting the modeling domain because breakup fumigation occurs when a plume (or gravitational settling and deposition as well
the urban influence in AERMOD is scaled multiple plumes) is emitted into a stable as wet deposition parameters already
based on a user-specified population. For layer of air and that layer is subsequently included as part of the chemistry scheme.
non-population oriented urban areas, or areas mixed to the ground through convective 7.2.2 Stationary Sources
influenced by both population and industrial transfer of heat from the surface or because
activity, the user will need to estimate an 7.2.2.1 Good Engineering Practice Stack
of advection to less stable surroundings.
equivalent population to adequately account Fumigation may cause excessively high Height
for the combined effects of industrialized concentrations, but is usually rather short- a. The use of stack height credit in excess
areas and populated areas within the lived at a given receptor. There are no of Good Engineering Practice (GEP) stack
modeling domain. Selection of the recommended refined techniques to model height or credit resulting from any other
appropriate population for these applications this phenomenon. There are, however, dispersion technique is prohibited in the
should be determined in consultation with screening procedures 40 that may be used to development of emissions limits by 40 CFR
the appropriate reviewing authority approximate the concentrations. 51.118 and 40 CFR 51.164. The definition of
(paragraph 3.0(b)) and the latest version of Considerable care should be exercised in GEP stack height and dispersion technique
the AERMOD Implementation Guide.81 using the results obtained from the screening are contained in 40 CFR 51.100. Methods and
e. It should be noted that AERMOD allows techniques. procedures for making the appropriate stack
for modeling rural and urban sources in a ii. Shoreline fumigation. Fumigation can be height calculations, determining stack height
single model run. For analyses of whole an important phenomenon on and near the credits and an example of applying those
techniques are found in several occur in agreement with the appropriate meteorological data used. This section
references,86 87 88 89 that provide a great deal reviewing authority (paragraph 3.0(b)) and attempts to minimize the uncertainty
of additional information for evaluating and approval by the EPA Regional Office based associated with database selection and use by
describing building cavity and wake effects. on the requirements of section 3.2.2. identifying requirements for input data used
b. If stacks for new or existing major 7.2.3 Mobile Sources in modeling. More specific data requirements
sources are found to be less than the height and the format required for the individual
defined by the EPA’s refined formula for a. Emissions of primary pollutants from models are described in detail in the user’s
determining GEP height, then air quality mobile sources can be modeled with an guide and/or associated documentation for
impacts associated with cavity or wake appropriate model identified in section 4.2. each model.
effects due to the nearby building structures Screening of mobile sources can be
should be determined. The EPA refined accomplished by using screening 8.1 Modeling Domain
formula height is defined as H + 1.5L.88 Since meteorology, e.g., worst-case meteorological 8.1.1 Discussion
the definition of GEP stack height defines conditions. Maximum hourly concentrations
computed from screening modeling can be a. The modeling domain is the geographic
excessive concentrations as a maximum area for which the required air quality
ground-level concentration due in whole or converted to longer averaging periods using
the scaling ratios specified in the analyses for the NAAQS and PSD increments
in part to downwash of at least 40 percent are conducted.
in excess of the maximum concentration AERSCREEN User’s Guide.37
without downwash, the potential air quality b. Mobile sources can be modeled in 8.1.2 Requirements
impacts associated with cavity and wake AERMOD as either line (i.e., elongated area) a. For a NAAQS or PSD increments
effects should also be considered for stacks sources or as a series of volume sources. Line assessment, the modeling domain or project’s
that equal or exceed the EPA formula height sources can be represented in AERMOD with impact area shall include all locations where
for GEP. The AERSCREEN model can be used the following source types: LINE, AREA, the emissions of a pollutant from the new or
to obtain screening estimates of potential VOLUME or RLINE. However, since mobile modifying source(s) may cause a significant
downwash influences, based on the PRIME source modeling usually includes an analysis ambient impact. This impact area is defined
downwash algorithm incorporated in the of very near-source impacts, the results can as an area with a radius extending from the
AERMOD model. If more refined be highly sensitive to the characterization of new or modifying source to: (1) the most
concentration estimates are required, the mobile emissions. Important distant location where air quality modeling
AERMOD should be used (section 4.2.2). characteristics for both line/area and volume predicts a significant ambient impact will
sources include the plume release height, occur, or (2) the nominal 50 km distance
7.2.2.2 Plume Rise source width, and initial dispersion considered applicable for Gaussian
a. The plume rise methods of Briggs 90 91 characteristics, and should also take into dispersion models, whichever is less. The
are incorporated in many of the preferred account the impact of traffic-induced required air quality analysis shall be carried
models and are recommended for use in turbulence that can cause roadway sources to out within this geographical area with
many modeling applications. In have larger initial dimensions than might characterization of source impacts, nearby
AERMOD,44 45 for the stable boundary layer, normally be used for representing line source impacts, and background
plume rise is estimated using an iterative sources. concentrations, as recommended later in this
approach, similar to that in the CTDMPLUS c. The EPA’s quantitative PM hot-spot section.
model. In the convective boundary layer, guidance 65 and Haul Road Workgroup Final b. For SIP attainment demonstrations for
plume rise is superposed on the Report 67 provide guidance on the ozone and PM2.5, or regional haze reasonable
displacements by random convective appropriate characterization of mobile progress goal analyses, the modeling domain
velocities.92 In AERMOD, plume rise is sources as a function of the roadway and is determined by the nature of the problem
computed using the methods of Briggs, vehicle characteristics. The EPA’s being modeled and the spatial scale of the
except in cases involving building quantitative PM hot-spot guidance includes emissions that impact the nonattainment or
downwash, in which a numerical solution of important considerations and should be Class I area(s). The modeling domain shall be
the mass, energy, and momentum consulted when modeling roadway links. designed so that all major upwind source
conservation laws is performed.93 No explicit Area and line sources, which can be areas that influence the downwind
provisions in these models are made for characterized as AREA, LINE, and RLINE nonattainment area are included in addition
multistack plume rise enhancement or the source types in AERMOD, or volume sources, to all monitor locations that are currently or
handling of such special plumes as flares. may be used for modeling mobile sources. recently violating the NAAQS or close to
b. Gradual plume rise is generally However, experience in the field has shown violating the NAAQS in the nonattainment
recommended where its use is appropriate: that area sources (characterized as AREA, area. Similarly, all Class I areas to be
(1) in AERMOD; (2) in complex terrain LINE, or RLINE source types) may be easier
evaluated in a regional haze modeling
screening procedures to determine close-in to characterize correctly compared to volume
application shall be included and sufficiently
impacts; and (3) when calculating the effects sources. If volume sources are used, it is
distant from the edge of the modeling
of building wakes. The building wake particularly important to ensure that roadway
domain. Guidance on the determination of
algorithm in AERMOD incorporates and emissions are appropriately spaced when
exercises the thermodynamically based the appropriate modeling domain for
using volume source so that the emissions
gradual plume rise calculations as described photochemical grid models in demonstrating
field is uniform across the roadway.
in paragraph (a) of this subsection. If the attainment of these air quality goals is
Additionally, receptor placement is
building wake is calculated to affect the available.64 Users should consult the latest
particularly important for volume sources
plume for any hour, gradual plume rise is version of this guidance for the most current
that have ‘‘exclusion zones’’ where
also used in downwind dispersion concentrations are not calculated for modeling guidance and the appropriate
calculations to the distance of final plume receptors located ‘‘within’’ the volume reviewing authority (paragraph 3.0(b)) for any
rise, after which final plume rise is used. sources, i.e., less than 2.15 times the initial application specific guidance that is beyond
Plumes captured by the near wake are re- lateral dispersion coefficient from the center the scope of this section.
emitted to the far wake as a ground-level of the volume.65 Therefore, placing receptors 8.2 Source Data
volume source. in these ‘‘exclusion zones’’ will result in
c. Stack tip downwash generally occurs 8.2.1 Discussion
underestimates of roadway impacts.
with poorly constructed stacks and when the a. Sources of pollutants can be classified as
ratio of the stack exit velocity to wind speed 8.0 Model Input Data point, line, area, and volume sources. Point
is small. An algorithm developed by Briggs 91 a. Databases and related procedures for sources are defined in terms of size and may
is the recommended technique for this estimating input parameters are an integral vary between regulatory programs. The line
situation and is used in preferred models for part of the modeling process. The most sources most frequently considered are
point sources. appropriate input data available should roadways and streets along which there are
d. On a case-by-case basis, refinements to always be selected for use in modeling well-defined movements of motor vehicles.
the preferred model may be considered for analyses. Modeled concentrations can vary They may also be lines of roof vents or
plume rise and downwash effects and shall widely depending on the source data or stacks, such as in aluminum refineries. Area
and volume sources are often collections of haze reasonable progress goal analyses, d. For stationary source applications,
a multitude of minor sources with emissions which reflect actual emissions changes in operating conditions that affect
individually small emissions that are during the base modeling year time period the physical emission parameters (e.g.,
impractical to consider as separate point or should be input to models for base year release height, initial plume volume, and exit
line sources. Large area sources are typically modeling. Emissions projections to future velocity) shall be considered to ensure that
treated as a grid network of square areas, years should account for key variables such maximum potential impacts are
with pollutant emissions distributed as growth due to increased or decreased appropriately determined in the assessment.
uniformly within each grid square. Generally, activity, expected emissions controls due to For example, the load or operating condition
input data requirements for air quality regulations, settlement agreements or consent for point sources that causes maximum
models necessitate the use of metric units. As decrees, fuel switches, and any other relevant ground-level concentrations shall be
necessary, any English units common to information. Guidance on emissions established. As a minimum, the source
engineering applications should be estimation techniques (including future year should be modeled using the design capacity
appropriately converted to metric. projections) for SIP attainment
(100 percent load). If a source operates at
b. For point sources, there are many source demonstrations is available.64 95
greater than design capacity for periods that
characteristics and operating conditions that b. For the purpose of SIP revisions for
could result in violations of the NAAQS or
may be needed to appropriately model the stationary point sources, the regulatory
modeling of inert pollutants shall use the PSD increments, this load should be
facility. For example, the plant layout (e.g.,
emissions input data shown in Table 8–1 for modeled. Where the source operates at
location of stacks and buildings), stack
parameters (e.g., height and diameter), boiler short-term and long-term NAAQS. To substantially less than design capacity, and
size and type, potential operating conditions, demonstrate compliance and/or establish the the changes in the stack parameters
and pollution control equipment parameters. appropriate SIP emissions limits, Table 8–1 associated with the operating conditions
Such details are required inputs to air quality generally provides for the use of ‘‘allowable’’ could lead to higher ground level
models and are needed to determine emissions in the regulatory dispersion concentrations, loads such as 50 percent and
maximum potential impacts. modeling of the stationary point source(s) of 75 percent of capacity should also be
c. Modeling mobile emissions from streets interest. In such modeling, these source(s) modeled. Malfunctions which may result in
and highways requires data on the road should be modeled sequentially with these excess emissions are not considered to be a
layout, including the width of each traveled loads for every hour of the year. As part of normal operating condition. They generally
lane, the number of lanes, and the width of a cumulative impact analysis, Table 8–1 should not be considered in determining
the median strip. Additionally, traffic allows for the model user to account for allowable emissions. However, if the excess
patterns should be taken into account (e.g., actual operations in developing the emissions are the result of poor maintenance,
daily cycles of rush hour, differences in emissions inputs for dispersion modeling of careless operation, or other preventable
weekday and weekend traffic volumes, and nearby sources, while other sources are best conditions, it may be necessary to consider
changes in the distribution of heavy-duty represented by air quality monitoring data. them in determining source impact. A range
trucks and light-duty passenger vehicles), as Consultation with the appropriate reviewing of operating conditions should be considered
these patterns will affect the types and authority (paragraph 3.0(b)) is advisable on in screening analyses. The load causing the
amounts of pollutant emissions allocated to the establishment of the appropriate highest concentration, in addition to the
each lane and the height of emissions. emissions inputs for regulatory modeling design load, should be included in refined
d. Emission factors can be determined applications with respect to SIP revisions for modeling.
through source-specific testing and stationary point sources. e. Emissions from mobile sources also have
measurements (e.g., stack test data) from c. For the purposes of demonstrating physical and temporal characteristics that
existing sources or provided from a NAAQS compliance in a PSD assessment, the should be appropriately accounted. For
manufacturing association or vendor. regulatory modeling of inert pollutants shall example, an appropriate emissions model
Additionally, emissions factors for a variety use the emissions input data shown in Table shall be used to determine emissions profiles.
of source types are compiled in an EPA 8–2 for short and long-term NAAQS. The
Such emissions should include speciation
publication commonly known as AP–42.94 new or modifying stationary point source
specific for the vehicle types used on the
AP–42 also provides an indication of the shall be modeled with ‘‘allowable’’ emissions
roadway (e.g., light duty and heavy duty
quality and amount of data on which many in the regulatory dispersion modeling. As
part of a cumulative impact analysis, Table trucks), and subsequent parameterizations of
of the factors are based. Other information the physical emissions characteristics (e.g.,
concerning emissions is available in EPA 8–2 allows for the model user to account for
actual operations in developing the release height) should reflect those emissions
publications relating to specific source sources. For long-term standards, annual
categories. The appropriate reviewing emissions inputs for dispersion modeling of
nearby sources, while other sources are best average emissions may be appropriate, but
authority (paragraph 3.0(b)) should be for short-term standards, discrete temporal
consulted to determine appropriate source represented by air quality monitoring data.
For purposes of situations involving representation of emissions should be used
definitions and for guidance concerning the
emissions trading, refer to current EPA policy (e.g., variations in weekday and weekend
determination of emissions from and
and guidance to establish input data. traffic or the diurnal rush-hour profile typical
techniques for modeling the various source
Consultation with the appropriate reviewing of many cities). Detailed information and
types.
authority (paragraph 3.0(b)) is advisable on data requirements for modeling mobile
8.2.2 Requirements the establishment of the appropriate sources of pollution are provided in the
a. For SIP attainment demonstrations for emissions inputs for regulatory modeling user’s manuals for each of the models
the purpose of projecting future year NAAQS applications with respect to PSD assessments applicable to mobile sources.65 67
attainment for ozone, PM2.5, and regional for a proposed new or modifying source. BILLING CODE 6560–50–P
Table 8-1. - Point Source Model Emission Inputs for SIP Revisions of Inert Pollutants 1
Emissions limit Operating level Operating factor
Averaging time X X
(lb/MMBtul 2 (MMBtu/hrl 2 (e.g., hr/yr, hr/day)
Stationary Point Source(s) Subject to SIP Emissions Limit(s) Evaluation for Compliance with Ambient Standards
(Including Areawide Demonstrations)
Maximum allowable emission Actual or design capacity

Actual operating factor averaged
Annual & quarterly .................... . limit or federally enforceable (whichever is greater), or federally
over the most recent 2 years. 4
permit limit. enforceable permit condition. 3
Continuous operation, i.e., all

Maximum allowable em1ss1on Actual or design capacity
hours of each time period under
Short term(:,; 24 hours) ............ . limit or federally enforceable (whichever is greater), or federally
consideration (for all hours of the
permit limit. enforceable permit condition. 3
meteorological database). 5
Nearby Source(s) 6
Maximum allowable emission Annual level when actually

Annual & quarterly .................... . limit or federally enforceable operating, averaged over the most
over the most recent 2 years. 4 • 8
permit limit. 6 recent 2 years. 4
Temporally representative level Continuous operation, i.e., all

Maximum allowable emission
when actually operating, hours of each time period under
Short term(:,; 24 hours) ............ . limit or federally enforceable
reflective of the most recent 2 consideration (for all hours of the
permit limit. 6
years. 4 ' 7 meteorological database). 5
Other Source(s) 6' 9
The ambient impacts from Non-nearby or Other Sources (e.g., natural, minor, distant major, and unidentified sources) can be represented by air
quality monitoring data unless adequate data do not exist.
1. For purposes of emissions trading, NSR, or PSD, other model input criteria may apply. See Section 8.2 for more information regarding attainment
demonstrations of primary PM2.5.
2. Terminology applicable to fuel burning sources; analogous terminology (e.g., lb/throughput) may be used for other types of sources.
3. Operating levels such as SO percent and 75 percent of capacity should also be modeled to determine the load causing the highest concentration.
4. Unless it is determined that this period is not representative.
5. If operation does not occur for all hours of the time period of consideration (e.g., 3 or 24-hours) and the source operation is constrained by a
federally enforceable permit condition, an appropriate adjustment to the modeled emission rate may be made (e.g., if operation is only 8 a.m. to 4
p.m. each day, only these hours will be modeled with emissions from the source. Modeled emissions should not be averaged across non-operating
time periods.)
6. See Section 8.3.3.
7. Temporally representative operating level could be based on Continuous Emissions Monitoring (CEM) data or other information and should be
determined through consultation with the appropriate reviewing authority (Paragraph 3.0(b)).
8. For those permitted sources not in operation or that have not established an appropriate factor, continuous operation (i.e., 8760) should be used.
EP23OC23.001</GPH>
Table 8-2. - Point Source Model Emission Inputs for NAAQS Compliance in PSD Demonstrations
Emissions limit Operating level Operating factor
Averaging time X X
(lb/MMBtul 1 (MMBtu/hrl 1 (e.g., hr/yr, hr/day)
Proposed Major New or Modified Source
Maximum allowable emission

Design capacity or federally Continuous operation (i.e., 8760
Annual & quarterly .................... . limit or federally enforceable
enforceable permit condition. 2 hours). 3
permit limit.
Continuous operation, i.e., all

Maximum allowable em1ss1on
Design capacity or federally hours of each time period under
Short term(::; 24 hours) ............ . limit or federally enforceable
enforceable permit condition. 2 consideration (for all hours of the
permit limit.
meteorological database). 3
Nearby Source(s) 4' 5
Maximum allowable emission Annual level when actually

Annual & quarterly .................... . limit or federally enforceable operating, averaged over the most
over the most recent 2 years. 6 • 8
permit limit. 5 recent 2 years. 6
Temporally representative level Continuous operation, i.e., all

Maximum allowable em1ss1on
when actually operating, hours of each time period under
Short term(::; 24 hours) ............ . limit or federally enforceable
reflective of the most recent 2 consideration (for all hours of the
permit limit. 5
years. 6 ' 7 meteorological database). 3
Other Source(s) 5' 9
The ambient impacts from Non-nearby or Other Sources (e.g., natural, minor, distant major, and unidentified sources) can be represented by air
quality monitoring data unless adequate data do not exist.
1. Terminology applicable to fuel burning sources; analogous terminology (e.g., lb/throughput) may be used for other types of sources.
2. Operating levels such as SO percent and 75 percent of capacity should also be modeled to determine the load causing the highest concentration.
3. If operation does not occur for all hours of the time period of consideration (e.g., 3 or 24-hours) and the source operation is constrained by a
federally enforceable permit condition, an appropriate adjustment to the modeled emission rate may be made (e.g., if operation is only 8 a.m. to 4
p.m. each day, only these hours will be modeled with emissions from the source. Modeled emissions should not be averaged across non-operating
time periods.
4. Includes existing facility to which modification is proposed if the emissions from the existing facility will not be affected by the modification.
Otherwise use the same parameters as for major modification.
6. Unless it is determined that this period is not representative.
7. Temporally representative operating level could be based on Continuous Emissions Monitoring (CEM) data or other information and should be
determined through consultation with the appropriate reviewing authority (Paragraph 3.0(b)).
8. For those permitted sources not in operation or that have not established an appropriate factor, continuous operation (i.e., 8760) should be used.
BILLING CODE 6560–50–C that should be used in parallel with the other unidentified sources in the vicinity of
8.3 Background Concentrations recommendations made in this section. the project, and regional transport
Generally, background air quality should not contributions from more distant sources
8.3.1 Discussion include the ambient impacts of the project (domestic and international). The ambient
a. Background concentrations are essential source under consideration. Instead, it contributions from these sources are typically
in constructing the design concentration, or should include: accounted for through use of ambient
total air quality concentration, as part of a i. Nearby sources: These are individual monitoring data or, in some cases, regional-
cumulative impact analysis for NAAQS and sources located in the vicinity of the scale photochemical grid modeling results.
PSD increments (section 9.2.3). To assist source(s) under consideration for emissions b. The monitoring network used for
applicants and reviewing authorities with limits that are not adequately represented by developing background concentrations is
appropriately characterizing background ambient monitoring data. The ambient expected to conform to the same quality
concentrations, EPA has developed the Draft contributions from these nearby sources are assurance and other requirements as those
Guidance on Developing Background thereby accounted for by explicitly modeling networks established for PSD purposes.97
Concentrations for Use in Modeling their emissions (section 8.2). Accordingly, the air quality monitoring data
Demonstrations.96 The guidance provides a ii. Other sources: That portion of the should be of sufficient completeness and
EP23OC23.002</GPH>
recommended framework composed of steps background attributable to natural sources, follow appropriate data validation
procedures. These data should be adequately ambient monitoring data and determining the contribute to the temporal and spatial
representative of the area to inform appropriate method for combining this variability of ambient concentrations across a
calculation of the design concentration for monitor-based background contribution to typical modeling domain on an hourly basis.
comparison to the applicable NAAQS the modeled impact of the project and other In most cases, the seasonal (or quarterly)
(section 9.2.2). nearby sources. For many cases, the best pairing of monitored and modeled
c. For photochemical grid modeling starting point would be use of the current concentrations should sufficiently address
conducted in SIP attainment demonstrations design value for the applicable NAAQS as a situations to which the impacts from
for ozone, PM2.5 and regional haze, the uniform monitored background contribution modeled emissions are not temporally
emissions from nearby and other sources are across the project area. However, there are correlated with background monitored levels.
included as model inputs and fully cases in which the current design value may e. In those cases where adequately
accounted for in the modeling application not be appropriate. Such cases include but representative monitoring data to
and predicted concentrations. The concept of are not limited to: characterize background concentrations are
adding individual components to develop a i. For situations involving a modifying not available, it may be appropriate to use
design concentration, therefore, do not apply source where the existing facility is results from a regional-scale photochemical
in these SIP applications. However, such determined to impact the ambient monitor, grid model, or other representative model
modeling results may then be appropriate for the background concentration at each application, as background concentrations
consideration in characterizing background monitor can be determined by excluding consistent with the considerations discussed
concentrations for other regulatory values when the source in question is above and in consultation with the
applications. Also, as noted in section 5, this impacting the monitor. In such cases, appropriate reviewing authority (paragraph
modeling approach does provide for an monitoring sites inside a 90° sector 3.0(b)).
appropriate atmospheric environment to downwind of the source may be used to 8.3.3 Recommendations for Multi-Source
assess single-source impacts for ozone and determine the area of impact. Areas
secondary PM2.5. ii. There may be other circumstances
d. For NAAQS assessments and SIP which would necessitate modifications to the a. In multi-source areas, determining the
attainment demonstrations for inert ambient data record. Such cases could appropriate background concentration
pollutants, the development of the include removal of data from specific days or involves: (1) characterization of contributions
appropriate background concentration for a hours when a monitor is being impacted by from other sources through adequately
cumulative impact analysis involves proper activities that are not typical or not expected representative ambient monitoring data, and
accounting of each contribution to the design (2) identification and characterization of
to occur again in the future (e.g.,
concentration and will depend upon whether contributions from nearby sources through
construction, roadway repairs, forest fires, or
the project area’s situation consists of either explicit modeling. A key point here is the
unusual agricultural activities). There may
an isolated single source(s) or a multitude of interconnectedness of each component in
also be cases where it may be appropriate to
sources. For PSD increment assessments, all that the question of which nearby sources to
scale (multiplying the monitored
impacts after the appropriate baseline dates include in the cumulative modeling is
concentrations with a scaling factor) or adjust
(i.e., trigger date, major source baseline date, inextricably linked to the question of what
(adding or subtracting a constant value the
and minor source baseline date) from all the ambient monitoring data represents
monitored concentrations) data from specific
increment-consuming and increment- within the project area.
days or hours. Such adjustments would make b. Nearby sources: All sources in the
expanding sources should be considered in the monitored background concentrations
the design concentration (section 9.2.2). vicinity of the source(s) under consideration
more temporally and/or spatially for emissions limits that are not adequately
8.3.2 Recommendations for Isolated Single representative of the area around the new or represented by ambient monitoring data
Sources modifying source for the purposes of the should be explicitly modeled. EPA’s
a. In areas with an isolated source(s), regulatory assessment. recommended framework for determining an
determining the appropriate background iii. For short-term standards, the diurnal or appropriate background concentration 96
concentration should focus on seasonal patterns of the air quality should be applied to identify such sources
characterization of contributions from all monitoring data may differ significantly from and accurately account for their ambient
other sources through adequately the patterns associated with the modeled impacts through explicit modeling.
representative ambient monitoring data. The concentrations. When this occurs, it may be i. The determination of nearby sources
application of EPA’s recommended appropriate to pair the air quality monitoring relies on the selection of adequately
framework for determining an appropriate data in a temporal manner that reflects these representative ambient monitoring data
background concentration should be patterns (e.g., pairing by season and/or hour (section 8.3.2). The EPA recommends
consistent with appropriate EPA modeling of day).98 determining the representativeness of the
guidance 63 96 and justified in the modeling iv. For situations where monitored air monitoring data through a visual assessment
protocol that is vetted with the appropriate quality concentrations vary across the of the modeling domain considering any
reviewing authority (paragraph 3.0(b)). modeling domain, it may be appropriate to relevant nearby sources and their respective
b. The EPA recommends use of the most consider air quality monitoring data from air quality data. The visual assessment
recent quality assured air quality monitoring multiple monitors within the project area. should consider any relevant air quality data
data collected in the vicinity of the source to d. Considering the spatial and temporal such as the proximity of nearby sources to
determine the background concentration for variability throughout a typical modeling the project source and the ambient monitor,
the averaging times of concern. In most cases, domain on an hourly basis and the the nearby source’s level of emissions with
the EPA recommends using data from the complexities and limitations of hourly respect to the ambient data, and the
monitor closest to and upwind of the project observations from the ambient monitoring dispersion environment (i.e., meteorological
area. If several monitors are available, network, the EPA does not recommend patterns, terrain, etc.) of the modeling
preference should be given to the monitor hourly or daily pairing of monitored domain.
with characteristics that are most similar to background and modeled concentrations ii. Nearby sources not adequately
the project area. If there are no monitors except in rare cases of relatively isolated represented by the ambient monitor through
located in the vicinity of the new or sources where the available monitor can be visual assessment should undergo further
modifying source, a ‘‘regional site’’ may be shown to be representative of the ambient qualitative and quantitative analysis before
used to determine background concentration levels in the areas of maximum being explicitly modeled. The EPA
concentrations. A regional site is one that is impact from the proposed new source. The recommends evaluating any modeling,
located away from the area of interest but is implicit assumption underlying hourly monitoring, or emissions data that may be
impacted by similar or adequately pairing is that the background monitored available for the identified nearby sources
representative sources. levels for each hour are spatially uniform and with respect to possible exceedances of the
c. Many of the challenges related to that the monitored values are fully appropriate SIL or violations to the NAAQS.
cumulative impact analyses arise in the representative of background levels at each iii. The number of nearby sources to be
context of defining the appropriate metric to receptor for each hour. Such an assumption explicitly modeled in the air quality analysis
characterize background concentrations from clearly ignores the many factors that is expected to be few except in unusual
situations. The determination of nearby 8.4 Meteorological Input Data processor, in most cases, should be used to
sources through the application of EPA’s 8.4.1 Discussion process 1-minute ASOS wind data for input
recommended framework calls for the to AERMET when processing NWS ASOS
a. This subsection covers meteorological sites in AERMET. When processing
exercise of professional judgment by the
input data for use in dispersion modeling for prognostic meteorological data for AERMOD,
appropriate reviewing authority (paragraph
regulatory applications and is separate from the Mesoscale Model Interface Program
3.0(b)) and should be consistent with recommendations made for photochemical
appropriate EPA modeling guidance.63 96 This (MMIF) 109 should be used to process data for
grid modeling. Recommendations for input to AERMET, both for land-based
guidance is not intended to alter the exercise meteorological data for photochemical grid applications and overwater applications.
of that judgment or to comprehensively modeling applications are outlined in the Other methods of processing prognostic
prescribe which sources should be included latest version of EPA’s Modeling Guidance meteorological data for input to AERMET
as nearby sources. for Demonstrating Attainment of Air Quality should be approved by the appropriate
c. For cumulative impact analyses of short- Goals for Ozone, PM2.5, and Regional Haze.64 reviewing authority. Additionally, the
term and annual ambient standards, the In cases where Lagrangian models are following meteorological preprocessors are
nearby sources as well as the project applied for regulatory purposes, appropriate recommended by the EPA: PCRAMMET,102
source(s) must be evaluated using an meteorological inputs should be determined MPRM,103 and METPRO.104 PCRAMMET is
appropriate addendum A model or approved in consultation with the appropriate the recommended meteorological data
alternative model with the emission input reviewing authority (paragraph 3.0(b)). preprocessor for use in applications of OCD
data shown in Table 8–1 or 8–2. b. The meteorological data used as input to employing hourly NWS data. MPRM is the
i. When modeling a nearby source that a dispersion model should be selected on the recommended meteorological data
does not have a permit and the emissions basis of spatial and climatological (temporal) preprocessor for applications of OCD
representativeness as well as the ability of employing site-specific meteorological data.
limits contained in the SIP for a particular
the individual parameters selected to METPRO is the recommended meteorological
source category is greater than the emissions
characterize the transport and dispersion data preprocessor for use with
possible given the source’s maximum conditions in the area of concern. The
physical capacity to emit, the ‘‘maximum CTDMPLUS.105
representativeness of the measured data is b. Regulatory application of AERMOD
allowable emissions limit’’ for such a nearby dependent on numerous factors including,
source may be calculated as the emissions necessitates careful consideration of the
but not limited to: (1) the proximity of the meteorological data for input to AERMET.
rate representative of the nearby source’s meteorological monitoring site to the area
maximum physical capacity to emit, Data representativeness, in the case of
under consideration; (2) the complexity of AERMOD, means utilizing data of an
considering its design specifications and the terrain; (3) the exposure of the appropriate type for constructing realistic
allowable fuels and process materials. meteorological monitoring site; and (4) the
boundary layer profiles. Of particular
However, the burden is on the permit period of time during which data are
importance is the requirement that all
applicant to sufficiently document what the collected. The spatial representativeness of
meteorological data used as input to
maximum physical capacity to emit is for the data can be adversely affected by large
AERMOD should be adequately
such a nearby source. distances between the source and receptors
representative of the transport and dispersion
ii. It is appropriate to model nearby sources of interest and the complex topographic
within the analysis domain. Where surface
only during those times when they, by their characteristics of the area. Temporal
conditions vary significantly over the
nature, operate at the same time as the representativeness is a function of the year-
analysis domain, the emphasis in assessing
primary source(s) or could have impact on to-year variations in weather conditions.
Where appropriate, data representativeness representativeness should be given to
the averaging period of concern. Accordingly, adequate characterization of transport and
should be viewed in terms of the
it is not necessary to model impacts of a dispersion between the source(s) of concern
appropriateness of the data for constructing
nearby source that does not, by its nature, and areas where maximum design
realistic boundary layer profiles and, where
operate at the same time as the primary applicable, three-dimensional meteorological concentrations are anticipated to occur. The
source or could have impact on the averaging fields, as described in paragraphs (c) and (d) EPA recommends that the surface
period of concern, regardless of an identified of this subsection. characteristics input to AERMET should be
significant concentration gradient from the c. The meteorological data should be representative of the land cover in the
nearby source. The burden is on the permit adequately representative and may be site- vicinity of the meteorological data, i.e., the
applicant to adequately justify the exclusion specific data (land-based or buoy data for location of the meteorological tower for
of nearby sources to the satisfaction of the overwater applications), data from a nearby measured data or the representative grid cell
appropriate reviewing authority (paragraph National Weather Service (NWS) or for prognostic data. Therefore, the model user
3.0(b)). The following examples illustrate two comparable station, or prognostic should apply the latest version
cases in which a nearby source may be meteorological data. The implementation of AERSURFACE,106 107 where applicable, for
shown not to operate at the same time as the NWS Automated Surface Observing Stations determining surface characteristics when
primary source(s) being modeled: (1) (ASOS) in the early 1990’s should not processing measured land-based
Seasonal sources (only used during certain preclude the use of NWS ASOS data if such meteorological data through AERMET. In
seasons of the year). Such sources would not a station is determined to be representative areas where it is not possible to use
be modeled as nearby sources during times of the modeled area.99 AERSURFACE output, surface characteristics
D. Model input data are normally obtained can be determined using techniques that
in which they do not operate; and (2)
either from the NWS or as part of a site- apply the same analysis as AERSURFACE. In
Emergency backup generators, to the extent
specific measurement program. State the case of measured meteorological data for
that they do not operate simultaneously with
climatology offices, local universities, FAA, overwater applications, AERMET calculates
the sources that they back up. Such the surface characteristics and AERSURFACE
emergency equipment would not be modeled military stations, industry, and pollution
control agencies may also be sources of such outputs are not needed. In the case of
as nearby sources. prognostic meteorological data, the surface
d. Other sources. That portion of the data. In specific cases, prognostic
meteorological data may be appropriate for characteristics associated with the prognostic
background attributable to all other sources meteorological model output for the
(e.g., natural, minor, distant major, and use and obtained from similar sources. Some
representative grid cell should be used.108 109
recommendations and requirements for the

unidentified sources) should be accounted Furthermore, since the spatial scope of each
use of each type of data are included in this
for through use of ambient monitoring data variable could be different,
subsection.
and determined by the procedures found in representativeness should be judged for each
section 8.3.2 in keeping with eliminating or 8.4.2 Recommendations and Requirements variable separately. For example, for a
reducing the source-oriented impacts from a. AERMET 100 shall be used to preprocess variable such as wind direction, the data
nearby sources to avoid potential double- all meteorological data, be it observed or should ideally be collected near plume
counting of modeled and monitored prognostic, for use with AERMOD in height to be adequately representative,
contributions. regulatory applications. The AERMINUTE 101 especially for sources located in complex
terrain. Whereas, for a variable such as model users. Although the NWS does not When such data are available for the NWS
temperature, data from a station several provide direct measurements of all the ASOS site being processed, the AERMINUTE
kilometers away from the source may be needed dispersion model input variables, processor should be used, in most cases, to
considered to be adequately representative. methods have been developed and calculate hourly average wind speed and
More information about meteorological data, successfully used to translate the basic NWS direction when processing NWS ASOS data
representativeness, and surface data to the needed model input. Site-specific for input to AERMOD.99
characteristics can be found in the AERMOD measurements of model input parameters e. Data from universities, FAA, military
Implementation Guide.81 have been made for many modeling studies, stations, industry and pollution control
c. Regulatory application of CTDMPLUS and those methods and techniques are agencies may be used if such data are
requires the input of multi-level becoming more widely applied, especially in equivalent in accuracy and detail (e.g., siting
measurements of wind speed, direction, situations such as complex terrain criteria, frequency of observations, data
temperature, and turbulence from an applications, where available NWS data are completeness, etc.) to the NWS data, they are
appropriately sited meteorological tower. The not adequately representative. However, judged to be adequately representative for the
measurements should be obtained up to the there are many modeling applications where particular application, and have undergone
representative plume height(s) of interest. NWS data are adequately representative and quality assurance checks.
Plume heights of interest can be determined the applications still rely heavily on the NWS f. After valid data retrieval requirements
by use of screening procedures such as data. have been met,110 large number of hours in
CTSCREEN. b. Many models use the standard hourly the record having missing data should be
d. Regulatory application of OCD requires weather observations available from the treated according to an established data
meteorological data over land and over water. National Centers for Environmental substitution protocol provided that
The over land or surface data, processed Information (NCEI).b These observations are adequately representative alternative data are
through PCRAMMET 102 or MPRM,103 that then preprocessed before they can be used in available. Data substitution guidance is
provides hourly stability class, wind the models. Prior to the advent of ASOS in provided in section 5.3 of reference 110.110
direction and speed, ambient temperature, the early 1990’s, the standard ‘‘hourly’’ If no representative alternative data are
and mixing height, are required. Data over weather observation was a human-based available for substitution, the absent data
water requires hourly mixing height, relative observation reflecting a single 2-minute should be coded as missing using missing
humidity, air temperature, and water surface average generally taken about 10 minutes data codes appropriate to the applicable
temperature. Missing winds are substituted before the hour. However, beginning in meteorological pre-processor. Appropriate
with the surface winds. Vertical wind January 2000 for first-order stations and in model options for treating missing data, if
direction shear, vertical temperature March 2005 for all stations, the NCEI has available in the model, should be employed.
gradient, and turbulence intensities are archived the 1-minute ASOS wind data (i.e.,
optional. the rolling 2-minute average winds) for the 8.4.4 Site-Specific Data
e. The model user should acquire enough NWS ASOS sites. The AERMINUTE 8.4.4.1 Discussion
meteorological data to ensure that worst-case processor 101 was developed to reduce the
number of calm and missing hours in a. Spatial or geographical
meteorological conditions are adequately
AERMET processing by substituting standard representativeness is best achieved by
represented in the model results. The use of
hourly observations with full hourly average collection of all of the needed model input
5 years of adequately representative NWS or
winds calculated from 1-minute ASOS wind data in close proximity to the actual site of
comparable meteorological data, at least 1
data. the source(s). Site-specific measured data are,
year of site-specific (either land-based or
therefore, preferred as model input, provided
overwater based), or at least 3 years of 8.4.3.2 Recommendations
prognostic meteorological data, are required. that appropriate instrumentation and quality
a. The preferred models listed in assurance procedures are followed, and that
If 1 year or more, up to 5 years, of site-
addendum A all accept, as input, the NWS the data collected are adequately
specific data are available, these data are
meteorological data preprocessed into model representative (free from inappropriate local
preferred for use in air quality analyses.
compatible form. If NWS data are judged to or microscale influences) and compatible
Depending on completeness of the data
be adequately representative for a specific with the input requirements of the model to
record, consecutive years of NWS, site-
modeling application, they may be used. The be used. It should be noted that, while site-
specific, or prognostic data are preferred.
NCEI makes available surface and upper air specific measurements are frequently made
Such data must be subjected to quality
meteorological data online and in CD–ROM ‘‘on-property’’ (i.e., on the source’s premises),
assurance procedures as described in section format. Upper air data are also available at
8.4.4.2. acquisition of adequately representative site-
the Earth System Research Laboratory Global specific data does not preclude collection of
f. Objective analysis in meteorological Systems Divisions website and from NCEI.
modeling is to improve meteorological data from a location off property. Conversely,
For the latest websites of available surface collection of meteorological data on a
analyses (the ‘‘first guess field’’) used as and upper air data see reference 100.
initial conditions for prognostic source’s property does not of itself guarantee
b. Although most NWS wind adequate representativeness. For help in
meteorological models by incorporating measurements are made at a standard height
information from meteorological determining representativeness of site-
of 10 m, the actual anemometer height specific measurements, technical
observations. Direct and indirect (using should be used as input to the preferred
remote sensing techniques) observations of guidance 110 is available. Site-specific data
meteorological processor and model. should always be reviewed for
temperature, humidity, and wind from c. Standard hourly NWS wind directions
surface and radiosonde reports are commonly representativeness and adequacy by an
are reported to the nearest 10 degrees. Due experienced meteorologist, atmospheric
employed to improve these analysis fields. to the coarse resolution of these data, a
For long-range transport applications, it is scientist, or other qualified scientist in
specific set of randomly generated numbers consultation with the appropriate reviewing
recommended that objective analysis has been developed by the EPA and should
procedures, using direct and indirect authority (paragraph 3.0(b)).
be used when processing standard hourly
meteorological observations, be employed in NWS data for use in the preferred EPA 8.4.4.2 Recommendations
preparing input fields to produce prognostic models to ensure a lack of bias in wind a. The EPA guidance 110 provides
meteorological datasets. The length of record direction assignments within the models. recommendations on the collection and use
of observations should conform to d. Beginning with year 2000, NCEI began of site-specific meteorological data.
recommendations outlined in paragraph archiving 2-minute winds, reported every Recommendations on characteristics, siting,
8.4.2(e) for prognostic meteorological model minute to the nearest degree for NWS ASOS and exposure of meteorological instruments
datasets. sites. The AERMINUTE processor was and on data recording, processing,
8.4.3 National Weather Service Data developed to read those winds and calculate completeness requirements, reporting, and
hourly average winds for input to AERMET. archiving are also included. This publication
8.4.3.1 Discussion should be used as a supplement to other
a. The NWS meteorological data are b Formerly the National Climatic Data Center limited guidance on these subjects.5 97 111 112
routinely available and familiar to most (NCDC). Detailed information on quality assurance is
also available.113 As a minimum, site-specific direct measurements of turbulence (wind prohibitive or infeasible to collect adequately
measurements of ambient air temperature, fluctuations) in the characterization of the representative site-specific data. For these
transport wind speed and direction, and the vertical and lateral dispersion (e.g., cases, it may be appropriate to use prognostic
variables necessary to estimate atmospheric CTDMPLUS or AERMOD). When turbulence meteorological data, if deemed adequately
dispersion should be available in data are used to directly characterize the representative, in a regulatory modeling
meteorological datasets to be used in vertical and lateral dispersion, the averaging application. However, if prognostic
modeling. Care should be taken to ensure time for the turbulence measurements should meteorological data are not representative of
that meteorological instruments are located be 1-hour. For technical guidance on transport and dispersion conditions in the
to provide an adequately representative processing of turbulence parameters for use area of concern, the collection of site-specific
characterization of pollutant transport in dispersion modeling, refer to the user’s data is necessary.
between sources and receptors of interest. guide to the meteorological processor for b. The EPA has developed a processor, the
The appropriate reviewing authority each model (see section 8.4.2(a)). MMIF,108 to process MM5 (Mesoscale Model
(paragraph 3.0(b)) is available to help ii. Stability categories. For dispersion 5) or WRF (Weather Research and
determine the appropriateness of the models that employ P–G stability categories Forecasting) model data for input to various
measurement locations. for the characterization of the vertical and models including AERMOD. MMIF can
i. Solar radiation measurements. Total lateral dispersion, the P–G stability process data for input to AERMET or
solar radiation or net radiation should be categories, as originally defined, couple near- AERMOD for a single grid cell or multiple
measured with a reliable pyranometer or net surface measurements of wind speed with grid cells. MMIF output has been found to
radiometer sited and operated in accordance subjectively determined insolation compare favorably against observed data
with established site-specific meteorological assessments based on hourly cloud cover and (site-specific or NWS).120 Specific guidance
guidance.110 113 ceiling height observations. The wind speed on processing MMIF for AERMOD can be
ii. Temperature measurements. measurements are made at or near 10 m. The found in reference 109109. When using
Temperature measurements should be made insolation rate is typically assessed using MMIF to process prognostic data for
at standard shelter height (2m) in accordance observations of cloud cover and ceiling regulatory applications, the data should be
with established site-specific meteorological height based on criteria outlined by Turner.77 processed to generate AERMET inputs and
guidance.110 It is recommended that the P–G stability the data subsequently processed through
iii. Temperature difference measurements. category be estimated using the Turner AERMET for input to AERMOD. If an
Temperature difference (DT) measurements method with site-specific wind speed alternative method of processing data for
should be obtained using matched measured at or near 10 m and representative input to AERMET is used, it must be
thermometers or a reliable thermocouple cloud cover and ceiling height. approved by the appropriate reviewing
system to achieve adequate accuracy. Siting, Implementation of the Turner method, as authority (paragraph 3.0(b)).
probe placement, and operation of DT well as considerations in determining
systems should be based on guidance found 8.4.5.2 Recommendations
representativeness of cloud cover and ceiling
in Chapter 3 of reference 110 and such height in cases for which site-specific cloud a. Prognostic model evaluation.
guidance should be followed when obtaining observations are unavailable, may be found Appropriate effort by the applicant should be
vertical temperature gradient data. AERMET in section 6 of reference 110. In the absence devoted to the process of evaluating the
may employ the Bulk Richardson scheme, of requisite data to implement the Turner prognostic meteorological data. The
which requires measurements of temperature method, the solar radiation/delta-T (SRDT) modeling data should be compared to NWS
difference, in lieu of cloud cover or method or wind fluctuation statistics (i.e., the observational data or other comparable data
insolation data. To ensure correct application sE and sA methods) may be used. in an effort to show that the data are
and acceptance, AERMOD users should iii. The SRDT method, described in section adequately replicating the observed
consult with the appropriate reviewing 6.4.4.2 of reference 110, is modified slightly meteorological conditions of the time periods
authority (paragraph 3.0(b)) before using the from that published from earlier work 114 and modeled. An operational evaluation of the
Bulk Richardson scheme for their analysis. has been evaluated with three site-specific modeling data for all model years (i.e.,
iv. Wind measurements. For simulation of databases.115 The two methods of stability statistical, graphical) should be completed.64
plume rise and dispersion of a plume emitted classification that use wind fluctuation The use of output from prognostic mesoscale
from a stack, characterization of the wind meteorological models is contingent upon the
statistics, the sE and sA methods, are also
profile up through the layer in which the concurrence with the appropriate reviewing
described in detail in section 6.4.4 of
plume disperses is desirable. This is authority (paragraph 3.0(b)) that the data are
reference 110 (note applicable tables in
especially important in complex terrain and/ of acceptable quality, which can be
section 6). For additional information on the
or complex wind situations where wind demonstrated through statistical comparisons
wind fluctuation methods, several references
measurements at heights up to hundreds of with meteorological observations aloft and at
are available.116 117 118 119
meters above stack base may be required in the surface at several appropriate locations.64
c. Missing data substitution. After valid
some circumstances. For tall stacks when b. Representativeness. When processing
data retrieval requirements have been met,110
site-specific data are needed, these winds MMIF data for use with AERMOD, the grid
hours in the record having missing data
have been obtained traditionally using cell used for the dispersion modeling should
meteorological sensors mounted on tall should be treated according to an established be adequately spatially representative of the
towers. A feasible alternative to tall towers is data substitution protocol provided that analysis domain. In most cases, this may be
the use of meteorological remote sensing adequately representative alternative data are the grid cell containing the emission source
instruments (e.g., acoustic sounders or radar available. Such protocols are usually part of of interest. Since the dispersion modeling
wind profilers) to provide winds aloft, the approved monitoring program plan. Data may involve multiple sources and the
coupled with 10-meter towers to provide the substitution guidance is provided in section domain may cover several grid cells,
near-surface winds. Note that when site- 5.3 of reference 110. If no representative depending on grid resolution of the
specific wind measurements are used, alternative data are available for substitution, prognostic model, professional judgment may
AERMOD, at a minimum, requires wind the absent data should be coded as missing, be needed to select the appropriate grid cell
observations at a height above ground using missing data codes appropriate to the to use. In such cases, the selected grid cells
between seven times the local surface applicable meteorological pre-processor. should be adequately representative of the
roughness height and 100 m. (For additional Appropriate model options for treating entire domain.
requirements for AERMOD and CTDMPLUS, missing data, if available in the model, c. Grid resolution. The grid resolution of
see addendum A.) Specifications for wind should be employed. the prognostic meteorological data should be
measuring instruments and systems are 8.4.5 Prognostic Meteorological Data considered and evaluated appropriately,
contained in reference 110. particularly for projects involving complex
b. All processed site-specific data should 8.4.5.1 Discussion terrain. The operational evaluation of the
be in the form of hourly averages for input a. For some modeling applications, there modeling data should consider whether a
to the dispersion model. may not be a representative NWS or finer grid resolution is needed to ensure that
i. Turbulence data. There are several comparable meteorological station available the data are representative. The use of output
dispersion models that are capable of using (e.g., complex terrain), and it may be cost from prognostic mesoscale meteorological
models is contingent upon the concurrence plume models assume that concentration is average. For annual averages, the sum of all
with the appropriate reviewing authority inversely proportional to wind speed, valid hourly concentrations is divided by the
(paragraph 3.0(b)) that the data are of depending on model formulations. number of non-calm hours during the year.
acceptable quality. Procedures have been developed to prevent AERMOD has been coded to implement these
8.4.6 Marine Boundary Layer Environments the occurrence of overly conservative instructions. For hours that are calm or
concentration estimates during periods of missing, the AERMOD hourly concentrations
8.4.6.1 Discussion calms. These procedures acknowledge that a will be zero. For other models listed in
a. Calculations of boundary layer steady-state Gaussian plume model does not addendum A, a post-processor computer
parameters for the marine boundary layer apply during calm conditions, and that our program, CALMPRO 121 has been prepared, is
present special challenges as the marine knowledge of wind patterns and plume available on the EPA’s SCRAM website
boundary layer can be very different from the behavior during these conditions does not, at (section 2.3), and should be used.
boundary layer over land. For example, present, permit the development of a better b. Stagnant conditions that include
convective conditions can occur in the technique. Therefore, the procedures extended periods of calms often produce
overnight hours in the marine boundary layer disregard hours that are identified as calm. high concentrations over wide areas for
while typically over land, stable conditions The hour is treated as missing and a relatively long averaging periods. The
occur at night. Also, surface roughness in the convention for handling missing hours is standard steady-state Gaussian plume models
marine environment is a function of wave recommended. With the advent of the are often not applicable to such situations.
height and wind speed and less static with AERMINUTE processor, when processing When stagnation conditions are of concern,
time than surface roughness over land. NWS ASOS data, the inclusion of hourly other modeling techniques should be
b. While the Offshore and Coastal averaged winds from AERMINUTE will, in considered on a case-by-case basis (see also
Dispersion Model (OCD) is the preferred some instances, dramatically reduce the section 7.2.1.2).
model for overwater applications, there are number of calm and missing hours, c. When used in steady-state Gaussian
applications where the use of AERMOD is especially when the ASOS wind are derived plume models other than AERMOD,
applicable. These include applications that from a sonic anemometer. To alleviate measured site-specific wind speeds of less
utilize features of AERMOD not included in concerns about these issues, especially those than 1 m/s but higher than the response
OCD (e.g., NO2 chemistry). Such use of introduced with AERMINUTE, the EPA threshold of the instrument should be input
AERMOD would require consultation with implemented a wind speed threshold in as 1 m/s; the corresponding wind direction
the Regional Office and appropriate AERMET for use with ASOS derived should also be input. Wind observations
reviewing authority to ensure that platform winds.99 100 Winds below the threshold will below the response threshold of the
downwash and shoreline fumigation are be treated as calms. instrument should be set to zero, with the
adequately considered in the modeling b. AERMOD, while fundamentally a input file in ASCII format. For input to
demonstration. steady-state Gaussian plume model, contains AERMOD, no such adjustment should be
c. For the reasons stated above, a algorithms for dealing with low wind speed made to the site-specific wind data, as
standalone pre-processor to AERMOD, called (near calm) conditions. As a result, AERMOD AERMOD has algorithms to account for light
AERCOARE 47 was developed to use the can produce model estimates for conditions or variable winds as discussed in section
Coupled Ocean Atmosphere Response when the wind speed may be less than 1 m/ 8.4.6.1(a). For NWS ASOS data, see the
Experiment (COARE) bulk-flux algorithms 48 s, but still greater than the instrument AERMET User’s Guide 100 for guidance on
to bypass AERMET and calculate the threshold. Required input to AERMET for wind speed thresholds. For prognostic data,
boundary layer parameters for input to site-specific data, the meteorological see the latest guidance 109 for thresholds.
AERMOD for the marine boundary layer. processor for AERMOD, includes a threshold Observations with wind speeds less than the
AERCOARE can process either measurements wind speed and a reference wind speed. The threshold are considered calm, and no
from water-based sites such as buoys or threshold wind speed is the greater of the concentration is calculated. In all cases
prognostic data. To better facilitate the use of threshold of the instrument used to collect involving steady-state Gaussian plume
the COARE algorithms for AERMOD, EPA the wind speed data or wind direction models, calm hours should be treated as
has included the COARE algorithms into sensor.110 The reference wind speed is missing, and concentrations should be
selected by the model as the lowest level of calculated as in paragraph (a) of this
AERMET thus eliminating the need for a
non-missing wind speed and direction data subsection.
standalone pre-processor and ensuring the
where the speed is greater than the wind
algorithms are updated as part of routine 9.0 Regulatory Application of Models
speed threshold, and the height of the
AERMET updates.
measurement is between seven times the 9.1 Discussion
8.4.6.2 Recommendations local surface roughness length and 100 m. If
a. Standardized procedures are valuable in
a. Measured data. For applications in the the only valid observation of the reference
the review of air quality modeling and data
marine environment that require the use of wind speed between these heights is less
analyses conducted to support SIP submittals
AERMOD, measured surface data, such as than the threshold, the hour is considered
and revisions, NSR, or other EPA
from a buoy or other offshore platform, calm, and no concentration is calculated.
requirements to ensure consistency in their
should be processed in AERMET with the None of the observed wind speeds in a
regulatory application. This section
COARE processing option following measured wind profile that are less than the
recommends procedures specific to NSR that
recommendations in the AERMET User’s threshold speed are used in construction of
facilitate some degree of standardization
Guide 100 and AERMOD Implementation the modeled wind speed profile in AERMOD.
while at the same time allowing the
Guide.81 For applications in the marine 8.4.7.2 Recommendations flexibility needed to assure the technically
environment that require the use of OCD, a. Hourly concentrations calculated with best analysis for each regulatory application.
users should use the recommended steady-state Gaussian plume models using For SIP attainment demonstrations, refer to
meteorological pre-processor MPRM. calms should not be considered valid; the the appropriate EPA guidance 53 64 for the
b. Prognostic data. For applications in the wind and concentration estimates for these recommended procedures.
marine environment that require the use of hours should be disregarded and considered b. Air quality model estimates, especially
AERMOD and prognostic data, the prognostic to be missing. Model predicted with the support of measured air quality
data should be processed via MMIF for input concentrations for 3-, 8-, and 24-hour data, are the preferred basis for air quality
to AERMET following recommendations in averages should be calculated by dividing the demonstrations. A number of actions have
paragraph 8.4.5.1(b) and the guidance found sum of the hourly concentrations for the been taken to ensure that the best air quality
in reference 109. period by the number of valid or non-missing model is used correctly for each regulatory
8.4.7 Treatment of Near-Calms and Calms hours. If the total number of valid hours is application and that it is not arbitrarily
less than 18 for 24-hour averages, less than imposed.
8.4.7.1 Discussion 6 for 8-hour averages, or less than 3 for 3- • First, the Guideline clearly recommends
a. Treatment of calm or light and variable hour averages, the total concentration should that the most appropriate model be used in
wind poses a special problem in modeling be divided by 18 for the 24-hour average, 6 each case. Preferred models are identified,
applications since steady-state Gaussian for the 8-hour average, and 3 for the 3-hour based on a number of factors, for many uses.
• Second, the preferred models have been compliance with NAAQS and PSD dispersion modeling address different issues
subjected to a systematic performance increments. and needs relative to each aspect of the
evaluation and a scientific peer review. overall air quality assessment.
Statistical performance measures, including 9.2 Recommendations c. The PSD increments for criteria
measures of difference (or residuals) such as 9.2.1 Modeling Protocol pollutants are listed in 40 CFR 52.21(c) and
bias, variance of difference and gross a. Every effort should be made by the 40 CFR 51.166(c). For short-term increments,
variability of the difference, and measures of appropriate reviewing authority (paragraph these maximum allowable increases in
correlation such as time, space, and time and 3.0(b)) to meet with all parties involved in pollutant concentrations may be exceeded
space combined, as described in section either a SIP submission or revision or a PSD once per year at each site, while the annual
2.1.1, were generally followed. permit application prior to the start of any increment may not be exceeded. The highest,
• Third, more specific information has work on such a project. During this meeting, second-highest increase in estimated
been provided for considering the a protocol should be established between the concentrations for the short-term averages, as
incorporation of new models into the preparing and reviewing parties to define the determined by a model, must be less than or
Guideline (section 3.1), and the Guideline procedures to be followed, the data to be equal to the permitted increment. The
contains procedures for justifying the case- collected, the model to be used, and the modeled annual averages must not exceed
by-case use of alternative models and analysis of the source and concentration data the increment.
obtaining EPA approval (section 3.2). to be performed. An example of the content d. Receptor sites for refined dispersion
c. Air quality modeling is the preferred for such an effort is contained in the Air modeling should be located within the
basis for air quality demonstrations. Quality Analysis Checklist posted on the modeling domain (section 8.1). In designing
Nevertheless, there are rare circumstances EPA’s SCRAM website (section 2.3). This a receptor network, the emphasis should be
where the performance of the preferred air checklist suggests the appropriate level of placed on receptor density and location, not
quality model may be shown to be less than detail to assess the air quality resulting from total number of receptors. Typically, the
reasonably acceptable or where no preferred the proposed action. Special cases may density of receptor sites should be
air quality model, screening model or progressively more resolved near the new or
require additional data collection or analysis
technique, or alternative model are suitable modifying source, areas of interest, and areas
and this should be determined and agreed
for the situation. In these unique instances, with the highest concentrations with
upon at the pre-application meeting. The
there is the possibility of assuring sufficient detail to determine where possible
protocol should be written and agreed upon
compliance and establishing emissions limits violations of a NAAQS or PSD increments are
by the parties concerned, although it is not
for an existing source solely on the basis of most likely to occur. The placement of
intended that this protocol be a binding,
observed air quality data in lieu of an air receptor sites should be determined on a
formal legal document. Changes in such a
quality modeling analysis. Comprehensive case-by-case basis, taking into consideration
protocol or deviations from the protocol are
air quality monitoring in the vicinity of the the source characteristics, topography,
often necessary as the data collection and
existing source with proposed modifications climatology, and monitor sites. Locations of
analysis progresses. However, the protocol particular importance include: (1) the area of
will be necessary in these cases. The same establishes a common understanding of how
attention should be given to the detailed maximum impact of the point source; (2) the
the demonstration required to meet area of maximum impact of nearby sources;
analyses of the air quality data as would be regulatory requirements will be made.
applied to a model performance evaluation. and (3) the area where all sources combine
d. The current levels and forms of the 9.2.2 Design Concentration and Receptor to cause maximum impact. Depending on the
NAAQS for the six criteria pollutants can be Sites complexities of the source and the
found on the EPA’s NAAQS website at a. Under the PSD permitting program, an environment to which the source is located,
https://www.epa.gov/criteria-air-pollutants. air quality analysis for criteria pollutants is a dense array of receptors may be required in
As required by the CAA, the NAAQS are required to demonstrate that emissions from some cases. In order to avoid unreasonably
subjected to extensive review every 5 years the construction or operation of a proposed large computer runs due to an excessively
and the standards, including the level and new source or modification will not cause or large array of receptors, it is often desirable
the form, may be revised as part of that contribute to a violation of the NAAQS or to model the area twice. The first model run
review. The criteria pollutants have either PSD increments. would use a moderate number of receptors
long-term (annual or quarterly) and/or short- i. For a NAAQS assessment, the design more resolved near the new or modifying
term (24-hour or less) forms that are not to concentration is the combination of the source and over areas of interest. The second
be exceeded more than a certain frequency appropriate background concentration model run would modify the receptor
over a period of time (e.g., no exceedance on (section 8.3) with the estimated modeled network from the first model run with a
a rolling 3-month average, no more than once impact of the proposed source. The NAAQS denser array of receptors in areas showing
design concentration is then compared to the potential for high concentrations and
per year, or no more than once per year
possible violations, as indicated by the
averaged over 3 years), are averaged over a applicable NAAQS.
results of the first model run. Accordingly,
period of time (e.g., an annual mean or an ii. For a PSD increment assessment, the
the EPA neither anticipates nor encourages
annual mean averaged over 3 years), or are design concentration includes impacts
that numerous iterations of modeling runs be
some percentile that is averaged over a occurring after the appropriate baseline date
made to continually refine the receptor
period of time (e.g., annual 99th or 98th from all increment-consuming and
network.
percentile averaged over 3 years). The 3-year increment-expanding sources. The PSD
period for ambient monitoring design values increment design concentration is then 9.2.3 NAAQS and PSD Increments
does not dictate the length of the data periods compared to the applicable PSD increment. Compliance Demonstrations for New or
recommended for modeling (i.e., 5 years of b. The specific form of the NAAQS for the Modifying Sources
NWS meteorological data, at least 1 year of pollutant(s) of concern will also influence a. As described in this subsection, the
site-specific, or at least 3 years of prognostic how the background and modeled data recommended procedure for conducting
meteorological data). should be combined for appropriate either a NAAQS or PSD increments
e. This section discusses general comparison with the respective NAAQS in assessment under PSD permitting is a multi-
recommendations on the regulatory such a modeling demonstration. Given the stage approach that includes the following
application of models for the purposes of potential for revision of the form of the two stages:
NSR, including PSD permitting, and NAAQS and the complexities of combining i. The EPA describes the first stage as a
particularly for estimating design background and modeled data, specific single-source impact analysis, since this stage
concentration(s), appropriately comparing details on this process can be found in the involves considering only the impact of the
these estimates to NAAQS and PSD applicable modeling guidance available on new or modifying source. There are two
increments, and developing emissions limits. the EPA’s SCRAM website (section 2.3). possible levels of detail in conducting a
This section also provides the criteria Modeled concentrations should not be single-source impact analysis with the model
necessary for considering use of an analysis rounded before comparing the resulting user beginning with use of a screening model
based on measured ambient data in lieu of design concentration to the NAAQS or PSD and proceeding to use of a refined model as
modeling as the sole basis for demonstrating increments. Ambient monitoring and necessary.
ii. The EPA describes the second stage as to a NAAQS or PSD increment violation. justifiable alternative model suitable for the
a cumulative impact analysis, since it takes This determination should be based on: (1) situation. In these unique instances, there
into account all sources affecting the air The appropriate design concentration for may be the possibility of establishing
quality in an area. In addition to the project each applicable NAAQS (and averaging emissions limits and demonstrating
source impact, this stage includes period); and (2) whether the source’s compliance with the NAAQS and PSD
consideration of background, which includes emissions cause or contribute to a violation increments solely on the basis of analysis of
contributions from nearby sources and other at the time and location of any modeled observed air quality data in lieu of an air
sources (e.g., natural, minor, distant major, violation (i.e., when and where the predicted quality modeling analysis. However, only in
and unidentified sources). design concentration is greater than the the case of a modification to an existing
b. Each stage should involve increasing NAAQS). For PSD increments, the source should air quality monitoring data
complexity and details, as required, to fully cumulative impact analysis should also alone be a basis for determining adequate
demonstrate that a new or modifying source consider the amount of the air quality emissions limits or for demonstration that the
will not cause or contribute to a violation of increment that has already been consumed modification will not cause or contribute to
any NAAQS or PSD increment. As such, by other sources, or, conversely, whether a violation of any NAAQS or PSD increment.
starting with a single-source impact analysis increment has expanded relative to the c. The following items should be
is recommended because, where the analysis baseline concentration. Therefore, the considered prior to the acceptance of an
at this stage is sufficient to demonstrate that applicant should model the existing or analysis of measured air quality data as the
a source will not cause or contribute to any permitted nearby increment-consuming and sole basis for an air quality demonstration or
potential violation, this may alleviate the increment-expanding sources, rather than determining an emissions limit:
need for a more time-consuming and using past modeling analyses of those i. Does a monitoring network exist for the
comprehensive cumulative modeling sources as part of background concentration. pollutants and averaging times of concern in
analysis. This would permit the use of newly acquired the vicinity of the existing source?
c. The single-source impact analysis, or data or improved modeling techniques if ii. Has the monitoring network been
first stage of an air quality analysis, should such data and/or techniques have become designed to locate points of maximum
begin by determining the potential of a available since the last source was permitted. concentration?
proposed new or modifying source to cause 9.2.3.1 Considerations in Developing iii. Do the monitoring network and the data
or contribute to a NAAQS or PSD increment Emissions Limits reduction and storage procedures meet EPA
violation. In certain circumstances, a monitoring and quality assurance
screening model or technique may be used a. Emissions limits and resulting control requirements?
instead of the preferred model because it will requirements should be established to iv. Do the dataset and the analysis allow
provide estimated worst-case ambient provide for compliance with each applicable impact of the most important individual
NAAQS (and averaging period) and PSD sources to be identified if more than one
impacts from the proposed new or modifying
increment. It is possible that multiple source or emission point is involved?
source. If these worst case ambient
emissions limits will be required for a source v. Is at least one full year of valid ambient
concentration estimates indicate that the
to demonstrate compliance with several
source will not cause or contribute to any data available?
criteria pollutants (and averaging periods)
potential violation of a NAAQS or PSD vi. Can it be demonstrated through the
and PSD increments. Case-by-case
increment, then the screening analysis comparison of monitored data with model
determinations must be made as to the
should generally be sufficient for the results that available air quality models and
appropriate form of the limits, i.e., whether
required demonstration under PSD. If the techniques are not applicable?
the emissions limits restrict the emission
ambient concentration estimates indicate that d. Comprehensive air quality monitoring in
factor (e.g., limiting lb/MMBTU), the
the source’s emissions have the potential to the area affected by the existing source with
emission rate (e.g., lb/hr), or both. The
cause or contribute to a violation, then the proposed modifications will be necessary in
appropriate reviewing authority (paragraph
use of a refined model to estimate the these cases. Additional meteorological
3.0(b)) and appropriate EPA guidance should
source’s impact should be pursued. The be consulted to determine the appropriate monitoring may also be necessary. The
refined modeling analysis should use a emissions limits on a case-by-case basis. appropriate number of air quality and
model or technique consistent with the meteorological monitors from a scientific and
Guideline (either a preferred model or 9.2.4 Use of Measured Data in Lieu of technical standpoint is a function of the
technique or an alternative model or Model Estimates situation being considered. The source
technique) and follow the requirements and a. As described throughout the Guideline, configuration, terrain configuration, and
recommendations for model inputs outlined modeling is the preferred method for meteorological variations all have an impact
in section 8. If the ambient concentration demonstrating compliance with the NAAQS on number and optimal placement of
increase predicted with refined modeling and PSD increments and for determining the monitors. Decisions on the monitoring
indicates that the source will not cause or most appropriate emissions limits for new network appropriate for this type of analysis
contribute to any potential violation of a and existing sources. When a preferred can only be made on a case-by-case basis.
NAAQS or PSD increment, then the refined model or adequately justified and approved e. Sources should obtain approval from the
analysis should generally be sufficient for the alternative model is available, model results, appropriate reviewing authority (paragraph
required demonstration under PSD. However, including the appropriate background, are 3.0(b)) and the EPA Regional Office for the
if the ambient concentration estimates from sufficient for air quality demonstrations and monitoring network prior to the start of
the refined modeling analysis indicate that establishing emissions limits, if necessary. In monitoring. A monitoring protocol agreed to
the source’s emissions have the potential to instances when the modeling technique by all parties involved is necessary to assure
cause or contribute to a violation, then a available is only a screening technique, the that ambient data are collected in a
cumulative impact analysis should be addition of air quality monitoring data to the consistent and appropriate manner. The
undertaken. The receptors that indicate the analysis may lend credence to the model design of the network, the number, type, and
location of significant ambient impacts results. However, air quality monitoring data location of the monitors, the sampling
should be used to define the modeling alone will normally not be acceptable as the period, averaging time, as well as the need
domain for use in the cumulative impact sole basis for demonstrating compliance with for meteorological monitoring or the use of
analysis (section 8.2.2). the NAAQS and PSD increments or for mobile sampling or plume tracking
d. The cumulative impact analysis, or the determining emissions limits. techniques, should all be specified in the
second stage of an air quality analysis, b. There may be rare circumstances where protocol and agreed upon prior to start-up of
should be conducted with the same refined the performance of the preferred air quality the network.
model or technique to characterize the model will be shown to be less than f. Given the uniqueness and complexities
project source and then include the reasonably acceptable when compared with of these rare circumstances, the procedures
appropriate background concentrations air quality monitoring data measured in the can only be established on a case-by-case
(section 8.3). The resulting design vicinity of an existing source. Additionally, basis for analyzing the source’s emissions
concentrations should be used to determine there may not be an applicable preferred air data and the measured air quality monitoring
whether the source will cause or contribute quality model, screening technique, or data, and for projecting with a reasoned basis
the air quality impact of a proposed 12. Zhou, W., Cohan, D.S., Pinder, R.W., 25. Hanna, S., M. Garrison and B. Turner,
modification to an existing source in order to Neuman, J.A., Holloway, J.S., Peischl, J., 1998. AERMOD Peer Review report.
demonstrate that emissions from the Ryerson, T.B., Nowak, J.B., Flocke, F., Prepared by SAI, Inc. under EPA
construction or operation of the modification Zheng, W.G., 2012. Observation and Contract No. 68–D6–0064/1–14 for the
will not cause or contribute to a violation of modeling of the evolution of Texas U.S. Environmental Protection Agency,
the applicable NAAQS and PSD increment, power plant plumes. Atmospheric Research Triangle Park, NC. 12pp. &
and to determine adequate emissions limits. Chemistry and Physics, 12: 455–468. appendices. (Docket No. A–99–05, II–A–
The same attention should be given to the 13. Chen, J., Lu, J., Avise, J.C., DaMassa, J.A., 6).
detailed analyses of the air quality data as Kleeman, M.J., Kaduwela, A.P., 2014. 26. Scire, J.S. and L.L. Schulman, 1981.
would be applied to a comprehensive model Seasonal modeling of PM 2.5 in Evaluation of the BLP and ISC Models
performance evaluation. In some cases, the California’s San Joaquin Valley. with SF6 Tracer Data and SO2
monitoring data collected for use in the Atmospheric Environment, 92: 182–190. Measurements at Aluminum Reduction
performance evaluation of preferred air 14. Russell, A.G., 2008. EPA Supersites Plants. APCA Specialty Conference on
quality models, screening technique, or program-related emissions-based Dispersion Modeling for Complex
existing alternative models may help inform particulate matter modeling: initial Sources, St. Louis, MO.
the development of a suitable new alternative applications and advances. Journal of the 27. U.S. Environmental Protection Agency,
model. Early coordination with the Air & Waste Management Association, 2003. AERMOD: Latest Features and
appropriate reviewing authority (paragraph 58: 289–302. Evaluation Results. Publication No.
3.0(b)) and the EPA Regional Office is 15. Tesche, T., Morris, R., Tonnesen, G., EPA–454/R–03–003. Office of Air
fundamental with respect to any potential McNally, D., Boylan, J., Brewer, P., 2006. Quality Planning and Standards,
use of measured data in lieu of model CMAQ/CAMx annual 2002 performance Research Triangle Park, NC.
estimates. evaluation over the eastern US. 28. ASTM D6589: Standard Guide for
Atmospheric Environment, 40: 4906– Statistical Evaluation of Atmospheric
10.0 References 4919. Dispersion Model Performance. (2010).
1. Code of Federal Regulations; Title 40 16. Fox, D.G., 1984. Uncertainty in air quality 29. U.S. Environmental Protection Agency,
(Protection of Environment); part 51; modeling. Bulletin of the American 1992. Protocol for Determining the Best
§§ 51.112, 51.117, 51.150, 51.160. Meteorological Society, 65(1): 27–36. Performing Model. Publication No. EPA–
2. U.S. Environmental Protection Agency, 17. Bowne, NE, 1981. Validation and 454/R–92–025. Office of Air Quality
1990. New Source Review Workshop Performance Criteria for Air Quality Planning and Standards, Research
Manual: Prevention of Significant Models. Appendix F in Air Quality Triangle Park, NC. (NTIS No. PB 93–
Deterioration and Nonattainment Area Modeling and the Clean Air Act: 226082).
Recommendations to EPA on Dispersion 30. Hanna, S.R., 1982. Natural Variability of
Permitting (Draft). Office of Air Quality
Modeling for Regulatory Applications. Observed Hourly SO2 and CO
Planning and Standards, Research
American Meteorological Society, Concentrations in St. Louis. Atmospheric
Triangle Park, NC.
Boston, MA; pp. 159–171. (Docket No. Environment, 16(6): 1435–1440.
3. Code of Federal Regulations; Title 40
A–80–46, II–A–106). 31. Pasquill, F., 1974. Atmospheric Diffusion,
(Protection of Environment); part 51;
18. Fox, D.G., 1981. Judging Air Quality 2nd Edition. John Wiley and Sons, New
§§ 51.166 and 52.21. Model Performance. Bulletin of the York, NY; 479pp.
4. Code of Federal Regulations; Title 40 American Meteorological Society, 62(5): 32. Rhoads, R.G., 1981. Accuracy of Air
(Protection of Environment); part 93; 599–609. Quality Models. Staff Report. U.S.
§§ 93.116, 93.123, and 93.150. 19. Simon, H., Baker, K.R., Phillips, S., 2012. Environmental Protection Agency,
5. Code of Federal Regulations; Title 40 Compilation and interpretation of Research Triangle Park, NC. (Docket No.
(Protection of Environment); part 58 photochemical model performance A–80–46, II–G–6).
(Ambient Air Quality Surveillance). statistics published between 2006 and 33. Hanna, S.R., 1993. Uncertainties in air
6. Code of Federal Regulations; Title 40 2012. Atmospheric Environment, 61: quality model predictions. Boundary-
(Protection of Environment); part 50 124–139. Layer Meteorology, 62: 3–20.
(National Primary and Secondary 20. Burton, C.S., 1981. The Role of 34. Hanna, S.R., 1989. Confidence limits for
Ambient Air Quality Standards). Atmospheric Models in Regulatory air quality model evaluations, as
7. Baker, K.R., Kelly, J.T., 2014. Single source Decision-Making: Summary Report. estimated by bootstrap and jackknife
impacts estimated with photochemical Systems Applications, Inc., San Rafael, resampling methods. Atmospheric
model source sensitivity and CA. Prepared under contract No. 68–01– Environment, 23(6): 1385–1398.
apportionment approaches. Atmospheric 5845 for the U.S. Environmental 35. Cox, W.M. and J.A. Tikvart, 1990. A
Environment, 96: 266–274. Protection Agency, Research Triangle statistical procedure for determining the
8. ENVIRON, 2012. Evaluation of Chemical Park, NC. (Docket No. A–80–46, II–M–6). best performing air quality simulation
Dispersion Models using Atmospheric 21. Olesen, H.R., 2001. Ten years of model. Atmospheric Environment,
Plume Measurements from Field Harmonisation activities: Past, present 24A(9): 2387–2395.
Experiments. ENVIRON International, and future. Introductory address and 36. U.S. Environmental Protection Agency,
Corp., Novato, CA. Prepared under paper presented at the 7th International 2016. Technical Support Document
contract No. EP–D–07–102 for the U.S. Conference on Harmonisation within (TSD) for AERMOD-Based Assessments
Environmental Protection Agency, Atmospheric Dispersion Modelling for of Long-Range Transport Impacts for
Research Triangle Park, NC. Regulatory Purposes, May 28–31, 2001, Primary Pollutants. Publication No.
9. McMurry, P.H., Shepherd, M.F., Vickery, Belgirate, Italy. EPA–454/B–16–007. Office of Air
J.S., 2004. Particulate matter science for 22. Weil, Sykes, and Venkatram, 1992. Quality Planning and Standards,
policy makers: A NARSTO assessment. Evaluating Air-Quality Models: Review Research Triangle Park, NC.
Cambridge University Press. and Outlook. Journal of Applied 37. U.S. Environmental Protection Agency,
10. Baker, K.R., Foley, K.M., 2011. A Meteorology, 31: 1121–1145. 2021. AERSCREEN User’s Guide.
nonlinear regression model estimating 23. U.S. Environmental Protection Agency, Publication No. EPA–454/B–21–005.
single source concentrations of primary 2016. Model Clearinghouse: Operational Office of Air Quality Planning and
and secondarily formed PM2.5. Plan. Publication No. EPA–454/B–16– Standards, Research Triangle Park, NC.
Atmospheric Environment, 45: 3758– 008. Office of Air Quality Planning and 38. U.S. Environmental Protection Agency,
3767. Standards, Research Triangle Park, NC. 2011. AERSCREEN Released as the EPA
11. Bergin, M.S., Russell, A.G., Odman, M.T., 24. American Meteorological Society, 1983. Recommended Screening Model.
Cohan, D.S., Chameldes, W.L., 2008. Synthesis of the Rural Model Reviews. Memorandum dated April 11, 2011,
Single-Source Impact Analysis Using Publication No. EPA–600/3–83–108. Office of Air Quality Planning and
Three-Dimensional Air Quality Models. Office of Research and Development, Standards, Research Triangle Park, NC.
Journal of the Air & Waste Management Research Triangle Park, NC. (NTIS No. 39. Perry, S.G., D.J. Burns and A.J. Cimorelli,
Association, 58: 1351–1359. PB 84–121037). 1990. User’s Guide to CTDMPLUS:
Volume 2. The Screening Mode 51. LEADPOST processor: https:// Goals for Ozone, PM2.5, and Regional
(CTSCREEN). Publication No. EPA–600/ gaftp.epa.gov/Air/aqmg/SCRAM/models/ Haze. Publication No. EPA 454/R–18–
8–90–087. U.S. Environmental preferred/aermod/leadpost.zip. 009. Office of Air Quality Planning and
Protection Agency, Research Triangle 52. U.S. Environmental Protection Agency, Standards, Research Triangle Park, NC.
Park, NC. (NTIS No. PB 91–136564). 1993. Lead Guideline Document. 65. U.S. Environmental Protection Agency,
40. U.S. Environmental Protection Agency, Publication No. EPA–452/R–93–009. 2015. Transportation Conformity
1992. Screening Procedures for Office of Air Quality Planning and Guidance for Quantitative Hot-Spot
Estimating the Air Quality Impact of Standards, Research Triangle Park, NC. Analyses in PM2.5 and PM10
Stationary Sources, Revised. Publication (NTIS No. PB 94–111846). Nonattainment and Maintenance Areas.
No. EPA–454/R–92–019. Office of Air 53. U.S. Environmental Protection Agency, Publication No. EPA–420–B–15–084.
Quality Planning and Standards, 2014. Guidance for 1-Hour SO2 Office of Transportation and Air Quality,
Research Triangle Park, NC. (NTIS No. Nonattainment Area SIP Submissions. Ann Arbor, MI.
PB 93–219095). Office of Air Quality Planning and 66. U.S. Environmental Protection Agency,
41. Burns, D.J., S.G. Perry and A.J. Cimorelli, Standards, Research Triangle Park, NC. 1987. PM10 SIP Development Guideline.
1991. An Advanced Screening Model for 54. U.S. Environmental Protection Agency, Publication No. EPA–450/2–86–001.
Complex Terrain Applications. Paper 2016. SO2 NAAQS Designations Office of Air Quality Planning and
presented at the 7th Joint Conference on Modeling Technical Assistance Standards, Research Triangle Park, NC.
Applications of Air Pollution Document. Office of Air Quality (NTIS No. PB 87–206488).
Meteorology (cosponsored by the Planning and Standards, Research 67. U.S. Environmental Protection Agency,
American Meteorological Society and the Triangle Park, NC. 2012. Haul Road Workgroup Final
Air & Waste Management Association), 55. Turner, D.B., 1964. A Diffusion Model for Report Submission to EPA–OAQPS.
January 13–18, 1991, New Orleans, LA. an Urban Area. Journal of Applied Memorandum dated March 2, 2012,
42. Mills, M.T., R.J. Paine, E.A. Insley and Meteorology, 3(1):83–91. Office of Air Quality Planning and
B.A. Egan, 1987. The Complex Terrain 56. U.S. Environmental Protection Agency, Standards, Research Triangle Park, NC.
Dispersion Model Terrain Preprocessor 2015. Technical Support Document 68. Seinfeld, J.H., Pandis, S.N., 2012.
System—User’s Guide and Program (TSD) for NO2-Related AERMOD Options Atmospheric chemistry and physics:
Description. Publication No. EPA–600/ and Modifications. Publication No. EPA– from air pollution to climate change.
8–88–003. U.S. Environmental 454/B–15–004. Office of Air Quality John Wiley & Sons.
Protection Agency, Research Triangle Planning and Standards, Research 69. Simon, H., Baker, K.R., Phillips, S., 2012.
Park, NC. (NTIS No. PB 88–162094). Triangle Park, NC. Compilation and interpretation of
43. Environmental Research and Technology, 57. Podrez, M., 2015. An Update to the
photochemical model performance
1987. User’s Guide to the Rough Terrain statistics published between 2006 and
Ambient Ratio Method for 1-h NO2 Air
Diffusion Model (RTDM), Rev. 3.20. ERT 2012. Atmospheric Environment, 61,
Quality Standards Dispersion Modeling.
Document No. P–D535–585. 124–139.
Atmospheric Environment, 103: 163–
Environmental Research and 70. U.S. Environmental Protection Agency,
170.
Technology, Inc., Concord, MA. (NTIS 2016. Guidance on the use of models for
No. PB 88–171467). 58. Cole, H.S. and J.E. Summerhays, 1979. A assessing the impacts of emissions from
44. U.S. Environmental Protection Agency, Review of Techniques Available for single sources on the secondarily formed
2023. AERMOD Model Formulation. Estimation of Short-Term NO2 pollutants ozone and PM2.5. Publication
Publication No. EPA–454/B–23–008. Concentrations. Journal of the Air No. EPA 454/R–16–005. Office of Air
Office of Air Quality Planning and Pollution Control Association, 29(8): Quality Planning and Standards,
Standards, Research Triangle Park, NC. 812–817. Research Triangle Park, NC.
45. Cimorelli, A., et al., 2005. AERMOD: A 59. Hanrahan, P.L., 1999. The Polar Volume 71. U.S. Environmental Protection Agency,
Dispersion Model for Industrial Source Polar Ratio Method for Determining 2019. Guidance on the Development of
Applications. Part I: General Model NO2/NOX Ratios in Modeling—Part I: Modeled Emission Rates for Precursors
Formulation and Boundary Layer Methodology. Journal of the Air & Waste (MERPs) as a Tier 1 Demonstration Tool
Characterization. Journal of Applied Management Association, 49: 1324– for Ozone and PM2.5 under the PSD
Meteorology, 44(5): 682–693. 1331. Permitting Program. Publication No.
46. L.L. Schulman, D.G. Strimaitis and J.S. 60. Carruthers, D.J.; Stocker, J.R.; Ellis, A.; EPA–454/R–19–003. Office of Air
Scire, 2002. Development and evaluation Seaton, M.D.; Smith, SE, 2017. Quality Planning and Standards,
of the PRIME plume rise and building Evaluation of an explicit NOX chemistry Research Triangle Park, NC.
downwash model. Journal of the Air & method in AERMOD; Journal of the Air 72. U.S. Department of the Interior, 2010.
Waste Management Association, 50: & Waste Management Association. 2017, Federal Land Managers’ Air Quality
378–390. 67 (6), 702–712; DOI:10.1080/ Related Values Work Group (FLAG)
47. U.S. EPA, 2012: User’s Manual 10962247.2017.1280096. Phase I Report—Revised 2010. Natural
AERCOARE Version 1.0. EPA–910–R– 61. Environmental Protection Agency, 2023. Resource Report NPS/NPRC/NRR–2010/
12–008. U.S. EPA, Region 10, Seattle, Technical Support Document (TSD) for 232.
WA. Adoption of the Generic Reaction Set 73. National Acid Precipitation Assessment
48. Fairall, C.W., E.F. Bradley, J.E. Hare, A.A. Method (GRSM) as a Regulatory Non- Program (NAPAP), 1991. Acid
Grachev, and J.B. Edson, 2003: ‘‘Bulk Default Tier-3 NO2 Screening Option, Deposition: State of Science and
Parameterization of Air-Sea Fluxes: Publication No. EPA–454/R–23–009. Technology. Volume III Terrestrial,
Updates and Verification for the COARE Office of Air Quality Planning & Materials, Health and Visibility Effects.
Algorithm.’’ Journal. Of Climate, 16, Standards, Research Triangle Park, NC. Report 24, Visibility: Existing and
571–591. 62. U.S. Environmental Protection Agency, Historical Conditions—Causes and
49. U.S. Environmental Protection Agency, 2004. The Particle Pollution Report. Effects. Edited by Patricia M. Irving.
1992. Guideline for modeling carbon Publication No. EPA–454/R–04–002. Washington, DC, 129pp.
monoxide from roadway intersections. Office of Air Quality Planning and 74. National Research Council, 1993.
Publication number EPA–454/R–92–005. Standards, Research Triangle Park, NC. Protecting Visibility in National Parks
Office of Air Quality Planning & 63. U.S. Environmental Protection Agency, and Wilderness Areas. National
Standards, Research Triangle Park, NC. 2022. Guidance for Ozone and PM2.5 Academy Press, Washington, DC, 446pp.
50. U.S. Environmental Protection Agency, Permit Modeling. Publication No. EPA– 75. U.S. Environmental Protection Agency,
1997. Guidance for Siting Ambient Air 454/R–22–005. Office of Air Quality 1992. Workbook for plume visual impact
Monitors around Stationary Lead Planning and Standards, Research screening and analysis (revised).
Sources. Publication No. EPA–454/R– Triangle Park, NC. Publication No. EPA–454/R–92–023.
92–009R. Office of Air Quality Planning 64. U.S. Environmental Protection Agency, Office of Air Quality Planning and
and Standards, Research Triangle Park, 2018. Modeling Guidance for Standards, Research Triangle Park, NC.
NC. (NTIS No. PB 97–208094). Demonstrating Attainment of Air Quality (NTIS No. PB 93–223592).
76. Nilsson, J., Grennfelt, P., Ministerråd, N., Good Engineering Practice Stack Height Air Quality Planning and Standards,
1988. Critical Loads for Sulphur and (Technical Support Document for the Research Triangle Park, NC.
Nitrogen: Report from a Workshop Held Stack Height Regulations), Revised. 99. U.S. Environmental Protection Agency,
at Skokloster, Sweden, 19–24 March, Publication No. EPA–450/4–80–023R. 2013. Use of ASOS meteorological data
1988. Nordic Council of Ministers. Office of Air Quality Planning and in AERMOD dispersion modeling.
77. Turner, D.B., 1969. Workbook of Standards, Research Triangle Park, NC. Memorandum dated March 8, 2013,
Atmospheric Dispersion Estimates. PHS (NTIS No. PB 85–225241). Office of Air Quality Planning and
Publication No. 999–AP–26. U.S. 89. Snyder, W.H. and R.E. Lawson, Jr., 1985. Standards, Research Triangle Park, NC.
Department of Health, Education and Fluid Modeling Demonstration of Good 100. U.S. Environmental Protection Agency,
Welfare, Public Health Service, Engineering-Practice Stack Height in 2023. User’s Guide for the AERMOD
Cincinnati, OH. (NTIS No. PB–191482). Complex Terrain. Publication No. EPA– Meteorological Preprocessor (AERMET).
78. McElroy, J.L. and F. Pooler, Jr., 1968. St. 600/3–85–022. Office of Research and Publication No. EPA–454/B–23–005.
Louis Dispersion Study, Volume II— Development, Research Triangle Park, Office of Air Quality Planning and
Analysis. National Air Pollution Control NC. (NTIS No. PB 85–203107). Standards, Research Triangle Park, NC.
Administration Publication No. AP–53, 90. Briggs, G.A., 1975. Plume Rise 101. U.S Environmental Protection Agency.
U.S. Department of Health, Education Predictions. Chapter 3 in Lectures on Air 2023. AERMINUTE User’s Guide.
and Welfare, Public Health Service, Pollution and Environmental Impact Publication No. EPA–454/B–23–007.
Arlington, VA. (NTIS No. PB–190255). Analyses. American Meteorological Office of Air Quality Planning and
79. Irwin, J.S., 1978. Proposed Criteria for Society, Boston, MA; pp. 59–111. Standards, Research Triangle Park, NC.
Selection of Urban Versus Rural 91. Hanna, S.R., G.A. Briggs and R.P. Hosker, 102. U.S. Environmental Protection Agency,
Dispersion Coefficients. (Draft Staff Jr., 1982. Plume Rise. Chapter 2 in 1993. PCRAMMET User’s Guide.
Report). Meteorology and Assessment Handbook on Atmospheric Diffusion. Publication No. EPA–454/R–96–001.
Division, U.S. Environmental Protection Technical Information Center, U.S. Office of Air Quality Planning and
Agency, Research Triangle Park, NC. Department of Energy, Washington, DC; Standards, Research Triangle Park, NC.
(Docket No. A–80–46, II–B–8). pp. 11–24. DOE/TIC–11223 (DE (NTIS No. PB 97–147912).
80. Auer, Jr., A.H., 1978. Correlation of Land 82002045). 103. U.S. Environmental Protection Agency,
Use and Cover with Meteorological 92. Weil, J.C., L.A. Corio and R.P. Brower, 1996. Meteorological Processor for
Anomalies. Journal of Applied 1997. A PDF dispersion model for Regulatory Models (MPRM). Publication
Meteorology, 17(5): 636–643. buoyant plumes in the convective No. EPA–454/R–96–002. Office of Air
81. U.S. Environmental Protection Agency, boundary layer. Journal of Applied Quality Planning and Standards,
2023. AERMOD Implementation Guide. Meteorology, 36: 982–1003. Research Triangle Park, NC. (NTIS No.
Publication No. EPA–454/B–23–009. 93. L.L. Schulman, D.G. Strimaitis and J.S. PB 96–180518).
Office of Air Quality Planning and
Scire, 2002. Development and evaluation 104. Paine, R.J., 1987. User’s Guide to the
Standards, Research Triangle Park, NC.
of the PRIME plume rise and building CTDM Meteorological Preprocessor
82. Pasquill, F., 1976. Atmospheric
downwash model. Journal of the Air & Program. Publication No. EPA–600/8–
Dispersion Parameters in Gaussian
Waste Management Association, 50: 88–004. Office of Research and
Plume Modeling, Part II. Possible
378–390. Development, Research Triangle Park,
Requirements for Change in the Turner
94. U.S. Environmental Protection Agency, NC. (NTIS No. PB–88–162102).
Workbook Values. Publication No. EPA–
600/4–76–030b. Office of Research and 1995. Compilation of Air Pollutant 105. Perry, S.G., D.J. Burns, L.H. Adams, R.J.
Development, Research Triangle Park, Emission Factors, Volume I: Stationary Paine, M.G. Dennis, M.T. Mills, D.G.
NC. (NTIS No. PB–258036/3BA). Point and Area Sources (Fifth Edition, Strimaitis, R.J. Yamartino and E.M.
83. Stull, R.B., 1988. An Introduction to AP–42: GPO Stock No. 055–000–00500– Insley, 1989. User’s Guide to the
Boundary Layer Meteorology. Kluwer 1), and Supplements A–D. Volume I can Complex Terrain Dispersion Model Plus
Academic Publishers, Boston, MA. be downloaded from EPA’s website at Algorithms for Unstable Situations
666pp. https://www.epa.gov/air-emissions- (CTDMPLUS). Volume 1: Model
84. U.S. Environmental Protection Agency, factors-and-quantification/ap-42- Descriptions and User Instructions.
1987. Analysis and Evaluation of compilation-air-emission-factors. Publication No. EPA–600/8–89–041. U.S.
Statistical Coastal Fumigation Models. 95. U.S. Environmental Protection Agency, Environmental Protection Agency,
Publication No. EPA–450/4–87–002. 2017. Emissions Inventory Guidance for Research Triangle Park, NC. (NTIS No.
Office of Air Quality Planning and Implementation of Ozone and Particulate PB 89–181424).
Standards, Research Triangle Park, NC. Matter National Ambient Air Quality 106. U.S. Environmental Protection Agency,
(NTIS No. PB 87–175519). Standards (NAAQS) and Regional Haze 2020. User’s Guide for AERSURFACE
85. Wesely, M.L, P.V. Doskey, and J.D. Regulations. Publication No. EPA–454/ Tool. Publication No. EPA–454/B–20–
Shannon, 2002: Deposition B–17–002. Office of Air Quality Planning 008. Office of Air Quality Planning and
Parameterizations for the Industrial and Standards, Research Triangle Park, Standards, Research Triangle Park, NC.
Source Complex (ISC3) Model. ANL NC. 107. Brode, R., K. Wesson, J. Thurman, and
Report ANL/ER/TRB01/003, DOE/W–31– 96. U.S. Environmental Protection Agency, C. Tillerson, 2008. AERMOD Sensitivity
109–Eng–38, Argonne National 2023. Draft Guidance on Developing to the Choice of Surface Characteristics.
Laboratory, Argonne, Illinois 60439. Background Concentrations for Use in Paper #811 presented at the 101st Air &
86. U.S. Environmental Protection Agency, Modeling Demonstrations. Publication Waste Management Association Annual
1981. Guideline for Use of Fluid No. EPA–454/P–23–001. Office of Air Conference and Exhibition, June 24–27,
Modeling to Determine Good Quality Planning and Standards, 2008, Portland, OR.
Engineering Practice (GEP) Stack Height. Research Triangle Park, NC. 108. Ramboll, 2023. The Mesoscale Model
Publication No. EPA–450/4–81–003. 97. U.S. Environmental Protection Agency, Interface Program (MMIF) Version 4.1
Office of Air Quality Planning and 1987. Ambient Air Monitoring User’s Manual.
Standards, Research Triangle Park, NC. Guidelines for Prevention of Significant 109. U.S. Environmental Protection Agency,
(NTIS No. PB 82–145327). Deterioration (PSD). Publication No. 2023. Guidance on the Use of the
87. Lawson, Jr., R.E. and W.H. Snyder, 1983. EPA–450/4–87–007. Office of Air Mesoscale Model Interface Program
Determination of Good Engineering Quality Planning and Standards, (MMIF) for AERMOD Applications.
Practice Stack Height: A Demonstration Research Triangle Park, NC. (NTIS No. Publication No. EPA–454/B–23–006.
Study for a Power Plant. Publication No. PB 90–168030). Office of Air Quality Planning and
EPA–600/3–83–024. Office of Research 98. U.S. Environmental Protection Agency, Standards, Research Triangle Park, NC.
and Development, Research Triangle 2011. Additional Clarification Regarding 110. U.S. Environmental Protection Agency,
Park, NC. (NTIS No. PB 83–207407). Application of Appendix W Modeling 2000. Meteorological Monitoring
88. U.S. Environmental Protection Agency, Guidance for the 1-hour NO2 National Guidance for Regulatory Modeling
1985. Guideline for Determination of Ambient Air Quality Standard. Office of Applications. Publication No. EPA–454/
R–99–005. Office of Air Quality Planning Addendum A to Appendix W of Part Part D: Transport and Environment, 25:
and Standards, Research Triangle Park, 51—Summaries of Preferred Air pp 93–105.
NC. (NTIS No. PB 2001–103606). Quality Models Heist, D., et al., 2023. Integration of RLINE
111. ASTM D5527: Standard Practice for dispersion model into EPA’s AERMOD:
Measuring Surface Winds and Table of Contents updated formulation and evaluations.
Temperature by Acoustic Means. (2011). A.0 Introduction and Availability Journal of the Air & Waste Management
112. ASTM D5741: Standard Practice for A.1 AERMOD (AMS/EPA Regulatory Association, Manuscript submitted for
Characterizing Surface Wind Using Wind Model) publication.
A.2 CTDMPLUS (Complex Terrain U.S. Environmental Protection Agency, 2023.
Vane and Rotating Anemometer. (2011).
Dispersion Model Plus Algorithms for User’s Guide for the AMS/EPA
113. U.S. Environmental Protection Agency, Regulatory Model (AERMOD).
1995. Quality Assurance for Air Unstable Situations)
A.3 OCD (Offshore and Coastal Dispersion Publication No. EPA–454/B–23–008.
Pollution Measurement Systems, Volume Office of Air Quality Planning and
IV—Meteorological Measurements. Model)
Standards, Research Triangle Park, NC.
Publication No. EPA600/R–94/038d. A.0 Introduction and Availability U.S. Environmental Protection Agency, 2023.
Office of Air Quality Planning and User’s Guide for the AERMOD
(1) This appendix summarizes key features
Standards, Research Triangle Park, NC. Meteorological Preprocessor (AERMET).
of refined air quality models preferred for
Note: for copies of this handbook, you specific regulatory applications. For each Publication No. EPA–454/B–23–005.
may make inquiry to ORD Publications, model, information is provided on Office of Air Quality Planning and
26 West Martin Luther King Dr., availability, approximate cost (where Standards, Research Triangle Park, NC.
Cincinnati, OH 45268. applicable), regulatory use, data input, U.S. Environmental Protection Agency, 2018.
114. Bowen, B.M., J.M. Dewart and A.I. Chen, output format and options, simulation of User’s Guide for the AERMOD Terrain
1983. Stability Class Determination: A atmospheric physics, and accuracy. These Preprocessor (AERMAP). Publication No.
Comparison for One Site. Proceedings, models may be used without a formal EPA–454/B–18–004. U.S. Environmental
Sixth Symposium on Turbulence and demonstration of applicability provided they Protection Agency, Office of Air Quality
Diffusion. American Meteorological satisfy the recommendations for regulatory Planning and Standards, Research
Society, Boston, MA; pp. 211–214. use; not all options in the models are Triangle Park, NC.
(Docket No. A–92–65, II–A–7). necessarily recommended for regulatory use. Schulman, L.L., D.G. Strimaitis and J.S. Scire,
115. U.S. Environmental Protection Agency, (2) These models have been subjected to a 2000. Development and evaluation of the
1993. An Evaluation of a Solar performance evaluation using comparisons PRIME plume rise and building
Radiation/Delta-T (SRDT) Method for with observed air quality data. Where downwash model. Journal of the Air &
Estimating Pasquill-Gifford (P–G) possible, the models contained herein have Waste Management Association, 50:
Stability Categories. Publication No. been subjected to evaluation exercises, 378–390.
including: (1) statistical performance tests Schulman, L.L., and Joseph S. Scire, 1980.
EPA–454/R–93–055. Office of Air
recommended by the American Buoyant Line and Point Source (BLP)
Quality Planning and Standards,
Meteorological Society, and (2) peer Dispersion Model User’s Guide.
Research Triangle Park, NC. (NTIS No.
scientific reviews. The models in this Document P–7304B. Environmental
PB 94–113958). Research and Technology, Inc., Concord,
116. Irwin, J.S., 1980. Dispersion Estimate appendix have been selected on the basis of
the results of the model evaluations, MA. (NTIS No. PB 81–164642).
Suggestion #8: Estimation of Pasquill
Stability Categories. U.S. Environmental experience with previous use, familiarity of Availability
Protection Agency, Office of Air Quality the model to various air quality programs,
The model codes and associated
Planning and Standards, Research and the costs and resource requirements for
documentation are available on EPA’s
Triangle Park, NC. (Docket No. A–80–46, use.
SCRAM website (paragraph A.0(3)).
(3) Codes and documentation for all
II–B–10).
models listed in this addendum are available Abstract
117. Mitchell, Jr., A.E. and K.O. Timbre,
from the EPA’s Support Center for Regulatory AERMOD is a steady-state plume
1979. Atmospheric Stability Class from
Air Models (SCRAM) website at https:// dispersion model for assessment of pollutant
Horizontal Wind Fluctuation. Presented www.epa.gov/scram. Codes and
at 72nd Annual Meeting of Air Pollution concentrations from a variety of sources.
documentation may also be available from AERMOD simulates transport and dispersion
Control Association, Cincinnati, OH; the National Technical Information Service
June 24–29, 1979. (Docket No. A–80–46, from multiple point, area, volume, and line
(NTIS), https://www.ntis.gov, and, when sources based on an up-to-date
II–P–9). available, are referenced with the appropriate
118. Smedman-Hogstrom, A. and V. characterization of the atmospheric boundary
NTIS accession number. layer. Sources may be located in rural or
Hogstrom, 1978. A Practical Method for
Determining Wind Frequency A.1 AERMOD (AMS/EPA Regulatory urban areas, and receptors may be located in
Distributions for the Lowest 200 m from Model) simple or complex terrain. AERMOD
Routine Meteorological Data. Journal of accounts for building wake effects (i.e.,
References plume downwash) based on the PRIME
Applied Meteorology, 17(7): 942–954.
U.S. Environmental Protection Agency, 2023. building downwash algorithms. The model
119. Smith, T.B. and S.M. Howard, 1972.
AERMOD Model Formulation. employs hourly sequential preprocessed
Methodology for Treating Diffusivity.
Publication No. EPA–454/B–23–010. meteorological data to estimate
MRI 72 FR–1030. Meteorology Research,
Office of Air Quality Planning and concentrations for averaging times from 1-
Inc., Altadena, CA. (Docket No. A–80– Standards, Research Triangle Park, NC. hour to 1-year (also multiple years).
46, II–P–8). Cimorelli, A., et al., 2005. AERMOD: A AERMOD can be used to estimate the
120. U.S. Environmental Protection Agency, Dispersion Model for Industrial Source concentrations of nonreactive pollutants from
2018. Evaluation of Prognostic Applications. Part I: General Model highway traffic. AERMOD also handles
Meteorological Data in AERMOD Formulation and Boundary Layer unique modeling problems associated with
Applications. Publication No. EPA–454/ Characterization. Journal of Applied aluminum reduction plants, and other
R–18–002. Office of Air Quality Planning Meteorology, 44(5): 682–693. industrial sources where plume rise and
and Standards, Research Triangle Park, Perry, S., et al., 2005. AERMOD: A downwash effects from stationary buoyant
NC. Dispersion Model for Industrial Source line sources are important. AERMOD is
121. U.S. Environmental Protection Agency, Applications. Part II: Model Performance designed to operate in concert with two pre-
1984. Calms Processor (CALMPRO) against 17 Field Study Databases. Journal processor codes: AERMET processes
User’s Guide. Publication No. EPA–901/ of Applied Meteorology, 44(5): 694–708. meteorological data for input to AERMOD,
9–84–001. Office of Air Quality Planning Heist, D., et al., 2013. Estimating near-road and AERMAP processes terrain elevation
and Standards, Region I, Boston, MA. pollutant dispersion: A model inter- data and generates receptor and hill height
(NTIS No. PB 84–229467). comparison. Transportation Research information for input to AERMOD.
a. Regulatory Use (2) Meteorological data: The AERMET and a morning upper air sounding, should
(1) AERMOD is appropriate for the meteorological preprocessor requires input of also be adequately representative of the
following applications: surface characteristics, including surface source area when using observed data.
• Point, volume, and area sources; roughness (zo), Bowen ratio, and albedo, as (ii) For applications involving the use of
• Buoyant, elevated line sources (e.g., well as, hourly observations of wind speed site-specific meteorological data that
aluminum reduction plants); between 7zo and 100 m (reference wind includes turbulences parameters (i.e., sigma-
• Mobile sources; speed measurement from which a vertical theta and/or sigma-w), the application of the
• Surface, near-surface, and elevated profile can be developed), wind direction, ADJ_U* option in AERMET would require
releases; cloud cover, and temperature between zo and approval as an alternative model application
• Rural or urban areas; 100 m (reference temperature measurement under section 3.2.
• Simple and complex terrain; from which a vertical profile can be (iii) For recommendations regarding the
• Transport distances over which steady- developed). Meteorological data can be in the length of meteorological record needed to
state assumptions are appropriate, up to 50 form of observed data or prognostic modeled perform a regulatory analysis with AERMOD,
km; data as discussed in paragraph 8.4.1(d). see section 8.4.2.
• 1-hour to annual averaging times, Surface characteristics may be varied by (3) Receptor data: Receptor coordinates,
• Continuous toxic air emissions; and, wind sector and by season or month. When elevations, height above ground, and hill
• Applications in the marine boundary using observed meteorological data, a height scales are produced by the AERMAP
layer environment where the effects of morning sounding (in National Weather terrain preprocessor for input to AERMOD.
shoreline fumigation and/or platform Service format) from a representative upper Discrete receptors and/or multiple receptor
downwash are adequately assessed or are not air station is required. Latitude, longitude, grids, Cartesian and/or polar, may be
applicable. and time zone of the surface, site-specific or employed in AERMOD. AERMAP requires
(2) For regulatory applications of prognostic data (if applicable) and upper air input of DEM or 3DEP terrain data produced
AERMOD, the regulatory default option meteorological stations are required. The by the USGS, or other equivalent data.
should be set, i.e., the parameter DFAULT wind speed starting threshold is also AERMAP can be used optionally to estimate
should be employed in the MODELOPT required in AERMET for applications source elevations.
record in the COntrol Pathway. The DFAULT involving site-specific data. When using
prognostic data, modeled profiles of c. Output
option requires the use of meteorological data
processed with the regulatory options in temperature and winds are input to Printed output options include input
AERMET, the use of terrain elevation data AERMET. These can be hourly or a time that information, high concentration summary
processed through the AERMAP terrain represents a morning sounding. Additionally, tables by receptor for user-specified
processor, stack-tip downwash, sequential measured profiles of wind, temperature, averaging periods, maximum concentration
date checking, and does not permit the use vertical and lateral turbulence may be summary tables, and concurrent values
of the model in the SCREEN mode. In the required in certain applications (e.g., in summarized by receptor for each day
regulatory default mode, pollutant half-life or complex terrain) to adequately represent the processed. Optional output files can be
decay options are not employed, except in meteorology affecting plume transport and generated for: a listing of occurrences of
the case of an urban source of sulfur dioxide dispersion. Optionally, measurements of exceedances of user-specified threshold
where a 4-hour half-life is applied. Terrain solar and/or net radiation may be input to value; a listing of concurrent (raw) results at
elevation data from the U.S. Geological AERMET. Two files are produced by the each receptor for each hour modeled, suitable
Survey (USGS) 7.5-Minute Digital Elevation AERMET meteorological preprocessor for for post-processing; a listing of design values
Model (DEM), or equivalent (approx. 30- input to the AERMOD dispersion model. that can be imported into graphics software
meter resolution and finer), (processed When using observed data, the surface file for plotting contours; a listing of results
through AERMAP) should be used in all contains observed and calculated surface suitable for NAAQS analyses including
applications. Starting in 2011, data from the variables, one record per hour. For NAAQS exceedances and culpability
3D Elevation Program (3DEP, https:// applications with multi-level site-specific analyses; an unformatted listing of raw
apps.nationalmap.gov/downloader), formerly meteorological data, the profile contains the results above a threshold value with a special
the National Elevation Dataset (NED), can observations made at each level of the structure for use with the TOXX model
also be used in AERMOD, which includes a meteorological tower (or remote sensor). component of TOXST; a listing of
range of resolutions, from 1-m to 2 arc When using prognostic data, the surface file concentrations by rank (e.g., for use in
seconds and such high resolution would contains surface variables calculated by the quantile-quantile plots); and a listing of
always be preferred. In some cases, prognostic model and AERMET. The profile concentrations, including arc-maximum
exceptions from the terrain data requirement file contains the observations made at each normalized concentrations, suitable for
may be made in consultation with the level of a meteorological tower (or remote model evaluation studies.
appropriate reviewing authority (paragraph sensor), the one-level observations taken d. Type of Model
3.0(b)). from other representative data (e.g., National
Weather Service surface observations), one AERMOD is a steady-state plume model,
b. Input Requirements record per level per hour, or in the case of using Gaussian distributions in the vertical
(1) Source data: Required inputs include prognostic data, the prognostic modeled and horizontal for stable conditions, and in
source type, location, emission rate, stack values of temperature and winds at user- the horizontal for convective conditions. The
height, stack inside diameter, stack gas exit specified levels. vertical concentration distribution for
velocity, stack gas exit temperature, area and (i) Data used as input to AERMET should convective conditions results from an
volume source dimensions, and source base possess an adequate degree of assumed bi-Gaussian probability density
elevation. For point sources subject to the representativeness to ensure that the wind, function of the vertical velocity.
influence of building downwash, direction- temperature and turbulence profiles derived e. Pollutant Types
specific building dimensions (processed by AERMOD are both laterally and vertically AERMOD is applicable to primary
through the BPIPPRM building processor) representative of the source impact area. The pollutants and continuous releases of toxic
should be input. Variable emission rates are adequacy of input data should be judged and hazardous waste pollutants. Chemical
optional. Buoyant line sources require independently for each variable. The values transformation is treated by simple
coordinates of the end points of the line, for surface roughness, Bowen ratio, and
exponential decay.
release height, emission rate, average line albedo should reflect the surface
source width, average building width, characteristics in the vicinity of the f. Source-Receptor Relationships
average spacing between buildings, and meteorological tower or representative grid AERMOD applies user-specified locations
average line source buoyancy parameter. For cell when using prognostic data, and should for sources and receptors. Actual separation
mobile sources, traffic volume; emission be adequately representative of the modeling between each source-receptor pair is used.
factor, source height, and mixing zone width domain. Finally, the primary atmospheric Source and receptor elevations are user input
are needed to determine appropriate model input variables, including wind speed and or are determined by AERMAP using USGS
inputs. direction, ambient temperature, cloud cover, DEM or 3DEP terrain data. Receptors may be
located at user-specified heights above j. Horizontal Dispersion Scheme in AERMOD. 13th Joint
ground level. Gaussian horizontal dispersion coefficients Conference on Applications of Air
g. Plume Behavior are estimated as continuous functions of the Pollution Meteorology, August 23–26,
parameterized (or measured) ambient lateral 2004; American Meteorological Society,
(1) In the convective boundary layer (CBL), Boston, MA.
turbulence and also account for buoyancy-
the transport and dispersion of a plume is U.S. Environmental Protection Agency, 2003.
induced and building wake-induced
characterized as the superposition of three AERMOD: Latest Features and
turbulence. Vertical profiles of lateral
modeled plumes: (1) the direct plume (from turbulence are developed from measurements Evaluation Results. Publication No.
the stack); (2) the indirect plume; and (3) the and similarity (scaling) relationships. EPA–454/R–03–003. Office of Air
penetrated plume, where the indirect plume Effective turbulence values are determined Quality Planning and Standards,
accounts for the lofting of a buoyant plume from the portion of the vertical profile of Research Triangle Park, NC.
near the top of the boundary layer, and the lateral turbulence between the plume height Heist, D., et al., 2013. Estimating near-road
penetrated plume accounts for the portion of and the receptor height. The effective lateral pollutant dispersion: A model inter-
a plume that, due to its buoyancy, penetrates turbulence is then used to estimate comparison. Transportation Research
above the mixed layer, but can disperse horizontal dispersion. Part D: Transport and Environment, 25:
downward and re-enter the mixed layer. In pp 93–105.
k. Vertical Dispersion
the CBL, plume rise is superposed on the Heist, D., et al., 2023. Integration of RLINE
displacements by random convective In the stable boundary layer, Gaussian dispersion model into EPA’s AERMOD:
velocities (Weil, et al., 1997). vertical dispersion coefficients are estimated updated formulation and evaluations.
(2) In the stable boundary layer, plume rise as continuous functions of parameterized Journal of the Air & Waste Management
is estimated using an iterative approach to vertical turbulence. In the convective Association, Manuscript submitted for
account for height-dependent lapse rates, boundary layer, vertical dispersion is publication.
similar to that in the CTDMPLUS model (see characterized by a bi-Gaussian probability Carruthers, D.J.; Stocker, J.R.; Ellis, A.;
A.2 in this appendix). density function and is also estimated as a Seaton, M.D.; Smith, SE Evaluation of an
continuous function of parameterized explicit NOx chemistry method in
(3) Stack-tip downwash and buoyancy
vertical turbulence. Vertical turbulence AERMOD; Journal of the Air & Waste
induced dispersion effects are modeled.
profiles are developed from measurements Management Association. 2017, 67 (6),
Building wake effects are simulated for stacks
and similarity (scaling) relationships. These 702–712; DOI:10.1080/
subject to building downwash using the turbulence profiles account for both
methods contained in the PRIME downwash 10962247.2017.1280096.
convective and mechanical turbulence. Environmental Protection Agency, 2023.
algorithms (Schulman, et al., 2000). For Effective turbulence values are determined
plume rise affected by the presence of a Technical Support Document (TSD) for
from the portion of the vertical profile of Adoption of the Generic Reaction Set
building, the PRIME downwash algorithm vertical turbulence between the plume height
uses a numerical solution of the mass, energy Method (GRSM) as a Regulatory Non-
and the receptor height. The effective vertical Default Tier-3 NO2 Screening Option.
and momentum conservation laws (Zhang turbulence is then used to estimate vertical
and Ghoniem, 1993). Streamline deflection Publication No. EPA–454/R–23–009.
dispersion. Office of Air Quality Planning &
and the position of the stack relative to the
l. Chemical Transformation Standards, Research Triangle Park, NC.
building affect plume trajectory and
dispersion. Enhanced dispersion is based on Chemical transformations are generally not A.2 CTDMPLUS (Complex Terrain
the approach of Weil (1996). Plume mass treated by AERMOD. However, AERMOD Dispersion Model Plus Algorithms for
captured by the cavity is well-mixed within does contain an option to treat chemical Unstable Situations)
the cavity. The captured plume mass is re- transformation using simple exponential
emitted to the far wake as a volume source. decay, although this option is typically not References
(4) For elevated terrain, AERMOD used in regulatory applications except for Perry, S.G., D.J. Burns, L.H. Adams, R.J.
incorporates the concept of the critical sources of sulfur dioxide in urban areas. Paine, M.G. Dennis, M.T. Mills, D.G.
dividing streamline height, in which flow Either a decay coefficient or a half-life is Strimaitis, R.J. Yamartino and E.M.
below this height remains horizontal, and input by the user. Note also that the Generic Insley, 1989. User’s Guide to the
flow above this height tends to rise up and Reaction Set Method, Plume Volume Molar Complex Terrain Dispersion Model Plus
Ratio Method and the Ozone Limiting Algorithms for Unstable Situations
over terrain (Snyder, et al., 1985). Plume
Method (section 4.2.3.4) for NO2 analyses are (CTDMPLUS). Volume 1: Model
concentration estimates are the weighted sum
available. Descriptions and User Instructions. EPA
of these two limiting plume states. However,
consistent with the steady-state assumption m. Physical Removal Publication No. EPA–600/8–89–041. U.S.
of uniform horizontal wind direction over the AERMOD can be used to treat dry and wet Environmental Protection Agency,
modeling domain, straight-line plume deposition for both gases and particles. Research Triangle Park, NC. (NTIS No.
trajectories are assumed, with adjustment in Currently, Method 1 particle deposition is PB 89–181424).
the plume/receptor geometry used to account available for regulatory applications. Method Perry, S.G., 1992. CTDMPLUS: A Dispersion
for the terrain effects. 2 particle deposition and gas deposition are Model for Sources near Complex
currently alpha options and not available for Topography. Part I: Technical
h. Horizontal Winds Formulations. Journal of Applied
regulatory applications
Vertical profiles of wind are calculated for Meteorology, 31(7): 633–645.
each hour based on measurements and n. Evaluation Studies
surface-layer similarity (scaling) American Petroleum Institute, 1998. Availability
relationships. At a given height above Evaluation of State of the Science of Air The model codes and associated
ground, for a given hour, winds are assumed Quality Dispersion Model, Scientific documentation are available on the EPA’s
constant over the modeling domain. The Evaluation, prepared by Woodward- SCRAM website (paragraph A.0(3)).
effect of the vertical variation in horizontal Clyde Consultants, Lexington,
Abstract
wind speed on dispersion is accounted for Massachusetts, for American Petroleum
through simple averaging over the plume Institute, Washington, DC 20005–4070. CTDMPLUS is a refined point source
Brode, R.W., 2002. Implementation and Gaussian air quality model for use in all
depth.
Evaluation of PRIME in AERMOD. stability conditions for complex terrain
i. Vertical Wind Speed applications. The model contains, in its
Preprints of the 12th Joint Conference on
In convective conditions, the effects of Applications of Air Pollution entirety, the technology of CTDM for stable
random vertical updraft and downdraft Meteorology, May 20–24, 2002; and neutral conditions. However,
velocities are simulated with a bi-Gaussian American Meteorological Society, CTDMPLUS can also simulate daytime,
probability density function. In both Boston, MA. unstable conditions, and has a number of
convective and stable conditions, the mean Brode, R.W., 2004. Implementation and additional capabilities for improved user
vertical wind speed is assumed equal to zero. Evaluation of Bulk Richardson Number friendliness. Its use of meteorological data
and terrain information is different from • Receptor information, and concentrations resulting from plume material
other EPA models; considerable detail for • Source-receptor location (line printer in each of these flows.
both types of input data is required and is map). (3) The model calculates on an hourly (or
supplied by preprocessors specifically (2) In addition, if the case-study option is appropriate steady averaging period) basis
designed for CTDMPLUS. CTDMPLUS selected, the listing includes: how the plume trajectory (and, in stable/
requires the parameterization of individual • Meteorological variables at plume height, neutral conditions, the shape) is deformed by
hill shapes using the terrain preprocessor and • Geometrical relationships between the each hill. Hourly profiles of wind and
the association of each model receptor with source and the hill, and temperature measurements are used by
a particular hill. • Plume characteristics at each receptor, CTDMPLUS to compute plume rise, plume
a. Regulatory Use i.e., penetration (a formulation is included to
—Distance in along-flow and cross flow handle penetration into elevated stable
CTDMPLUS is appropriate for the direction layers, based on Briggs (1984)), convective
following applications: —Effective plume-receptor height difference scaling parameters, the value of Hc, and the
• Elevated point sources; —Effective sy & sz values, both flat terrain Froude number above Hc.
• Terrain elevations above stack top; and hill induced (the difference shows the
• Rural or urban areas; h. Horizontal Winds
effect of the hill)
• Transport distances less than 50 —Concentration components due to WRAP, CTDMPLUS does not simulate calm
kilometers; and LIFT and FLAT. meteorological conditions. Both scalar and
• 1-hour to annual averaging times when (3) If the user selects the TOPN option, a vector wind speed observations can be read
used with a post-processor program such as summary table of the top four concentrations by the model. If vector wind speed is
CHAVG. at each receptor is given. If the ISOR option unavailable, it is calculated from the scalar
b. Input Requirements is selected, a source contribution table for wind speed. The assignment of wind speed
every hour will be printed. (either vector or scalar) at plume height is
(1) Source data: For each source, user done by either:
(4) A separate output file of predicted (1-
supplies source location, height, stack
hour only) concentrations (‘‘CONC’’) is • Interpolating between observations
diameter, stack exit velocity, stack exit above and below the plume height, or
written if the user chooses this option. Three
temperature, and emission rate; if variable
forms of output are possible: • Extrapolating (within the surface layer)
emissions are appropriate, the user supplies from the nearest measurement height to the
(i) A binary file of concentrations, one
hourly values for emission rate, stack exit plume height.
value for each receptor in the hourly
velocity, and stack exit temperature.
sequence as run; i. Vertical Wind Speed
(2) Meteorological data: For applications of
(ii) A text file of concentrations, one value Vertical flow is treated for the plume
CTDMPLUS, multiple level (typically three
for each receptor in the hourly sequence as component above the critical dividing
or more) measurements of wind speed and
run; or streamline height (Hc); see ‘‘Plume
direction, temperature and turbulence (wind (iii) A text file as described above, but with
fluctuation statistics) are required to create Behavior.’’
a listing of receptor information (names,
the basic meteorological data file positions, hill number) at the beginning of j. Horizontal Dispersion
(‘‘PROFILE’’). Such measurements should be the file. Horizontal dispersion for stable/neutral
obtained up to the representative plume (5) Hourly information provided to these conditions is related to the turbulence
height(s) of interest (i.e., the plume height(s) files besides the concentrations themselves velocity scale for lateral fluctuations, sv, for
under those conditions important to the includes the year, month, day, and hour which a minimum value of 0.2 m/s is used.
determination of the design concentration). information as well as the receptor number Convective scaling formulations are used to
The representative plume height(s) of interest with the highest concentration. estimate horizontal dispersion for unstable
should be determined using an appropriate conditions.
complex terrain screening procedure (e.g., d. Type of Model
CTSCREEN) and should be documented in CTDMPLUS is a refined steady-state, point k. Vertical Dispersion
the monitoring/modeling protocol. The source plume model for use in all stability Direct estimates of vertical dispersion for
necessary meteorological measurements conditions for complex terrain applications. stable/neutral conditions are based on
should be obtained from an appropriately e. Pollutant Types observed vertical turbulence intensity, e.g.,
sited meteorological tower augmented by sw (standard deviation of the vertical
SODAR and/or RASS if the representative CTDMPLUS may be used to model non- velocity fluctuation). In simulating unstable
plume height(s) of interest is above the levels reactive, primary pollutants. (convective) conditions, CTDMPLUS relies
represented by the tower measurements. f. Source-Receptor Relationship on a skewed, bi-Gaussian probability density
Meteorological preprocessors then create a Up to 40 point sources, 400 receptors and function (pdf) description of the vertical
SURFACE data file (hourly values of mixed 25 hills may be used. Receptors and sources velocities to estimate the vertical distribution
layer heights, surface friction velocity, are allowed at any location. Hill slopes are of pollutant concentration.
Monin-Obukhov length and surface assumed not to exceed 15°, so that the l. Chemical Transformation
roughness length) and a RAWINsonde data linearized equation of motion for Boussinesq Chemical transformation is not treated by
file (upper air measurements of pressure, flow are applicable. Receptors upwind of the CTDMPLUS.
temperature, wind direction, and wind impingement point, or those associated with
speed). any of the hills in the modeling domain, m. Physical Removal
(3) Receptor data: receptor names (up to require separate treatment. Physical removal is not treated by
400) and coordinates, and hill number (each CTDMPLUS (complete reflection at the
g. Plume Behavior
receptor must have a hill number assigned). ground/hill surface is assumed).
(4) Terrain data: user inputs digitized (1) As in CTDM, the basic plume rise
algorithms are based on Briggs’ (1975) n. Evaluation Studies
contour information to the terrain
preprocessor which creates the TERRAIN recommendations. Burns, D.J., L.H. Adams and S.G. Perry, 1990.
data file (for up to 25 hills). (2) A central feature of CTDMPLUS for Testing and Evaluation of the
neutral/stable conditions is its use of a CTDMPLUS Dispersion Model: Daytime
c. Output critical dividing-streamline height (Hc) to Convective Conditions. U.S.
(1) When CTDMPLUS is run, it produces separate the flow in the vicinity of a hill into Environmental Protection Agency,
a concentration file, in either binary or text two separate layers. The plume component in Research Triangle Park, NC.
format (user’s choice), and a list file the upper layer has sufficient kinetic energy Paumier, J.O., S.G. Perry and D.J. Burns,
containing a verification of model inputs, i.e., to pass over the top of the hill while 1990. An Analysis of CTDMPLUS Model
• Input meteorological data from streamlines in the lower portion are Predictions with the Lovett Power Plant
‘‘SURFACE’’ and ‘‘PROFILE,’’ constrained to flow in a horizontal plane Data Base. U.S. Environmental
• Stack data for each source, around the hill. Two separate components of Protection Agency, Research Triangle
• Terrain information, CTDMPLUS compute ground-level Park, NC.
Paumier, J.O., S.G. Perry and D.J. Burns, (ii) Over water: Hourly values for mixing (3) Wind speed profiles are estimated using
1992. CTDMPLUS: A Dispersion Model height, relative humidity, air temperature, similarity theory (Businger, 1973). Surface
for Sources near Complex Topography. and water surface temperature are required; layer fluxes for these formulas are calculated
Part II: Performance Characteristics. if wind speed/direction are missing, values from bulk aerodynamic methods.
Journal of Applied Meteorology, 31(7): over land will be used (if available); vertical i. Vertical Wind Speed
646–660. wind direction shear, vertical temperature
gradient, and turbulence intensities are Vertical wind speed is assumed equal to
A.3 OCD (Offshore and Coastal Dispersion) optional. zero.
Model (3) Receptor data: Location, height above j. Horizontal Dispersion
Reference local ground-level, ground-level elevation (1) Lateral turbulence intensity is
above the water surface. recommended as a direct estimate of
DiCristofaro, DC and S.R. Hanna, 1989. OCD:
c. Output horizontal dispersion. If lateral turbulence
The Offshore and Coastal Dispersion
Model, Version 4. Volume I: User’s (1) All input options, specification of intensity is not available, it is estimated from
Guide, and Volume II: Appendices. sources, receptors and land/water map boundary layer theory. For wind speeds less
Sigma Research Corporation, Westford, including locations of sources and receptors. than 8 m/s, lateral turbulence intensity is
MA. (NTIS Nos. PB 93–144384 and PB (2) Summary tables of five highest assumed inversely proportional to wind
concentrations at each receptor for each speed.
93–144392).
averaging period, and average concentration (2) Horizontal dispersion may be enhanced
Availability for entire run period at each receptor. because of obstructions near the source. A
The model codes and associated (3) Optional case study printout with virtual source technique is used to simulate
documentation are available on EPA’s hourly plume and receptor characteristics. the initial plume dilution due to downwash.
Optional table of annual impact assessment (3) Formulas recommended by Pasquill
SCRAM website (paragraph A.0(3)).
from non-permanent activities. (1976) are used to calculate buoyant plume
Abstract (4) Concentration output files can be used enhancement and wind direction shear
(1) OCD is a straight-line Gaussian model by ANALYSIS postprocessor to produce the enhancement.
developed to determine the impact of highest concentrations for each receptor, the (4) At the water/land interface, the change
offshore emissions from point, area or line cumulative frequency distributions for each to overland dispersion rates is modeled using
sources on the air quality of coastal regions. receptor, the tabulation of all concentrations a virtual source. The overland dispersion
OCD incorporates overwater plume transport exceeding a given threshold, and the rates can be calculated from either lateral
and dispersion as well as changes that occur manipulation of hourly concentration files. turbulence intensity or Pasquill-Gifford
as the plume crosses the shoreline. Hourly d. Type of Model curves. The change is implemented where
meteorological data are needed from both the plume intercepts the rising internal
OCD is a Gaussian plume model boundary layer.
offshore and onshore locations. These constructed on the framework of the MPTER
include water surface temperature, overwater model. k. Vertical Dispersion
air temperature, mixing height, and relative (1) Observed vertical turbulence intensity
e. Pollutant Types
humidity. is not recommended as a direct estimate of
(2) Some of the key features include OCD may be used to model primary
vertical dispersion. Turbulence intensity
platform building downwash, partial plume pollutants. Settling and deposition are not
treated. should be estimated from boundary layer
penetration into elevated inversions, direct theory as default in the model. For very
use of turbulence intensities for plume f. Source-Receptor Relationship stable conditions, vertical dispersion is also
dispersion, interaction with the overland (1) Up to 250 point sources, 5 area sources, a function of lapse rate.
internal boundary layer, and continuous or 1 line source and 180 receptors may be (2) Vertical dispersion may be enhanced
shoreline fumigation. used. because of obstructions near the source. A
a. Regulatory Use (2) Receptors and sources are allowed at virtual source technique is used to simulate
OCD is applicable for overwater sources any location. the initial plume dilution due to downwash.
where onshore receptors are below the lowest (3) The coastal configuration is determined (3) Formulas recommended by Pasquill
source height. Where onshore receptors are by a grid of up to 3600 rectangles. Each (1976) are used to calculate buoyant plume
above the lowest source height, offshore element of the grid is designated as either enhancement.
plume transport and dispersion may be land or water to identify the coastline. (4) At the water/land interface, the change
modeled on a case-by-case basis in g. Plume Behavior to overland dispersion rates is modeled using
consultation with the appropriate reviewing a virtual source. The overland dispersion
(1) The basic plume rise algorithms are rates can be calculated from either vertical
authority (paragraph 3.0(b)). based on Briggs’ recommendations. turbulence intensity or the Pasquill-Gifford
b. Input Requirements (2) Momentum rise includes consideration coefficients. The change is implemented
of the stack angle from the vertical. where the plume intercepts the rising
(1) Source data: Point, area or line source
(3) The effect of drilling platforms, ships, internal boundary layer.
location, pollutant emission rate, building
or any overwater obstructions near the source
height, stack height, stack gas temperature, l. Chemical Transformation
are used to decrease plume rise using a
stack inside diameter, stack gas exit velocity,
revised platform downwash algorithm based Chemical transformations are treated using
stack angle from vertical, elevation of stack
on laboratory experiments. exponential decay. Different rates can be
base above water surface and gridded
(4) Partial plume penetration of elevated specified by month and by day or night.
specification of the land/water surfaces. As
inversions is included using the suggestions
an option, emission rate, stack gas exit m. Physical Removal
of Briggs (1975) and Weil and Brower (1984).
velocity and temperature can be varied Physical removal is also treated using
(5) Continuous shoreline fumigation is
hourly. parameterized using the Turner method exponential decay.
(2) Meteorological data: PCRAMMET is the where complete vertical mixing through the
recommended meteorological data n. Evaluation Studies
thermal internal boundary layer (TIBL)
preprocessor for use in applications of OCD DiCristofaro, D.C. and S.R. Hanna, 1989.
occurs as soon as the plume intercepts the

employing hourly NWS data. MPRM is the TIBL. OCD: The Offshore and Coastal
recommended meteorological data Dispersion Model. Volume I: User’s
preprocessor for applications of OCD h. Horizontal Winds Guide. Sigma Research Corporation,
employing site-specific meteorological data (1) Constant, uniform wind is assumed for Westford, MA.
(i) Over land: Surface weather data each hour. Hanna, S.R., L.L. Schulman, R.J. Paine and
including hourly stability class, wind (2) Overwater wind speed can be estimated J.E. Pleim, 1984. The Offshore and
direction, wind speed, ambient temperature, from overland wind speed using relationship Coastal Dispersion (OCD) Model User’s
and mixing height are required. of Hsu (1981). Guide, Revised. OCS Study, MMS 84–
0069. Environmental Research & Coastal Dispersion (OCD) Model. Journal 4461, American Petroleum Institute,
Technology, Inc., Concord, MA. (NTIS of the Air Pollution Control Association, Washington, DC.
No. PB 86–159803). 35: 1039–1047. [FR Doc. 2023–22876 Filed 10–20–23; 8:45 am]
Hanna, S.R., L.L. Schulman, R.J. Paine, J.E. Hanna, S.R. and D.C. DiCristofaro, 1988.
BILLING CODE 6560–50–P
Pleim and M. Baer, 1985. Development Development and Evaluation of the
and Evaluation of the Offshore and OCD/API Model. Final Report, API Pub.

Appw 23 2023-22876

Uploaded by

Copyright:

Available Formats

Appw 23 2023-22876

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Appw 23 2023-22876

Uploaded by

Copyright:

Available Formats

72826 Federal Register / Vol. 88, No.

203 / Monday, October 23, 2023 / Proposed Rules

III. Public Participation air-quality-modeling.

consideration under the Guideline to

update has met the criteria required for

regulatory assessments. Climate, 16, 571–591.

content in section 8.3.1(a)(i) due to the

AERMOD. We also propose to clarify

Register. Information on the current status of

have been met.

concentrations are equal to the 1-hour

from highways, terminals, and other

but are common to dispersion modeling

Maximum allowable emission Actual or design capacity

Continuous operation, i.e., all

Maximum allowable emission Annual level when actually

Temporally representative level Continuous operation, i.e., all

Other Source(s) 6' 9

Proposed Major New or Modified Source

Maximum allowable emission

Continuous operation, i.e., all

Nearby Source(s) 4' 5

Maximum allowable emission Annual level when actually

Temporally representative level Continuous operation, i.e., all

Other Source(s) 5' 9

recommendations and requirements for the

occurs as soon as the plume intercepts the

You might also like