Office of Research and Development
CESER/HSMMD/DCB
Dispersion Modeling Systems
Relevant to Homeland
Security Preparedness and
Response
EMgaawwE
flMbnfLiARk.
J&SsSnfei.'Ss
• kfa&t wEim
'Mlw i .t ., a-:

-------
EPA/600/R-20/338
October 2020
Dispersion Modeling Systems
Relevant to Homeland
Security Preparedness and
Response
by
Michael Pirhalla
Office of Research and Development (ORD)
Research Triangle Park, NC 2771 1
HSRP RAP Product ID: HS19-01.01 - 4497
Project Officer
Center for Environmental Solutions and Emergency Response
(CESER)
Homeland Security and Materials Management Division (HSMMD)
Disaster Characterization Branch (DCB)
Research Triangle Park, NC 2771 1

-------
Disclaimer
The U.S. Environmental Protection Agency (EPA) through its Office of Research and Development
funded and managed this project through intramural research. It has been subjected to the Agency' s review
and has been approved for publication. Note that approval does not signify that the contents necessarily
reflect the views of the Agency. Any mention of trade names, products, or services does not imply an
endorsement by the U.S. Government or EPA. The EPA does not endorse any commercial products,
services, or enterprises.
Questions concerning this document, or its application should be addressed to:
Michael Pirhalla, M.S.
U.S. Environmental Protection Agency (US EPA)
Office of Research and Development (ORD)
Center for Environmental Solutions and Emergency Response (CESER)
Homeland Security and Materials Management Division (HSMMD)
Disaster Characterization Branch (DCB)
109 T.W. Alexander Dr. (MD E343-06)
Research Triangle Park, NC 27711
Phone: 919-541-0782
Pirhalla.michael@epa.eov
ii

-------
Abstract
As one of its core research focuses, the U.S. Environmental Protection Agency's (EPA's) Homeland
Security Research Program (HSRP) is interested in refining its tools and methodologies to better
characterize the fate and transport of hazardous contaminants during all phases of an emergency response.
Atmospheric dispersion modeling is one tool that can be used for effective emergency preparation or
response from hazardous chemical, biological, radiological, nuclear, and explosive (CBRNe) releases,
especially in urban areas where population densities are high and wind flow becomes altered between
buildings and street canyons. The goal of this report is to explain the fundamental concepts of atmospheric
transport and dispersion and provide a comprehensive database of dispersion models that can be used for
emergency preparation and response to facilitate discussion between public, private, academic, and/or
government sectors. The abundance of available modeling options creates confusion and results in
challenging decisions regarding the type of model to be used during different scenarios. A comprehensive
dispersion model review of this magnitude has also not occurred recently. This report provides a literature
review of previous model review efforts to lay the foundation for this updated database, provides
introductory concepts on boundary layer meteorology and the types of dispersion models available (e.g.
Gaussian Plume or Puff, Lagrangian, or CFD models), and outlines a comprehensive list of 96 dispersion
models that could be considered for wide-area release risks. Sixteen of those models were selected for a
more detailed two-page review due to their potential applicability and usefulness for emergency response.
This model review is not meant to recommend or endorse a specific model, but to provide users with a
resource of available modeling options. Even though no single model tends to have all the capabilities that
are beneficial during the consequence management of a wide area release, this report is meant to identify
the strengths and limitations so users can make informed decisions.
This report covers a research period from September 2018 to June 2020 and work was completed as of
July 2020 as part of the author's Ph.D. dissertation.
iii

-------
Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the
Nation's land, air, and water resources. Under a mandate of national environmental laws, the Agency
strives to formulate and implement actions leading to a compatible balance between human activities and
the ability of natural systems to support and nurture life. To meet this mandate, EPA's research program
is providing data and technical support for solving environmental problems today and building a science
knowledge base necessary to manage our ecological resources wisely, understand how pollutants affect
our health, and prevent or reduce environmental risks in the future.
The Center for Environmental Solutions and Emergency Response (CESER) within the Office of
Research and Development (ORD) conducts applied, stakeholder-driven research and provides responsive
technical support to help solve the Nation's environmental challenges. The Center's research focuses on
innovative approaches to address environmental challenges associated with the built environment. We
develop technologies and decision-support tools to help safeguard public water systems and groundwater,
guide sustainable materials management, remediate sites from traditional contamination sources and
emerging environmental stressors, and address potential threats from terrorism and natural disasters.
CESER collaborates with both public and private sector partners to foster technologies that improve the
effectiveness and reduce the cost of compliance, while anticipating emerging problems. We provide
technical support to EPA regions and programs, states, tribal nations, and federal partners, and serve as
the interagency liaison for EPA in homeland security research and technology. The Center is a leader in
providing scientific solutions to protect human health and the environment.
Gregory Sayles, Director
Center for Environmental Solutions and Emergency Response
iv

-------
Table of Contents
Disclaimer	ii
Abstract	iii
Foreword	iv
Table of Contents	v
List of Figures	viii
List of Tables	ix
Acronyms and Abbreviations	x
Acknowledgments	xvi
1.0 Introduction	1
2.0 Project Background and Goals	3
2.1	Proj ect Moti vati on	3
2.2	Background and Previous Model Review Efforts	4
2.3	Project Goals	5
2.4	Literature Quality Assurance	6
3.0 Emergency Response and Dispersion Modeling	7
3.1	Dispersion Modeling Definition	7
3.2	CBRNe Terminology	7
3.2.1	Chemical	7
3.2.2	Biological	8
3.2.3	Radiological	8
3.2.4	Nuclear	8
3.2.5	Explosive	9
3.3	Stages of an Emergency Response	9
3.3.1	Prevention Framework	9
3.3.2	Protecti on F ram ework	10
3.3.3	Mitigation Framework	10
3.3.4	Response Framework	10
3.3.5	Recovery F ram ework	10
3.4	Operational Dispersion Modeling and Reach Back	11
3.4.1	Interagency Modeling and Atmospheric Assessment Center (IMAAC)	11
3.4.2	Models Used in IMAAC Responses	13
3.5	Hazardous Release Mitigation Programs and Risk Evaluation	14
3.5.1	Emergency Planning and Community Right-to-Know Act (EPCRA)	14
3.5.2	EPA's Risk Management Plan Rule	14
v

-------
3.5.3	BioWatch	15
3.5.4	Chemical Facility Antiterrorism Standards Program	16
3.6 EPA's Contributions to Emergency Response Initiatives and Dispersion Modeling	16
3.6.1	EPA Support Center for Regulatory Atmospheric Modeling	17
3.6.2	RMP*Comp	17
3.6.3	CAMEO/ALOHA	18
4.0 Atmospheric and Micrometeorological Fundamentals in Dispersion Modeling	19
4.1	Atmospheric Turbulence	19
4.2	Planetary Boundary Layer	20
4.3	Modifications to Urban Flow from Building Structures	22
4.4	Vertical Wind Profile	24
5.0 Types of Atmospheric Dispersion Models	26
5.1	Box Models	26
5.2	Gaussian Plume Models	27
5.3	Gaussian Puff Models	28
5.4	Lagrangian Stochastic Particle Models	28
5.5	Eulerian Grid Models	29
5.6	Higher Order Models	29
5.7	Street Network Models	30
5.8	Comparisons, Strengths, and Limitations for Atmospheric Dispersion Models	32
6.0 Model Review Process	35
6.1	Quick Reference Table	35
6.2	Expanded Model Description	36
7.0 Dispersion Models- Quick Reference Table	39
8.0 Expanded Model Descriptions	56
8.1	ADAM Tool	56
8.2	ADAPT/I .01)1	58
8.3	Aeolus	60
8.4	AERMOD	62
8.5	ALOHA (CAMEO)	65
8.6	CALPUFF	67
8.7	CASRAM	69
8.8	CT-Analyst	71
8.9	DEGADIS	73
8.10	HotSpot	75
8.11	HP AC	77
vi

-------
8.12	HYSPLIT	80
8.13	JEM	82
8.14	MELCOR and MACCS	84
8.15	QUIC	87
8.16	SHARC/ERAD	89
9.0 Concluding Remarks	91
10.0 References	93
vii

-------
List of Figures
Figure 1: Summary of IMAAC's agency responsibilities with the typical dispersion model used, based
on information from DTRA (Runyon 2017)	14
Figure 2: The common, but idealized, diurnal evolution of the Planetary Boundary Layer (PBL) adapted
from Stull (1988)	21
Figure 3: Schematic of a) cavity and wake flow zones associated with a building or other square
obstacle in the mean flow and b) its relationship with the vertical wind profile, after Halitsky (1968). . 23
Figure 4: Schematic of streamlines when perpendicular flow encounters a street canyon, based on
Dabberdt et al. (1973)	23
Figure 5: Various perpendicular flows for urban canyons with a) isolated roughness, b) wake
interference, and c) skimming flow aspect ratios, based on Oke (1988)	24
Figure 6: Components of an atmospheric dispersion model, modified after OFCM (2002) and Turner
(1979)	27
Figure 7: Visual comparison of dispersion models that can be applied for homeland security, emergency
preparedness, and emergency response. As the models increase in complexity, so do their computational
and user requirements. Many have special applications for urban use. Figure adapted from presentation:
National Atmospheric Release Advisory Center's Urban Plume Dispersion Modeling Capability for
Radiological Sources by Gowardhan et al. (2018) with additional modifications	32
viii

-------
List of Tables
Table 1: Summary of strengths and limitations for different types of atmospheric dispersion models. . 33
Table 2: Model classification criteria for inclusion or omission in detailed model review	37
Table 3: Model criteria and explanation of information provided in the expanded model descriptions. 38
ix

-------
Acronyms and Abbreviations
2D	Two-Dimensional
3D	Three-Dimensional
ABC	Atomic, Biological, and Chemical
ADAM	Air Force Dispersion Assessment Model; Accident Damage Analysis Module
ADAPT	Atmospheric Data Assimilation and Parameterization Tool
ADEME	French Ministry and Environmental Agency
ADMLC	Atmospheric Dispersion Modeling Liaison Committee
ADMS	Atmospheric Dispersion (and Dose Assessment) Modeling System
AER	Atmospheric and Environmental Research
AERMIC	American Meteorological Society/EPA Regulatory Model Improvement
Committee
AERMOD	American Meteorological Society/Environmental Protection Agency
Regulatory Model
AES	Atmospheric Environment Service
AFTOX	Air Force Toxics Model
AIR	Atmosphere, Impact, and Risk
AMS	American Meteorological Society
ANL	Argonne National Laboratory
APGEMS	Air Pollutant Graphical Environmental Monitoring System
AQPAC	Air Quality Package
ARA	Applied Research Associates
ARAC	Atmospheric Release Advisory Capability
ARCHIE	Automated Resource for Chemical Hazard Incident Evaluation
ARCON	Atmospheric Relative Concentrations
ARGOS	Accident Reporting and Guidance System
ARL	Air Resources Laboratory
ASPEN	Assessment System for Population Exposure Nationwide
ATD	Atmospheric Transport and Diffusion
BAR	BioWatch Actionable Result
BERT	BioWatch Event Reconstruction Tool
BLP	Buoyant Line and Point (Source Model)
BNL	Brookhaven National Laboratory
BNLGPM	Brookhaven National Laboratory Gaussian Plume Model
CAA	Clean Air Act
CALINE	California Line (Source Dispersion Model)
CALPUFF	California Puff (Model)
CAMEO/ALOHA Computer-Aided Management of Emergency Operations/Areal Locations of
Hazardous Atmospheres
CAPARS	Computer-Assisted Protective Action Recommendations System
CASRAM	Chemical Accident Stochastic Risk Assessment Model
CATS-JACE	Consequence Assessment Tool Set/Joint Assessment of Catastrophic Events
x

-------
CBRNe
Chemical, Biological, Radiological, Nuclear, and Explosive
CERC
Cambridge Environmental Research Consultants
CERCLA
Comprehensive Environmental Response, Compensation, and Liability Act
CESER
Center for Environmental Solutions and Emergency Response
CFATS
Chemical Facility Antiterrorism Standards
CFD
Computational Fluid Dynamics
CISRO
Commonwealth Scientific and Industrial Research Organisation
CMAQ
Community Multiscale Air Quality (Modeling System)
CMI
Christian Michelsen Institute
CML
Convective Mixed Layer
CO
Carbon Monoxide
COOP
Continuity of Operations Plan
CSA
Combat Support Agency
CsCl
Cesium Chloride
CTDMPLUS
Complex Terrain Dispersion Model Plus (Algorithms for Unstable Situations)
CUDM
Canadian Urban Dispersion Model
DCB
Disaster Characterization Branch
DEGADIS
Dense Gas Dispersion (Model)
DELFIC/FPTool
Defense Land Fallout Interpretive Code/ Fallout Planning Tool
DERMA
Danish Emergency Response Model of the Atmosphere
DHHS
U.S. Department of Health and Human Services
DHS
U.S. Department of Homeland Security
DNS
Direct Numerical Simulation
DOC
U.S. Department of Commerce
DOD
U.S. Department of Defense
DOE
U.S. Department of Energy
DOT
U.S. Department of Transportation
DRIFT
Dispersion of Releases Involving Flammables or Toxics
DSTL
Defence Science and Technology Laboratory
DTRA
Defense Threat Reduction Agency
EC
European Commission
EMS
Emergency Medical Service
EOC
Emergency Operations Center
EPA
U.S. Environmental Protection Agency
EPCRA
Emergency Planning and Community Right-to-Know Act
EPICode
Emergency Prediction Information Code
EPRI
Electric Power Research Institute
ERG
Environmental Response Guidebook
ERT
Environmental Response Team; Environmental Research and Technology, Inc
ESCAPE
Expert System for Consequence Analysis and Preparing for Emergencies
EU
European Union
EULAG
EUlerian LAGrangian (Model)
FEM3MP
Finite Element Model in 3-Dimensions and Massively Parallelized
xi

-------
FEMA	Federal Emergency Management Agency
FLACS	FLame Acceleration Simulator
FLEXPART	Flexible Particle (Dispersion Model)
FOI	Swedish Defence Research Agency
FRERP	Federal Radiological Emergency Response Plan
FRMAC	Federal Radiological Monitoring and Assessment Center
FRP	Federal Response Plan
GAO	Government Accountability Office
GENII	Generalized Environmental Radiation Dosimetry Software System - Hanford
Dosimetry System (Gen. II)
GIS	Geographic Information System
GMU	George Mason University
GUI	Graphical User Interface
HASP	Hazard Assessment Simulation and Prediction
HIGRAD/	High-Resolution Model for Strong Gradient Applications Fire Behavior
FIRETEC	(Model)
HOTMAC	Higher Order Turbulence Model for Atmospheric Circulation
HPAC	Hazard Prediction and Assessment Capability
HSE	Health and Safety Executive
HSIN	Homeland Security Information Network
HSMMD	Homeland Security and Materials Management Division
HSPD	Homeland Security Presidential Directive
HSRP	Homeland Security Research Program
HYROAD	Hybrid Roadway (Intersection Model)
HYSPLIT	Hybrid Single-Particle Lagrangian Integrated Trajectory
IBL	Internal Boundary Layer
ICS	Incident Command System
IED	Improvised Explosive Device
IEM	Innovative Emergency Management, Inc
IMAAC	Interagency Modeling and Atmospheric Assessment Center
INL	Idaho National Laboratory
INPUFF	(Gaussian) Integrated Puff (Model)
ISC	Industrial Source Complex (Model)
JEM	Joint Effects Model
JOULES	Joint Outdoor-indoor Urban Large-Eddy Simulation
JRC	Joint Research Centre
JRTT	Jack Rabbit II
KBERT	Knowledge-Based-system for Estimating hazards of Radioactive material
release Transients
KDFOC	"K" Division (Defense Nuclear) Fallout Code
LANL	Los Alamos National Laboratory
LAPMOD	LAgrangian Particle MODel
LBNL	Lawrence Berkeley National Laboratory
xii

-------
LES	Large Eddy Simulation
LFA	Lead Federal Agency
LLNL	Lawrence Livermore National Laboratory
LODI	Lagrangian Operational Dispersion Integrator
LPDM	Lagrangian Particle Dispersion Model
MACCS	MELCOR Accident Consequence Code System
MAHB	Maj or Accident Hazards Bureau
MATHEW/	Mass-Adjusted Three-Dimensional Wind Field/Atmospheric Diffusion Particle-
ADPIC	in-Cell
MIDAS-AT	Meteorological Information Dispersion and Assessment System Anti-Terrorism
MOU	Memorandum of Understanding
MSS	Micro-Swift Spray
NAAQS	National Ambient Air Quality Standards
NAM	North American Model
NAME	Numerical Atmospheric-Dispersion Modeling Environment
NARAC	National Atmospheric Release Advisory Center
NBC	Nuclear, Biological, and Chemical
NCAR	National Center for Atmospheric Research
NCEP	National Centers for Environmental Prediction
NCHRP	National Cooperative Highway Research Program
NEF	National Essential Function
NHSRC	National Homeland Security Research Center
NOAA	National Oceanic and Atmospheric Administration
NOx	Nitrogen Oxides
NRC	Nuclear Regulatory Commission
NRF	National Response Framework
NWP	Numerical Weather Prediction
NWS	National Weather Service
O3	Ozone
OBODM	Open Burn/Open Detonation Dispersion Model
OCD	Offshore and Coastal Dispersion (Model)
OECD	Organisation for Economic Co-operation and Development
OFCM	Office of the Federal Coordinator for Meteorology
OMEGA	Operational Multiscale Environment (Model) with Grid Adaptivity
ORD	Office of Research and Development
ORNL	Oak Ridge National Laboratory
OSC	On Scene Coordinator
OSHA	Occupational Safety and Health Administration
OSPM	Operational Street Pollution Model
PANACHE	Atmosphere Pollution and Industrial Risk Analysis
PBL	Planetary Boundary Layer
PHAST	Process Hazard Analysis Software
PI	Principal Investigator
xiii

-------
PLUVUE	Plume Visibility (Model)
PM	Particulate Matter
PMEF	Primary Mission Essential Function
PNNL	Pacific Northwest National Laboratory
PPE	Personal Protective Equipment
PUMA	Puff Model of Atmospheric Dispersion
QUIC	Quick Urban Industrial Complex
RA/HA	Risk Assessment/Hazard Assessment
RANS	Reynolds-averaged Navier-Stokes
RAPTAD	Random Puff Transport and Diffusion
RASCAL	Radiological Assessment System for Consequence Analysis
RCRA	Resource Conservation and Recovery Act
RDD	Radiological Dispersal Device
RIMPUFF	Ris0 Mesoscale Puff Model
RL	Residual Layer
RLINE	Research Line-source (Dispersion Model)
RMP	Risk Management Plan
RSAC	Radiological Safety Analysis Computer (Program)
RTDM	Rough Terrain Dispersion Model
RTVSM	Real-time Volume Source Model
SAIC	Science Applications International Corporation
SBL	Stable Boundary Layer
SCIPUFF	Second-order Closure Integrated Puff (Model)
SCRAM	Support Center for Regulatory and Atmospheric Modeling
SDM	Shoreline Dispersion Model
SHARC/ERAD	Specialized Hazard Assessment Response Capability/ Explosive Release
Atmospheric Dispersion
SIP	State Implementation Plan
SL	Surface Layer
SNL	Sandia National Laboratory
SOARCA	State-of-the-Art Reactor Consequence Analyses
SRC	Sigma Research Corporation
SRS	Savannah River Site
STILT	Stochastic Time-Inverted Lagrangian Transport (Model)
TAPM	The Air Pollution Model
TIC	Toxic Industrial Chemical
TKE	Turbulent Kinetic Energy
TOPOFF	(National) Top Officials (Exercise)
TRAC	Terrain Responsive Atmospheric Code
TRACE	Toxic Release Analysis of Chemical Emissions
U.S.	United States
UBL	Urban Boundary Layer
UDM	Urban Dispersion Model
xiv

-------
UHI
Urban Heat Island
UK
United Kingdom
UoR-SNM
University of Reading Street Network Model
USD A
U.S. Department of Agriculture
USGS
U.S. Geological Survey
VAFTAD
Volcanic Ash Forecast Transport and Dispersion
VAPO
Vulnerability Analysis and Protection Option
VLSTRACK
Vapor, Liquid, and Solid Tracking
WADOCT
Wind and Diffusion Over Complex Terrain
WINDS
Weather Information and Display System
WMD
Weapon of Mass Destruction
WRF
Weather Research and Forecasting Model
YSA
Yamada Science and Art (Corporation)
XV

-------
Acknowledgments
This effort was developed and written by the EPA principal investigator (PI). Contributions of the
following individuals are gratefully acknowledged:
EPA PI
Michael Pirhalla, CESER/HSMMD/DCB
Pirhalla.michael@epa.eov
EPA Technical Reviewers
David Heist, CEMM/AESMB/ESAB
Heist, david@epa.eov
Brian Eder, CEMM/AESMB/ADMB
Eder.brian@epa.eov
External Technical Reviewers
Steven Perry, US EPA, Retired
Lakeaum an@em ail. com
S. Pal Arya, North Carolina State University, Marine, Earth, and Atmospheric Sciences Department
Sparva@ncsu.edu
EPA Quality Assurance
Ramona Sherman, CESER
Eletha Brady Roberts, CESER
Joan Bursey, CESER/HSMMD
EPA Technical Editor
Joan Bursey, CESER/HSMMD
xv i

-------
1.0 Introduction
After the events of September 11,2001, the United States quickly became aware of its vulnerability
to external terrorism-related threats. To minimize and prepare for potentially adverse future situations, the
Federal government established the U.S. Department of Homeland Security (DHS). The agency's goal
was to prepare and protect the country against terrorism, instill a presence of border security, and prepare
for and manage disaster scenarios. Although the responsibilities of DHS are broad, an important
component of its homeland security efforts is the preparation, detection, response, and mitigation of
hazardous substance releases into the ambient atmosphere, as these situations have the potential to affect
the health and welfare of the American people.
To help fulfill the core objectives, DHS works alongside various government entities, including
the United States (US) Environmental Protection Agency (EPA). In September 2002, the US EPA formed
the National Homeland Security Research Center (NHSRC) to lead scientific-based research and provide
technical expertise for a variety of environmental and human health-related homeland security threats. In
2019, EPA's Office of Research and Development (ORD) reorganized and much of the same research is
now handled the Homeland Security Research Program (HSRP). EPA's HSRP and its partners work to
develop risk-prevention strategies that strengthen the country's ability to withstand and recover from
future disasters and wide-area incidents, whether the hazards stem from natural, accidental, or terrorist-
related sources. These wide-area events could span the spatial distance of several city blocks or more,
such as within an urban area like lower Manhattan, or throughout a municipality's drinking water
distribution system (EPA 2020). Situations that involve the release, or potential release, of hazardous
chemicals, microbial pathogens, or radiological materials further complicate disaster scenarios and require
specialized expertise during the response and recovery process. As one of its core research focuses, HSRP
is particularly interested in refining its tools and methodologies for a better understanding of the fate and
transport of hazardous wide-area contaminants (EPA 2020). This need for tool and methodology
refinement extends to all phases of an emergency response, from the near-term to the extended remediation
and recovery stages.
Upon its creation, DHS was slated to develop new countermeasures for chemical, biological,
radiological, nuclear, and explosive (CBRNe) releases, which would include improved knowledge of
atmospheric transport and diffusion (ATD) through computer dispersion modeling. A dispersion model is
a mathematical representation of the transport of air pollutants in the ambient atmosphere which is used
to calculate concentrations at various locations away from the emission source(s) (Holmes and Morawska
2006). The equations governing pollutant dispersion are frequently based on a Gaussian (bell-shaped)
downwind concentration distributions and are solved through computer modeling software.
Understanding complex atmospheric flow and dispersion processes, especially in urban areas, is important
when modeling hazardous air quality scenarios. These efforts are supported by the U.S. Department of
Commerce (DOC) Office of the Federal Coordinator for Meteorology (OFCM) and several collaborating
federal agencies that developed guidance for dispersion modeling implementation (OFCM 2002).
In the federal government, operational dispersion modeling is a multiagency approach. Hazardous
accidental release scenarios may arise from accidental industrial and transportation-related contaminant
spills and intentional acts of terrorism. To provide a single point for the coordination and dissemination
of hazard prediction products, DHS established a multiagency working group in 2004 called the
Interagency Modeling and Atmospheric Assessment Center (IMAAC). IMAAC was not intended to
replace individual dispersion modeling efforts but is able to be activated quickly if a hazardous release
occurs and an emergency plume estimation is required under tight time constraints. Dispersion modeling
in its research, regulatory, or academic role, is a multifaceted scientific tool used, developed, and improved
by many private, university, state, and federal government entities, including the EPA.
1

-------
While considerable research and development effort has been leveraged in dispersion modeling
improvements over the past several decades, especially for CBRNe releases, there is still room for further
development. The critical need for advancements in atmospheric modeling and plume prediction has been
rekindled from the events on September 11, as well as other numerous hazardous situations, including the
threat of wide-area Bacillus anthracis (anthrax) releases (Amerithrax), the 2011 Fukushima nuclear
reactor accident, the 2017 Portland, Oregon asbestos fire, or more recently the 2020 Visakhapatnam
styrene gas leak, which could have had a result similar to the 1984 Bhopal disaster. These scenarios are
just a few cases demonstrating the critical need for dispersion models to be continuously tested, developed,
and improved, usually by evaluating their performance against extensive field and laboratory data. Since
it is impossible to predict the timing and location of the next catastrophic incident, emergency responders
must be prepared for a multitude of hazardous releases. Dispersion modeling offers a critical insight in
emergency preparation or planning so responders can become better equipped for various release
scenarios. It is also a critical component or tool for efficient and precise emergency response (Leitl et al.
2016), especially for determining the extent of a toxic plume and informing where to evacuate, sample,
decontaminate surfaces, and manage waste.
Building on the fundamental concepts and physical understanding of atmospheric transport and
diffusion that emerged in the early-to-mid twentieth century when the foundation of ATD research was
laid (Richardson 1922; Taylor 1921; Pasquill 1961, and others), today's model development activities
focus on more complex circumstances. These situations are particularly challenging in urban areas with
high population densities and the potential for acute exposure effects (Schmidtgoessling 2009). The
complex nature of a cityscape results in substantial challenges in determining pollution dispersion
throughout the urban canopy (Garbero 2008). Wind flow patterns become altered by the urban geometry,
and turbulent flows are generated between buildings and streets (Belcher et al. 2013; Barlow and Coceal
2009; Britter and Hanna 2003). The urban canopy also tends to modify the local boundary layer by
reducing wind speeds, increasing turbulent intensities and turbulent kinetic energy (TKE) in the lee of
buildings, and increasing episodes of neutral stability instead of extreme stability through added
turbulence and heat fluxes (Arya 2001; Briggs 1973). The simplifying assumptions in many dispersion
models make urban, industrial, and small-scale modeling quick and efficient for rapid results but also
introduce errors that could propagate to poor model performance (Chang et al. 2005). An acceptable
balance between speed, model performance, ease of use, and purpose of application must be established
when employing a dispersion model for emergency response. As a result, the use, analysis, and
implementation of atmospheric dispersion models, along with improvements and more in-depth
understanding of micro- and mesoscale transport processes, are key research priorities within the EPA.
Improved dispersion research can also aid the EPA's emergency response mission in preparing for and
responding to large-scale CBRNe incidents as part of the Homeland Security Research Program (HSRP)
(EPA 2020).
This document first outlines the project background, justification, and goals in Section 2.0, along
with a short literature review of previous dispersion model compilations. Section 3.0 identifies the role of
dispersion modeling in emergency preparation and response, details the available operational dispersion
modeling resources, and defines EPA's role in emergency response. Section 4.0 provides an overview of
atmospheric turbulence and the fundamentals of dispersion within the Planetary Boundary Layer (PBL)
and urban areas. The types and corresponding strengths and limitations for different dispersion models
are covered in Section 5.0. Section 6.0 describes the model review process, specific details included in
the review, and the criteria used to determine inclusion or omission of the model in the detailed review.
An extensive quick reference table for 96 different dispersion models is provided in Section 7.0, and 16
of those models are selected for additional review in Section 8.0 due to their applicability and usefulness
for emergency response.
2

-------
2.0	Project Background and Goals
Dispersion model users often must make challenging decisions on the type of model to select for
their unique scenario depending on release type, terrain, urban geometry, and time considerations. The
abundance of publicly available, proprietary, or no-longer-supported atmospheric dispersion models (or
simply "dispersion models") often creates considerable confusion for investigators attempting to select
and use a model for their purpose, especially as research and scientific knowledge of atmospheric
dispersion continues to advance. Although the assessment is now somewhat dated, the OFCM noted that
there were over 140 types of public and proprietary dispersion models developed for a variety of purposes
(OFCM 2002). However, only a small subset of those models is readily accessible or still being used in
regulatory efforts and urban emergency planning initiatives while others are not designed for emergency
planning or response.
2.1	Project Motivation
The purpose of this report is to provide a comprehensive database of dispersion models while also
briefly explaining the fundamental concepts of atmospheric transport and dispersion incorporated in each
model. This document outlines and alphabetically sorts dispersion modeling systems and acts as a
comprehensive guide for modelers to rapidly relay risk, sampling, and various model choices to decision
makers. The ATD background information (Section 4.0) and model type summaries (Section 5.0) are
intended to provide a quick reference for those new to air dispersion modeling or for those seeking to
expand their knowledge base but is not meant to replace primary literature sources such as textbooks.
Literature is introduced from various academic journal articles, textbooks, government documents, and
various reports to provide a diverse synthesis of information. Currently available dispersion models,
situations where they are most applicable, model availability and runtime, and notable studies and
publications from academic articles are also detailed. Many models can simulate atmospheric transport
and diffusion; however, this report emphasizes dispersion models that have specific emergency
preparation and response applications, urban or complex environment capabilities, and those that can
simulate scenarios related to a variety of hazardous CBRNe cases. Because individual models are
necessarily limited in scope and may not have all the components required to be beneficial during
consequence management of a wide area response (Mikelonis et al., 2018), this report aims to document
the capabilities of dispersion models within the framework of emergency response and preparedness.
The goal is to provide EPA researchers, emergency response planners (federal, regional, and/or
local), and policymakers an additional resource to make informed decisions regarding dispersion model
use. Emergency planners may find this document a useful reference and ideal starting point when learning
about potential modeling resources. This document may also be beneficial when attempting to select a
dispersion model to assess local emergency planning exercises such as within areas with high levels of
potential human exposure. EPA scientists may use this document as a resource when developing research
projects or field studies that involve an airborne release. This report is also intended to facilitate discussion
between public, private, academic, and/or government sectors to aid in the selection of a useful dispersion
model during the preparation, response, or recovery phases.
The need to periodically review the state of dispersion modeling arises from the continuous growth
in our understanding of boundary layer turbulence, dispersion, ongoing model development, and the sheer
number of dispersion model variations. Some of these models have not been updated recently and are
retired, while others are proprietary and may be used only by their developers or paid subscribers. Other
dispersion models are only suitable for specific releases (i.e., radiological release, explosions, dense gas).
In a charge recommended by the Atmospheric Dispersion Modeling Liaison Committee (ADMLC)
appointed in the United Kingdom, "a qualitative assessment of the 'use-ability' of a model should be
3

-------
undertaken, considering the extent to which the model is user-friendly, the data requirements of the model,
and accessibility and availability of such models" (ADMLC 2013). A similar request was raised in the
U.S. at the 22nd Annual George Mason University (GMU) Conference of Atmospheric Transport and
Dispersion Modeling in June 2018. The OFCM called upon a joint action working group of various
agencies to revise the 2002 and 2004 atmospheric modeling guidebooks for homeland security
applications (see: OFCM 2002; 2004), but the status of this update is not known. As these documents are
16-18 years old (at the time of this report), a considerable number of changes may be warranted.
Additionally, OFCM used to publish a directory of consequence assessment dispersion models, but that
document has not been updated in over two decades (OFCM 1999).
The U.S. Government Accountability Office (GAO) also noted in a 2008 report for DHS that
confusion and lack of coordination between government agencies has existed when these agencies respond
to simulated homeland security incidents, demonstrating that the federal government struggles to
efficiently "coordinate and properly use atmospheric transport and dispersion models" (US GAO 2008).
From discussions with on-scene coordinators and EPA researchers, as well as literature and guidance
documents, considerable confusion still exists regarding the options for current dispersion models, their
capabilities, and applicability of use. The need for a coordinated and centralized response has led to the
establishment of IMAAC and the National Atmospheric Release Advisory Center (NARAC). A resource
such as this report may be useful for individuals who want to select a dispersion model for research or
planning scenarios.
2.2 Background and Previous Model Review Efforts
The OFCM, which leads a collaboration with at least 14 federal agencies including the National
Oceanic and Atmospheric Administration (NOAA), U.S. Department of Defense (DOD), and U.S.
Department of Energy (DOE), and EPA developed a series of comprehensive dispersion modeling
directories (OFCM 1993, 1999, 2002, 2004). Those reviews, last published in the late 1990s, provided an
in-depth compilation and description of atmospheric dispersion models available to those with
consequence assessment requirements, especially those requiring real-time information on chemical,
radiological, or biological weapon emergencies. The first version of the report published by OFCM in
1993 was titled "Directory of Atmospheric Transport and Diffusion Models, Equipment, and Projects''
(FCM-I3-1993) (OFCM 1993), followed by an update in 1995. The last and final version broadened the
consequence assessment scope of the model directory to incorporate fires and explosions (OFCM 1999).
The OFCM then issued a report after the 9/11 tragedy noting that while there were currently over 140
dispersion modeling systems used for regulatory, research, and emergency response, only approximately
29 non-proprietary models were used by first responders, with even fewer used operationally in quick-
response modeling facilities (OFCM 2002). The report featured these models sorted by scenario to address
end users' needs.
The U.S. DOE Emergency Management Advisory Committee and Subcommittee on Consequence
Assessment and Protective Actions also published an early report logging 93 dispersion models known to
be used within the DOE consequence assessment community (Mazzola et al. 1995*). Many of the models
on the survey list are not used frequently or are specific to individual DOE sites. In addition, another
review of existing dispersion modeling software concluded that no one system had all the features that
were deemed critical for emergency preparation and response (National Research Council 2003). Some
of the models fell short on the confidence in predicted dosages and urban and complex topography. The
1 The document entitled "Atmospheric Dispersion Modeling Resources" can be accessed online at:
fat tp ://www6 .uniovi. es/gma/adm r. pdf
4

-------
report suggested that users focus on models with short run times for response applications and more
accurate but slower models for preparedness and recovery phases.
Vardoulakis et al. (2003) compiled a comprehensive review paper of urban dispersion models
capable of use in street canyons following gaseous releases. The focus was to document the effects of
buildings within the urban street canyon and then to provide information on the 47 dispersion models
found to simulate some form of gaseous release within the urban canyon, particularly for traffic-related
emissions. Holmes and Morawska (2006) produced the first overview of dispersion models capable of
characterizing particle dispersion. The authors reviewed 18 commercial and publicly available box,
Gaussian, and Lagrangian/Eulerian dispersion models, as well as 11 aerosol dynamics models or modules,
noting that substantial differences existed between the models, and that considerable thought must be
given when selecting a model for each application. It was not possible to rank the models based on
performance or usefulness due to large model differences and the lack of particle evaluation field studies
to test the integrity of each model (Holmes and Morawska 2006).
The ADMLC working group called upon its membership of government departments, utilities, and
research organizations to develop a review of urban dispersion modeling efforts, current advances, and
future needs, including a section detailing currently available models and a review of dispersion modeling
advances from accidental releases in urban areas since the previous review in 2002 (ADMLC 2013). A
follow-on publication by Belcher et al. (2013) addressed this request, but the document does not generally
describe more than ten modeling systems or provide a useful and comprehensive table as seen in other
resources.
Most recently and within the homeland security realm, Van Leuken et al. (2016) conducted a
review of dispersion modeling studies that assessed pathogenic bioaerosols to humans and livestock. The
authors identified 16 models capable of simulating bioaerosol dispersion, provided background
information on dispersion modeling and potential bioaerosols, and developed a comprehensive table of
atmospheric pathogen dispersion studies, including the models employed. Most emergency preparedness
models only considered B. anthracis as their bioagent of focus, and all studies lacked full quantitative risk
assessments (most were simply qualitative). Hertwig et al. (2018) evaluated eight variations of
atmospheric dispersion models (including Gaussian, Lagrangian, large eddy simulations (LES), and street
network models) in mock urban areas with building obstacles. The goal was to compare results based on
the necessary balance of model speed and accuracy during emergency scenarios. The authors suggested
that the emerging, simple variations of street network models may provide accurate results comparable to
complex Lagrangian models (Hertwig et al. 2018), but the emerging models are not relatively well known
or extensively tested.
2.3 Project Goals
The objectives of this report are as follows:
1.	Introduce CBRNe terminology and illustrate how atmospheric dispersion models can simulate
these releases (Section 3.2).
2.	Define the responsibilities of government agencies and EVLAAC during a hazardous atmospheric
release (Section 3.4-3.5),
3.	Document EPA's contribution to dispersion modeling, including its history, preferred and
recommended models, role in emergency response operations and modeling efforts (Section 3.6).
4.	Provide a brief background on PBL processes that control the dispersion of hazardous releases,
including an overview of urban flow phenomenology from city structures (Section 4.0).
5

-------
5.	Introduce and discuss the advantages and disadvantages of current dispersion models including
Gaussian Plume and Puff models, Eulerian grid models, Lagrangian stochastic models, and
computational fluid dynamic (CFD) models (Section 5.1-5.7),
6.	Briefly identify dispersion model uncertainties and potential sources of error (Section 5.8).
7.	From peer-reviewed literature, technical reports, and developer websites, identify dispersion
models used by private companies, universities, and federal, state, and local agencies that are
designed for CBRNe applications. Summarize the models that are primarily used and developed
within the U.S. into a concise reference table (Section 7.0).
8.	Develop a quick reference guide (less than three pages each) for a selection of models that are
recommended for use in emergency preparation or response scenarios by expanding upon the
model's specifications, usefulness, and applicability to CBRNe releases (Section 8.0).
2.4 Literature Quality Assurance
EPA quality assurance policies and procedures were followed for this research effort. Any
literature obtained and cited within this report has been subject to a rigorous selection process. As also
outlined in the dispersion model selection process (Section 6.1), the sources of secondary data and any
cited literature in the report included peer reviewed journal articles, federal agency reports, technical
documents, model manuals, and published books (including textbooks). The topics pertained to dispersion
modeling systems and well-documented micrometeorological concepts. Additional information was
gathered through reputable websites associated with the models' developers. Due to the complexity and
lengthy history of dispersion model use, previous model applications and well-established concepts were
introduced, but emphasis was placed on recent publications, documents, and website information. Older
peer-reviewed journal articles referenced in more recent articles were also considered if deemed to contain
relevant background information for introducing the material. Some dispersion models that were
documented previously in other resources that currently could not be found with a relatively in-depth
internet search were not included in the model review and reference table (Section 8.0).
During the literature search, secondary data sources were qualitatively assessed according to the
source document type. Knowledge of the document type provided an indication of trustworthiness of the
information and secondary data contained therein, based on general professional judgment of each
document type. Each source of information and/or secondary data was also considered according to the
following categories: focus, verity, integrity, rigor, utility, clarity, soundness, uncertainty and variability,
and evaluation and review. Additionally, the literature search was limited to articles, websites, and
documents published in the English language. An emphasis was placed on models developed or used
within the U.S., although some well-known and flexible international models were featured.
Internet search criteria included lists of strategic keywords anticipated to elicit identification of
relevant secondary data and information, and the arrangement of the keywords with Boolean operators
were used to execute the searches. Boolean searches were performed using strategically selected keywords
with the operators AND and OR. After each search, the resulting identified literature was reviewed to
determine the effectiveness of the search and the relevancy of the results. Based on the search run results,
the Boolean search strategy was revised, and another run was performed. Internet searches were also run
using parentheses (" ") to ensure certain keywords were obtained.
6

-------
3.0	Emergency Response and Dispersion Modeling
3.1	Dispersion Modeling Definition
In the simplest terms, a dispersion model is a mathematical representation of the transport and
diffusion of air pollutants in the ambient atmosphere that is used to calculate effluent concentrations at
various locations away from a source (Holmes and Morawska 2006; Turner 1979). Equations governing
the dispersion of pollutants, frequently based on a Gaussian downwind concentration distribution, can be
calculated manually or through a variety of computer algorithms. Computing programs and software
permit thousands (or millions) of calculations in a short period of time, resulting in rapid estimates of
downwind concentrations that can inform policymakers, regulatory entities, researchers, or emergency
responders following the release or potential release of a hazardous substance. Most dispersion models
have limitations related to their simplified meteorology, terrain, and release assumptions, basic physics,
and parameterizations (mathematical simplification) of complex processes (Arya 1999). These
assumptions can propagate errors in the dispersion calculation, but the errors are oftentimes outweighed
when considering the computational speed and relative accuracy of their prediction.
This section introduces CBRNe terminology, describes the stages of an emergency response, and
identifies the available federal resources for emergency dispersion modeling, including IMAAC. The
section also clarifies EPA's responsibilities in prevention and mitigation, as well as roles during an
emergency response scenario. EPA is not technically a first-responding agency and generally does not
mobilize to a scene until 72 hours after the event, once state and local partners have addressed immediate
lifesaving operations. An emphasis is placed on EPA's role in hazardous release mitigation strategies and
risk evaluation, as well as the recommended models developed by the agency. This section satisfies
objectives 1 through 3 as described in Section 2.3.
3.2	CBRNe Terminology
Dispersion modeling provides significant insight to understand the fate and transport of CBRNe
(pronounced "see-burn-e'\ or simply CBRN or CBR) releases, as these situations pose significant
environmental and human exposure risks (Schmidtgoessling 2009). CBRNe's (chemical, biological,
radiological, nuclear, and explosive's) modern etymology is adapted from the Cold War acronyms ABC
(Atomic, Biological, Chemical) and NBC (Nuclear, Biological, Chemical), which were used to describe
agents intentionally (or accidentally) released that inflict harm (Hendricks and Hall 2007a). These agents
are sometimes referred to as weapons of mass destruction (WMDs), but warfighting, emergency response,
and scientific professionals use the CBRN or CBRNe identification to better characterize the release
agents. CBRNe releases are often associated with terrorist-related events intended to inflict mass
casualties and/or cause major infrastructural and systematic disruptions. However, most CBRNe releases
are inadvertent and typically the result of poor maintenance or structural upkeep, vehicular accidents, or
human error. Oftentimes, hazardous substances such as toxic industrial chemicals (TICs) are transported
near or through cities and stored close to inhabited locations (Brown 2014), where they have the potential
to enter the environment accidentally or intentionally. The following subsections define terminology used
to describe a broad overview of CBRNe releases to set the stage for their application within dispersion
models.
3.2.1 Chemical
Gaseous chemical releases may refer to various types of chemical weapons or TICs such as nerve,
choking, and blister agents that can incapacitate an individual. Certain pesticides, mustard gas, sarin, and
7

-------
chlorine are examples of chemical agents. In high concentrations, these chemicals can kill or directly harm
their target. The release of sarin gas (extremely toxic at low concentrations) in Nazi Germany was an
instance of an intentional chemical release. The 1984 Bhopal disaster in India is often considered one of
the world's worst unintentional industrial chemical-related disasters. Over a half million people were
exposed to deadly concentrations of methyl isocyanate gas used in the production of carbamate pesticides.
The incident led to almost 4,000 immediate deaths with several thousands more dying from complications
(Broughton 2005).
3.2.2 Biological
The intention of biological warfare is to release pathogens such as bacteria, viruses, or toxins so
that a person contracting the agent will have adverse health effects. This release can be achieved through
the poisoning of fomites (an inanimate object such as clothing or utensils) or food or water supplies with
infectious materials (Hendricks and Hall 2007b). One of the best-known intentional biological threats was
the release of weaponized particles of Bacillus anthracis (anthrax) spores in the United States mail stream
in 2001, commonly referred to as Amerithrax. The Yersinia pestis bacteria or Black Death during the
1300s, which killed over a third of the European population, can be considered an example of an
unintentional biological event (Perry and Fetherston 1997). The pandemic was spread by fleas carried by
rodents that resided among the population. The current COVID-19 pandemic may also fit this realm, as
well.
3.2.3 Radiological
A radiological release combines radioactive material with explosives but without the detonation
of a nuclear device. This type of release is generally achieved in the form of improvised explosives
containing radioactive materials such as a dirty bomb, also known as a radiological dispersal device (RDD)
containing an agent like 137Cs and 137cesium chloride (137CsCl). The objective of an RDD is not necessarily
to inflict mass casualties, but to cause widespread structural and systematic disruption that is costly to
repair and decontaminate, often akin to a "weapon of mass disruption" (USNRC 2018). The blast or initial
release is likely to cause more psychological than physical harm, as levels of radiation from the RDD are
not likely to be high enough to cause illness or death, especially far from the blast zone. Most of these
events, such as the foiled attempts of Chechen terrorists who tried to explode an improvised 137Cs RDD
in a park in Moscow in 1995 are intentional (Stewart 2014). An RDD has not been successfully detonated
by a militant group thus far. Accidental radiological spills may occur at laboratories and hospitals that use
radioactive chemicals, especially during radiation therapy, but are not expected to precipitate wide-area
incidents.
3.2.4 Nuclear
Nuclear releases involve accidents at nuclear power plants, the detonation of a nuclear device, or
a weapon caused by an explosion due to nuclear fission, where atoms undergo unstoppable division. A
nuclear weapon releases an incredible amount of energy over a short period of time, immediately killing
individuals close to the blast site and sickening others through radiation poisoning. Although many
intentional tests have been carried out, a nuclear device has only been detonated twice as a WMD, i.e., the
U.S.-led World War II bombs dropped on Hiroshima and Nagasaki, Japan. Some of the better-known
nuclear accidents occurred in 1986 at Chernobyl, 1979 at Three Mile Island, and, more recently, in 2011
at the Fukushima Daiichi Nuclear Plant, all of which happened at nuclear power plants. Nuclear releases
8

-------
pose some of the most effective measures for inflicting mass casualties and are the subject of numerous
emergency planning measures throughout the world.
3.2.5 Explosive
Improvised Explosive Devices (IEDs) are explosive weapons carrying conventional, non-
radioactive materials with the primary intention of inflicting harm on a subject. Commonly used as a
warfighting or suicide bombing tactic, IEDs are sometimes deployed by terrorists to injure soldiers and
other individuals. The 2013 Boston Marathon bombing and the 1995 Oklahoma City bombing are primary
examples of intentional explosive events. Various accidental explosions occur occasionally, which may
lead to homeland security concerns.
3.3 Stages of an Emergency Response
Most disasters occur at the local level, requiring municipalities to be prepared for a wide range of
scenarios. If a local jurisdiction does not have the proper resources to respond to a disaster, state or federal
government assistance may be required. These efforts must be escalated by a state's governor, who then
applies to the President for federal relief. Under the Federal Response Plan (FRP), the Federal Emergency
Management Agency (FEMA) coordinates and activates the response effort with collaboration from the
appropriate federal agencies. Regardless of the degree of emergency, FEMA states that emergency
management is generally broken down into four phases: 1) mitigation, 2) preparedness, 3) response, and
4) recovery, although adherence to the four stages is not a state-level requirement for grant funding
(FEMA 2010). Mitigation and preparedness occur before or in anticipation of a release scenario while
response and recovery ensue during or after the event.
While every emergency response effort is somewhat different, a similar structure for developing
and maintaining plans for emergency operations should be kept standard. The federal government uses a
five-category emergency response framework approach that differs slightly from the standard emergency
management process. In 2011, the Presidential Policy Directive 8: National Preparedness (PPD-8)
replaced the Homeland Security Presidential Directive 8 (HSPD-8) and was intended to be a federal level
guide on how the nation can prevent, respond to, and recover from homeland security threats (Lindsay
2012). The HSPD-8 establishes five emergency management frameworks that are intended to assign roles
to various federal agencies fitting mission-specific areas (Lindsay 2012). The five frameworks are: 1)
prevention, 2) protection, 3) mitigation, 4) response, and 5) recovery, and are expanded in this section.
This report uses the five frameworks when referring to stages of an emergency response. Dispersion
modeling may therefore occur at any stage of the response and provide vital decision-making guidance
caused by the effluent release.
3.3.1 Prevention Framework
The key to preventing an emergency scenario from occurring in the first place or to minimize its
disastrous effects is to practice mitigation activities (FEMA 1998). In the general emergency management
sense, these activities include taking actions to reduce the chance of the impact of an emergency on human
life, property, and the environment, including short- and long-term exposure effects. These activities
ensure individuals and authorities are trained and prepared to handle an emergency before it happens,
which includes establishing evacuation plans, stocking up on food and supplies, and planning how to
respond and rescue lives (FEMA 1998). The National Prevention Framework assigns roles and
responsibilities to federal agencies to help prevent imminent terrorist threats (Brown 2011). It helps
coordinate information sharing and intelligence among agencies and assists in detecting terrorist threats
9

-------
before they occur. The DOD and DHS have many specific roles related to the surveillance, prevention,
and detection of potential terrorism and WMDs.
3.3.2	Protection Framework
The National Protection Framework provides guidance on how to secure the country against
homeland security threats from acts of terrorism or natural disasters (Brown 2011). This would include
the defense against WMD threats, critical infrastructure protection (including transportation, utilities, and
agriculture), border security, and cybersecurity (Brown 2011). This framework relies on the coordination
of existing capabilities to protect the homeland.
3.3.3	Mitigation Framework
The National Mitigation Framework presents a risk management strategy to reduce the loss of life,
property, and impacts following natural or manmade disasters (Lindsay 2012). Since mitigation exists at
all levels of the emergency response process, and most notably at the local level, some examples of
mitigation efforts might involve: the procurement of insurance policies to mitigate financial impacts,
retrofitting building structures to withstand severe weather or external conditions and making informed
decisions on where to build or how to design structures (FEMA 2010). The federal National Mitigation
Framework establishes large scale risk reduction strategies, initiatives to improve homeland resiliency,
and efforts to reduce future risks (Brown 2011).
3.3.4	Response Framework
Immediately following the manmade or natural disaster, the response stage puts any preparedness
plans into place and encompasses actions that are taken to save and sustain lives, reduce the loss of
property, and support critical infrastructure after the incident has occurred (FEMA 2010). The National
Response Framework provides a foundational guide informing how the country will respond to all types
of disasters and emergencies by initiating the flexible National Incident Command System (ICS) and then
aligning roles to various federal agencies (Brown 2011). If federal emergency response support is
approved through a presidential order, FEMA will organize the response through its partner government
agencies through the FRP and deploy individuals following ICS. This framework closely mimics the
general emergency response phase, but below the federal level, the immediate response will involve the
deployment and mobilization of emergency first responders such as firefighters, police, and medical
services. External response support will also be activated in local, regional, or federal emergency
operations centers (EOCs) to coordinate and direct logistics for those deployed in the field.
3.3.5	Recovery Framework
After the immediate danger of the episode has passed, the National Recovery Framework
encompasses the short- and long-term efforts of rebuilding, restoring, and bringing the affected area back
to pre-disaster conditions or better (Brown 2011; FEMA 2010). Depending on the situation (such as
Hurricane Katrina in New Orleans), the recovery stage could take years, or the affected area may never
be fully remediated. Recovery may also happen concurrently with response efforts. Specific efforts
include rebuilding critical infrastructure, providing housing to survivors, restoring community services,
and promoting economic development (Brown 2011). During a federal emergency response activation,
EPA may be deployed for certain spills or releases that pose risks to human health and the environment
10

-------
(more details are presented in Section 3.6). EPA is generally involved with the recovery and remediation
phases of an emergency response since EPA is not a first-responding agency.
Three important components within the remediation role of consequence management are:
sampling, decontamination, and waste management (Mikelonis et al. 2018). These areas are typically
where the remediation methodologies developed within EPA's HSRP are implemented. Sampling is first
conducted by air monitors or surface-based methods through vacuuming or swabbing (Calfee et al. 2013),
which is time-consuming and expensive. Predictive modeling of atmospheric dispersion or stormwater
runoff (Mikelonis et al. 2018) may be used to inform sampling locations. Decontamination methods such
as spraying, fogging, or washdowns are then employed to mitigate the effects of the hazardous material
(Ryan et al. 2010). Decontamination approaches can lead to large amounts of contaminated byproducts
from supplies and personal protective equipment (PPE), along with the waste generated from the actual
disaster. Effective waste management is the final critical component in the recovery phase to minimize
contaminant exposure and remediate the affected area (Boe et al. 2013).
3.4 Operational Dispersion Modeling and Reach Back
Historically, the first dispersion models were developed to design control strategies for air
pollutants released from industrial exhaust stacks (Beychok 1979). The development led to the
implementation of dispersion models in the federal government for regulatory use in new stack
construction or compliance with air quality standards such as the National Ambient Air Quality Standards
(NAAQS). However, dispersion modeling was also found to be a useful, multifaceted scientific tool to
inform public officials during accidental or intentional environmental releases, particularly due to their
fast run times. Dispersion modeling for both regulatory and emergency response applications are part of
the analyses and planning required by the Clean Air Act (CAA).
During times of crisis, first responders (local firefighters and police) are usually the first on the
scene, typically within fifteen minutes of notification. Guidance information for evacuation and sampling
is often needed within thirty minutes of a release, when containment is most critical (van de Walle and
Turoff 2008), while other key decisions are generally made within one hour. To take advantage of
dispersion model results, this timeframe would require emergency responders to 1) have the necessary
knowledge to run a dispersion model, and to 2) know most of the physical details of the release and
atmospheric parameters, as well as the model domain and boundary conditions. Local meteorology and
release-source strength (mass of the release per unit time) are some of the most important components to
initialize a dispersion model (OFCM 2002), in addition to secondary sources caused by the trapping and
re-release of material in the wake of obstructions, leading to critical variations in contaminant
concentration over short distances (Coceal et al. 2014). These variables are usually not known immediately
and could introduce significant bias into the modeling results.
Since emergency responders generally do not have time to setup, run, and process results from a
dispersion model, operational modeling options are available for large-scale and potentially high-impact
situations when time is critical, and a modeled plume cannot feasibly be generated without external
assistance. These operational modeling services are dubbed as "reach back" support, which is known in
the military as the opportunity to reach outside a unit's traditional information flow to obtain additional
intelligence and remain well-informed on specific matters (Radzikowski 2008). This section explains
those choices, as well as federal reach back responsibilities and then identifies EPA's connection to
dispersion modeling for emergency preparation and response.
3.4.1 Interagency Modeling and Atmospheric Assessment Center (IMAA C)
11

-------
To provide a single point for the coordination and dissemination of hazard prediction products
during an actual or potential incident involving federal government coordination, DHS established
IMAAC as part of the National Response Framework (NRF; U.S. Department of Homeland Security
2016) to prepare and provide a more unified Federal response to disasters and emergencies. Currently led
by FEMA, IMAAC is a collaboration between seven federal agencies, including DHS, DOE, DOD,
NOAA, EPA, Nuclear Regulatory Commission (NRC), and the U.S. Department of Health and Human
Services (DHHS), each with its own responsibilities for atmospheric plume modeling or a support role.
IMAAC was formally recognized in April 2004 through a memorandum of understanding (MOU) through
the Homeland Security Council Deputies Committee of DHS and is governed under the Homeland
Security Act of 2002 for DHS to respond to and prepare for natural and manmade crises. FEMA assumes
a key role in IMAAC under the Post-Katrina Emergency Response Management Reform Act of 2006 to
prepare, plan, respond, recover, and mitigate effects to key infrastructure and resources following
catastrophic incidents.
One of the primary reasons for IMAAC's creation was the inability of some federal agencies to
properly coordinate dispersion modeling efforts. A noteworthy example occurred during the 1999
National Top Officials Exercise (TOPOFF) when various modeling systems produced confusing and
contradictory results (US GAO 2008). TOPOFF2, a five-day full-scale simulated exercise, was intended
to realistically test federal, state, and local emergency response systems should a high-profile CBRNe
situation occur. During the exercise, dispersion models produced conflicting results because actual and
mock meteorological inputs were used by different modelers without prior discussion. The resulting
dispersion plumes, which were meant to inform emergency responders and the general public of
potentially affected areas following a hypothetical release scenario, had the plumes impacting two
different areas: one over the Pacific Ocean and another over the city of Seattle. The dispersion plumes led
to an embarrassing appearance by the Washington State governor on television and ended with blame
being placed on top officials for the misunderstanding (Mongeon 2018). The goal of IMAAC was
therefore to provide a single official government dispersion prediction using the best possible government
resources to lessen the chances of a repeated embarrassing scenario and to deliver the information more
efficiently to first responders.
When requested by any state, local, federal, or tribal agency, IMAAC will organize a rapid, around
the clock Federal response to produce a simulated atmospheric dispersion model plume. These products
can be used by first responders to make informed decisions following an actual or potential hazardous
release scenario. Plumes can be requested by email or through the Homeland Security Information
Network (HSIN) helpdesk (phone number: (703) 767-2003) under significant real-world emergencies.
IMAAC can also provide dispersion plumes for planning scenarios or national exercises, but a planning
request will not take priority if another situation requires immediate attention.
IMAAC was also developed to support emergency responders in field response efforts. Following
a hazardous release, the three primary questions from emergency personnel typically are: 1) when and
where the greatest impacts could occur, 2) which areas are confidently out of danger from exposure, and
3) how long the response and remediation efforts could take. IMAAC results can be disseminated to
emergency responders in as little as 20 minutes after the request with simple information such as where,
when, and what was released. This preliminary information can provide the initial best guess needed to
identify regions in the hot zone. The first model predictions tend to overestimate the actual event because
details of the source characteristics are rarely known in emergency situations (OFCM 2002). The model
results are then refined as additional information is reported from the scene.
Another justification for IMAAC was to reduce confusion over who should perform modeling in
emergency situations and which models should be chosen. While IMAAC was not intended to replace
agency-specific dispersion modeling activities, no permanent IMAAC dispersion provider has been
12

-------
officially identified. Instead, IMAAC fulfills its mission by providing a variety of dispersion resources
through reach back support, where agencies or centers with specific scientific expertise (e.g., chemical,
nuclear, biological) become activated and assume operational responsibility (Dadosky 2010). These
agencies establish a baseline set of dispersion models that are regularly implemented and improved
through research and development initiatives. NARAC, operated out of DOE's Lawrence Livermore
National Laboratory (LLNL), has been designated as the primary IMAAC emergency operational hub
with round-the-clock staff support, modelers, nuclear experts, and radiological monitoring teams.
Currently, NARAC is the default reach back option for radiological and nuclear events but may also
provide airborne hazard predictions for chemical and biological releases (Nasstrom et al. 2007). The
Defense Threat Reduction Agency (DTRA), the DOD's official Combat Support Agency (CSA) for
countering WMDs, is the official technical reach back for biological and chemical agent releases.
3.4.2 Models Used in IMAAC Responses
NARAC employs a comprehensive suite of proprietary modeling systems that integrate multiple
LLNL models during a radiological or nuclear incident. Source-term models are fed into rapid effects-
processing models or NARAC's own three-dimensional (3D) atmospheric dispersion models. For most
cases, NARAC will use the Atmospheric Data Assimilation and Parameterization Tool (ADAPT) model
to construct 3D meteorology fields for use in the dispersion model, Lagrangian Operational Dispersion
Integrator (LODI) (Nasstrom et al. 2007). NARAC simulates the input meteorology for ADAPT's initial
conditions through basic parameters obtained from the National Weather Service (NWS), or if there is an
ongoing atmospheric release that is expected to continue for an extensive period of time, a research-grade
mesoscale weather prediction model called the Weather Research and Forecasting (WRF) Model may be
used (Nasstrom et al. 2007). Through a Lagrangian stochastic Monte Carlo approach (which calculates an
average based on a nearly Gaussian distribution of atmospheric turbulence), LODI then solves the 3D
advection-diffusion equation to produce a rapid and detailed plume within 5-15 minutes.
IMAAC reach back through DTRA uses the Hazard Prediction and Assessment Capability
(HP AC) dispersion model suite. The main dispersion code for HP AC is built upon the Lagrangian Second-
order Closure Integrated Puff (SCIPUFF) model but has many advanced capabilities, including
atmospheric transport and dispersion calculations, urban parameterizations, deposition, dose, and human
effects-hazards. Real-time meteorology from the NWS is automatically pulled in through DTRA's
meteorological data servers, which also host worldwide numerical weather predication (NWP) products
from climate reanalysis data such as the National Centers for Environmental Prediction (NCEP) or the
North American Model (NAM). These products can also be used to initiate WRF. HPAC's stochastic,
second-order closure Puff model ensures that computations take only a few minutes, and an initial
response can be sent to the requestor within 20 minutes.
Other advanced dispersion models such as Los Alamos National Laboratory's (LANL's) Quick
Urban Industrial Complex (QUIC) model (Nelson and Brown 2013) are not part of IMAAC since QUIC
lacks the integration of the SCIPUFF dispersion model. If EPA or NOAA is activated based on release
type (such as inland oil spills), either agency may use the relatively simple co-developed
CAMEO/ALOHA (Computer-Aided Management of Emergency Operations/Areal Locations of
Hazardous Atmospheres) model (see Section 8.6), which is already used by many emergency responders.
Dose projections for atmospheric radiological releases could be calculated by NRC's Radiological
Assessment System for Consequence Analysis (RASCAL) model. The progression of radiological
releases in nuclear reactors may be calculated through the MELCOR Accident Consequence Code System
(MACCS) code. A summary of IMAAC agency responsibilities and models typically used for each release
type is shown in Figure 1. More detailed information on each of these models will be presented later in
this report.
13

-------
• DTRA (DOD)
O
Model Support:
HP AC

-------
and expedite remediation. The RMP rule considers the use, storage, manufacturing, handling, or
movement of an extremely hazardous substance at a stationary source site based on an extensive list of
regulated chemicals (EPA 2009a). A facility's RMP must include several elements based on the type of
program or processes that occur at the site. Generally, all facilities must include a description of the site
and its regulated substances or processes that occur, the five-year accident history, the hazard assessment
for the process, the potential worst-case scenario, and any site-specific emergency response programs
should there be an environmental release (EPA 2009b).
A key element of an RMP is completion of a hazard assessment of the potential impacts from a
release at the facility - effectively called an offsite consequence analysis. EPA provides guidance for this
consequence analysis in EPA (2009c), which requires two elements: a worst-case scenario release and an
alternative release scenario (i.e., the effect of a hypothetical accident under more realistic circumstances)
(EPA 2004). The EPA guidance document offers several methods to carry out this analysis, including the
use of dispersion models. However, the guidance is optional if the methodology or models can be
substantiated (EPA 2004). The simplest guidance relies on lookup tables to provide conservative estimates
of downwind risk and does not require computer modeling. More accurate site-specific consequences
could be generated through dispersion modeling results. In this capacity, facilities have the option to
choose their own dispersion model, fire or explosion model, EPA-established model, or another
computational method (EPA 2009c). EPA's RMP*Comp dispersion modeling tool is one option that can
be employed and is discussed in more detail in Section 3.6.2,
3.5.3 Bio Watch
The mechanism, location, and timing of a chemical, radiological/nuclear, or explosive release may
be possible to pinpoint (and thus remediate, model, evacuate residents, and decontaminate). However,
biological vectors are often more challenging. Bacteria, spores, and other biological agents may be
passively dispersed through the air or from person to person. During the 2001 Amerithrax events, it was
unclear when, where, and how any further biological releases could occur. As a result, the DHS launched
the BioWatch program in 2003 with the primary goal of detecting the presence of bioterrorism agents in
large, densely populated urban areas. Deployed as a series of samplers, typically adjacent to the EPA's air
quality monitors or in high human-traffic indoor and outdoor locations, the BioWatch program is the
nation's first early warning system to detect certain known biological threats (NAS 2018). While the
premise of the program is beneficial, it has received criticism for its monitoring methodology and high
false-alarm rate. The BioWatch program does not operate any real-time samplers and therefore has
minimal value for first responders seeking to obtain real-time information on biological pathogens (US
GAO 2008). The only way to determine the presence of a harmful agent is to analyze the filter sample by
completing a full laboratory analysis, which may take 24 hours or more after the sample is collected (NAS
2018). If a sample tests positive for any of a suite of biological agents, a BioWatch Actionable Result
(BAR) is created and can trigger federal, state, and local response through the means of teleconferences
and potentially the activation of a consequence management plan (e.g., sampling, public communication,
environmental surveillance, event reconstruction) (NAS 2018). More than 50 BARs have been generated
since the program's inception, but all have been false alarms due to naturally occurring organisms in the
environment (NAS 2018). These BARs have led to mixed response and criticism for the program.
Even though LANL is not part of IMAAC, an around-the-clock modeling team may be activated
directly by the BioWatch program if a noteworthy BAR is detected. LANL's BioWatch Event
Reconstruction Tool (BERT) is run to simulate source inversion (determining a source location and
strength based on dispersion modeling) or forward plume calculations. The QUIC dispersion model has
also been employed to simulate biological releases for special events where BioWatch detectors are
deployed. The future goal is to link QUIC with Argonne National Laboratory's (ANL's) Below Ground
15

-------
Model so below- and above-ground biological-agent transport and dispersion can be captured and
simulated (Michael Brown, personal communication, 2018).
3.5.4 Chemical Facility Antiterrorism Standards Program
The nature of certain facilities such as industrial chemical plants producing and using hazardous
materials, power plants, or storage facilities makes them high profile targets for potential sabotage or
accidental releases. To identify and help prevent and mitigate chemical releases from these types of high-
risk facilities, the DHS has developed the Chemical Facility Antiterrorism Standards (CFATS) program.
The purpose of CFATS is to identify and regulate high risk facilities to ensure that security measures and
contingency plans are in place to reduce the effects of a terrorist-related attack associated with TICS. More
than 300 chemicals, manufactured at plants that produce certain quantities, have been identified and must
be regulated by CFATS. Based on the level of potential impact, CFATS focuses its efforts on the highest
risk facilities and then employs a risk-based tiering system (6 CFR, Chapter I, Part 27). For example, any
site that possesses more than 10,000 pounds of anhydrous ammonia must comply by reporting their
holdings to the DHS CFATS program. The importance of the CFATS program for the interests of this
report is that established, physics-based dispersion models may be used as part of the risk analysis
methodology instead of employing EPA's RMP*Comp dispersion tool (81 CFRNo. 71). Potential release
scenarios could be run by IMAAC for planning initiatives, although DHS does not disclose what types of
dispersion models might be employed. The CFATS program is important for EPA emergency response
efforts in that it identifies these high-risk facilities, tiers their potential impact, and provides an inventory
on the amount of hazardous TIC present should an incident occur.
3.6 EPA's Contributions to Emergency Response Initiatives and Dispersion Modeling
A wide-area CBRNe incident undoubtedly causes large-scale human safety, environmental,
infrastructure, and economic concerns. If a local or state jurisdiction does not have the appropriate
resources needed to respond to an emergency, federal emergency response may be required. In this
capacity, FEMA is responsible for coordination and activation of the FRP. Response to wide-area
incidents requires coordinated, multiagency approaches. Under the Homeland Security Presidential
Directive 10 (HSPD-10), DHS coordinates with the appropriate federal agencies to respond to homeland
security incidents. Before, during, and following natural disasters, terrorist attacks, or accidental
emergencies, the federal government may enact a Continuity of Operations Plan (COOP) to sustain
operational order. While the federal government has a series of eight National Essential Functions (NEFs)
to continue leading the country during times of hardship, each agency is responsible for its Primary
Mission Essential Functions (PMEFs). The EPA has one PMEF: to prevent, limit, and/or contain chemical,
radiological, biological, oil, and other hazardous contamination incidents and provide environmental
monitoring, assessment, and reporting for the incident. As a result, the EPA may act as the primary agency
responsible for organizing the event with on-scene actions and local governments, or in some capacity to
be the lead federal agency (LFA). These responsibilities are outlined under the DHS National Response
Framework (U.S. Homeland Security 2016). In most cases, the EPA is LFA when the response is related
to the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA: Superfund),
EPCRA (planning for chemical emergencies), the Clean Water Act, the Safe Drinking Water Act, the Oil
Pollution Act, the CAA, and the Resource Conservation and Recovery Act (RCRA: hazardous and general
solid waste disposal). Additional responsibilities involve the cleanup of impacted buildings or natural
areas, recovery from terrorist attacks or natural disasters, and support to drinking water systems.
For initial radiological and monitoring assessments, the DOE is generally considered the LFA, but
EPA would assume intermediate and long-term response once the initial threat has subsided (U.S.
16

-------
Homeland Security 2016). EPA would also be the LFA with accidents involving shipments of radiological
materials not licensed or owned by a federal agency, as dictated by the Federal Radiological Emergency
Response Plan (FRERP). When EPA is the LFA, on-scene coordinators (OSCs) assume the position of
the lead federal officials. At the site level, EPA's OSC personnel assess, monitor, and control the response
with the incident command and employ research developed by scientists in the agency (Mikelonis et al.
2018) to streamline the response process.
Even though EPA is not technically a first-responding agency as opposed to local or regional
police, fire, and emergency medical service (EMS), dispersion modeling may play an important role in
agency actions after EPA assumes responsibility at least 72 hours after the release. Various tools and
strategies such as predictive computer modeling can be employed to protect human health and the
environment and respond to wide area incidents (Mikelonis et al. 2018). These plume predictions could
help guide field sampling efforts, and the sampling results could then be used to update the plume and
refine estimates. The dispersion models developed, used, and improved by the EPA are expanded in the
following sections.
3.6.1	EPA Support Center for Regulatory Atmospheric Modeling
EPA implements dispersion and predictive modeling tools for regulatory applications, emergency
preparation and response, and research and development purposes. EPA's Support Center for Regulatory
Atmospheric Modeling (SCRAM) website provides public access to air quality and dispersion models and
resources developed by the agency. SCRAM also delivers training and resources, reports and journal
articles, and other modeling guidance. The EPA has also established a list of agency-preferred and
recommended models for screening purposes, state implementation plans (SIPs), and downwind
calculations from source to receptor for regulatory and permitting use (Appendix W 40 CFR part 51). The
models identified as preferred agency options are assessed for quality assurance so that criteria are met
for scientific rigor, model development and evaluation, peer reviewed theory, and the ability to provide
transparent, reliably disseminated information (www.epa.gov/quality). Some of EPA's preferred and
recommended dispersion models, all of which are Gaussian plume models, are AERMOD (American
Meteorological Society/Environmental Protection Agency Regulatory Model; Cimorelli et al. 2005; see
Section 8.5), CTDMPLUS (Complex Terrain Dispersion Model Plus Algorithms for Unstable Situations),
and OCD (Offshore and Coastal Dispersion Model). These models are identified later in this report, but
only AERMOD is expanded in further detail since these models are not generally used for emergency
response. EPA also suggests alternative and screening models on the SCRAM website for other
applications.
3.6.2	RMP*Comp
As part of the EPA's RMP rule, the agency provides resources to chemical facilities so they can
develop their site-specific consequence analyses. The facilities may employ their own public or
proprietary dispersion model of choice if they are willing to share the results, the model is recognized
within the industry, the model is applicable for the chemical release being simulated, and it defines the
appropriate parameters and worst-case scenarios (EPA 2009c). However, EPA also provides a free, web
and desktop computer-based RMP*Comp dispersion modeling tool, which was co-developed by EPA and
NOAA. The program incorporates database tables of regulated materials including 77 acutely toxic
substances and 63 flammable gases and volatile liquids. Based on the amount released, RMP*Comp
determines the downwind distances to an endpoint location for probable and worst-case scenario events
for standard atmospheric temperatures (25°C), low winds (1.5 m/s), and stable conditions (stability class
F). The program can handle vapor cloud fires for flammable gases that are liquefied under pressure (EPA
17

-------
2009c). However, RMP*Comp is not a model to be used for emergency response as it only allows the user
to select the amount released, liquid temperature or chemical physical state, and sometimes the
surrounding terrain type. Moderate ambient air temperatures, low-to-moderate wind speeds, and stable
atmospheric conditions are assumed during the planning scenario, which does not always reflect the
atmospheric state during an actual release, potentially lofting the chemical or allowing it to travel farther
downwind.
3.6.3 CAMEO/ALOHA
RMP*Comp is meant to easily plan and identify high-priority hazards for an RMP, but more
sophisticated co-EPA-developed modeling tools are available for emergency response use such as the
CAMEO Software Suite. CAMEO (also see Section 8.6), which includes four distinct entities: 1) CAMEO
Chemicals, 2) CAMEO/tw, 3) ALOHA, and 4) MARPLOT. CAMEO Chemicals is a comprehensive,
proprietary database of hazardous chemical datasheets and physical properties, which provides
information similar to the information seen in the classic orange US Department of Transportation (DOT)
Emergency Response Guidebook (ERG). The ERG is still a go-to resource for most first responders, as it
provides basic guidance to determine the extent of a and its approximate evacuation distances (Christine
Wagner, personal communication, 2018), as well as chemical-specific hazards. ALOHA, CAMEO's
simple Gaussian plume model, can be used for on-scene chemical releases since results are generated
within seconds from only a few details about the release and current meteorology. ALOHA was first
developed by EPA and NOAA in the late 1980s specifically for use by first responders, including EPA's
Environmental Response Team (ERT). The plotting software in CAMEO is MARPLOT. The entire
CAMEO software package is available as a free download for laptops and mobile phones from NOAA's
Office of Response and Restoration.
ALOHA is identified in this section because it can also be used to perform RMP guidance. Since
ALOHA performs more sophisticated dispersion analyses due to specific and refined input parameters, it
may provide results that do not closely match RMP*Comp. However, ALOHA cannot be used to inform
EPCRA hazard analyses because the quick and simple calculation methodologies in EPA et al. (1987)
(i.e., "The Green Book") are used instead. The Green Book capabilities are available in the CAMEO/tw,
which is used to manage planning data such as details about a particular facility, chemical transportation
routes, and emergency response procedures about chemicals in a local community. The Green Book may
also provide results that do not closely match ALOHA due to its simplifying assumptions. While also
simplified, ALOHA can account for different atmospheric stabilities, dispersion parameters based on
terrain, temperatures, liquid evaporation rates, and buoyant or dense gases (Jones et al. 2013).
18

-------
4.0	Atmospheric and Micrometeorological Fundamentals
in Dispersion Modeling
The purpose of this section is to outline fundamental components of boundary layer meteorology
and micrometeorology (meteorological processes on a spatial scale of-1-10 km and a temporal scale of
~ 1 hour to 1 day) that dictate the transport and dispersion of a hazardous release. This section is not meant
to be a comprehensive explanation of PBL processes, but more of a resource to introduce salient concepts
employed within dispersion models. More in-depth explanations of micrometeorology and its associated
processes are covered in Stull (1988) and Arya (2001; 1999). Dispersion models simplify the complex
atmospheric state by using equations that govern the dispersion of pollutants under the assumptions of
stationarity and horizontal homogeneity, frequently based on Gaussian downwind concentration patterns.
However, the atmosphere is very turbulent, especially close to the ground surface where roughness
elements (e.g., trees, buildings, vegetation, topography) create complex and variable flows, and physical
variables (due to solar heating, moisture, and the overall large-scale atmospheric state) are in constant
flux. Since the fate and transport of contaminants is significantly influenced by turbulent exchange in the
PBL, the fundamental concepts are introduced here. This section satisfies Objective 4 as described in
Section 2.3.
4.1	Atmospheric Turbulence
The dispersion (transport and diffusion) of atmospheric contaminants is strongly influenced by
microscale physics. While pollutant transport is primarily a function of the mean wind, small-scale
atmospheric turbulence is the fundamental driver of plume dispersion. According to Arya (2001),
atmospheric turbulence is defined as the highly irregular, random, and almost unpredictable chaotic
fluctuations of wind velocity, temperature, and scalar concentrations around their mean values. The
irregular fluctuations in a turbulent flow are functions of time at fixed points in space. Turbulence occurs
in the PBL where the earth's surface strongly influences small scale motion, temperature, water vapor,
and pollutants. Turbulent flow constantly undergoes random changes in both magnitude and direction,
visualized as irregular vortex-like swirls of motion called eddies. These vortices are not clearly defined
structures or features, but more of a concept to qualitatively describe turbulence on the order of 0.001 m
to 1000 m in diameter in the PBL (Stull 1988). Turbulence consists of many different eddies superimposed
on each other where the strengths of eddies of various scales define the turbulence spectrum. However,
turbulence is not as easy to precisely define, as certain wave motions in the atmosphere may be irregular
and chaotic but not necessarily turbulent.
Useful rules of thumb to describe atmospheric turbulence can be defined by the five following
criteria provided by Arya (2001) where the flow is:
1.	Irregular and random: These motions make turbulence nearly unpredictable and irreproducible,
meaning that a statistical description of turbulence must be employed (i.e., wind fluxes are
described by means, variances, fluxes, etc.).
2.	3D and rotational. Three-dimensional velocity fields and the presence of vorticity or rotation that
are highly variable in space and time.
3.	Ability to mix: turbulent diffusivity is responsible for the dispersion and spread of pollutants in the
PBL and is also effective at exchanging momentum and heat. This is often regarded as the most
important property of turbulence.
4.	Ability to dissipate: turbulent motion is continuously dissipated and turned into heat or internal
energy by viscosity, meaning that turbulence will decay if it is not produced continuously.
19

-------
5. Multiple scales of motion: turbulent flows contain a wide range of scales, and the ability to transfer
energy from one scale to the other is important and key to larger atmospheric processes.
Turbulent eddies create fluctuations in wind velocity, temperature, humidity, and scalar
concentrations causing their components to vary irregularly in time or space around their mean
quantities. This can be thought of as a fluctuating component superimposed on the mean quantity. The
basis of Reynolds averaging (Reynolds 1894) for wind describes the wind's components (w, v, and w
for longitudinal, lateral, and vertical, respectively) consisting of their characteristic mean (u) and
fluctuating (u') portions, called Reynolds decomposition. Many meteorological applications use
average wind over time, but in dispersion modeling, the fluctuating components are key considerations
that point to the PBL's vertical structure (Pasquill and Smith 1983). Simple Gaussian dispersion
models may only require the wind speed and direction averaged over a longer time period. However,
more complex models that track and simulate a puff or particle over numerous time iterations may
benefit from the wind's fine-scale fluctuations.
4.2 Planetary Boundary Layer
Dispersion models are commonly used for multiple simulations over short durations (on the order
of one-hour averages) and relatively small areas of spatial interest (< 20-50 km) in the atmospheric region
closest to the earth's surface called the PBL. This 0.2-3 km vertical layer of atmosphere connects larger
scale weather patterns (mesoscale: ~10s-100s of km, and synoptic: >1000 km) to those driven by surface-
related effects (Figure 2). The PBL is a dynamic and constantly changing portion of the atmosphere that
is influenced by terrain, vegetation, water bodies, heat and moisture fluxes, and human-introduced
influences such as anthropogenic emissions, urban canopies, and alterations to the ground surface. The
PBL varies diurnally based on solar heating and seasonality.
During the middle of the day, the PBL has typically grown to its thickest point due to surface
heating from the sun and the formation of buoyant thermals. This daytime layer is referred to as the
convective mixed layer (CML). The CML grows by entrainment, or the downward mixing of air from a
more stable layer above. It begins to grow just after sunrise, reaches its maximum point during the day,
and shrinks in height (collapses) by sunset. Conditions that include an unstable boundary layer and high
air flow during the daytime hours (due to increased wind speeds, buoyancy, or turbulence) usually mean
that a release will disperse faster than during the evening, although detrimental effects can be transported
farther downwind.
As the sun slowly sets, solar heating to the surface is cut off and the PBL slowly transforms into a
stable (nocturnal) boundary layer (SBL) nearest to the surface and a residual layer (RL) above. The RL is
a neutrally stratified zone where turbulence exists from the previous day's CML and tends to be equal in
intensity from all directions (Stull 1988). The SBL is characterized by stable atmospheric conditions
(where vertical motions are suppressed) with light or calm winds and minimal turbulence directly above
the surface. The lower part of the RL is transformed into the evening SBL due to its contact with the
ground and radiational cooling. A capping temperature inversion usually develops above the RL. Air in
the SBL is statically stable with the potential for intermittent weak turbulence (low-level or nocturnal jets),
making the evening boundary layer perhaps one of the most difficult to predict in dispersion modeling
due to complex transport, diffusion, and spurts of turbulence in combination with natural and
anthropogenic surface effects (OFCM 2002). The increase in atmospheric stability and a general decrease
in wind speed during the evening and overnight hours may mean that a release will disperse more slowly
than during the day but also remain concentrated near the emission source.
As the sun rises in the morning hours, solar heating once again grows turbulence and starts to mix
the two layers by entrainment and gradually builds the convective boundary layer, resulting in the
20

-------
entrainment zone extending to the surface as the SBL erodes. During all periods of the day and night, the
surface layer (SL) is generally defined as approximately the lower 10% of the PBL and is a complicated
area where turbulent fluxes, stresses, surface roughness, and other perturbations affect wind flow directly
above the ground surface (Stull 1988). While the SL is arguably one of the most important regions for
understanding pollutant dispersal, particularly at breathing levels after a release, it is also one of the most
difficult to predict, simulate, or otherwise parameterize in dispersion models. Complex PBL relationships
are often simplified when parameterized in dispersion models, which can translate into lower model
performance. Specifically, complex terrain effects, coastal influences, and urban canopies introduce fine
scale meteorological variations that complicate modeling efforts.
1000
Free Atmosphere
Capping Inversion
Entrainment Zone
Convective
Mixed Layer
Residual Layer
Convective
Mixed Layer
Surface Layer
Stable (nocturnal) boundary layer
t
Midnight
SI
S2
Local Time
S3
S4 S5 S6
Figure 2: The common, but idealized, diurnal evolution of the Planetary Boundary Layer (PBL) adapted
from Stull (1988).
The PBL is also significantly modified over urban areas due to increased roughness effects from
the presence of buildings and other man-made landscape alterations. With lower surface albedo due to
dark pavement, concrete, and lack of vegetation, as well as increased anthropogenic emissions, urban
areas frequently result in localized regions of higher temperatures called the urban heat island (UHI) effect.
This phenomenon results in higher overall temperatures both laterally and vertically around the urban area
based on the density of buildings. The urban boundary layer (UBL) creates a gradient or "dome" of locally
warmer temperatures in comparison to more rural locations, where a "cliff or a steep rise in temperature
is seen between the transition of rural to suburban locals, followed by a "plateau" over suburban portions
of the outer city, and a "peak" over the city center (Oke 1988). Diurnal changes in temperatures may be a
few degrees warmer due to the increase of incoming longwave radiation, as well as absorption and re-
emission by the polluted urban atmosphere (Arya 2001). A decease in surface albedo may cause a decrease
in outgoing longwave reflection or shortwave radiation emission, causing a greater daytime heat storage
and decreased evaporation. During clear skies and calm wind conditions, thermal modifications of the
UBL tend to dominate over roughness effects created by the buildings (Arya 2001). As the boundary layer
erodes during the evening and overnight hours, the UBL tends to decrease in altitude but remains higher
than in rural locations where strong atmospheric stability suppresses turbulent vertical mixing.
21

-------
The sections of the PBL are therefore not always homogeneous. When the approach flow is
modified by changes in roughness or temperature differences over the surface, a change in the mean wind
profile and turbulence may be seen near the surface. This modified layer is called an internal boundary
layer (IBL) because it grows within another boundary layer associated with the approach flow (Arya
2001). Notable examples occur when there is a dramatic change of surface roughness where the flow
suddenly moves between a grassy field to a stand of trees, from a water body to a shoreline (leading to
land and sea breezes), or upon entering an urban area to a rural location.
4.3 Modifications to Urban Flow from Building Structures
While the UBL influences the local mesoscale flow regime, complex wind behaviors exist within
the urban canopy. Buildings and streets alter the overall wind flow patterns and cause complicated
turbulence that could disperse harmful pollutants and affect the exposure for urban inhabitants. In addition
to the effect of industrial emissions on poor air quality, transportation-related emissions associated with
vehicles and other releases have a large effect on localized urban air pollution. Especially under low wind
conditions, buildings can restrict ventilation and keep localized concentrations of pollutants very high at
breathing level or in isolated recirculation zones.
A street flanked with buildings on either side is classically referred to as a street canyon (Nicholson
1975), although some urban geometry may contain discontinuous street canyons with intersections and
building breaks. The schematic in Figure 3 from Halitsky (1968), as described in Arya (2001), shows how
the mean velocity profile and background flow become altered and separated when the windward end of
the building is encountered. In the lower portion of the UBL, flow on the windward side of the obstacle
creates a clockwise turbulent eddy due to pockets of low pressure and reversed flow at various locations
around the building (Monbureau et al. 2018). Once the flow is forced up and over the building, high
turbulence intensity and reverse flow form a recirculating "cavity" zone on the lee side of the building
(Monbureau et al. 2018), which may bring contaminants to the street surface and lead to downwind
stagnation, especially when the mean wind speed is < 1.5 m/s (DePaul and Sheih 1986). This effect can
entrain outside pollution and/or accumulate contaminants emitted inside the cavity, leading to high
concentrations as wind speeds remain low but wind shear and turbulence intensity is high (Arya 2001).
The size of the cavity depends on the length, width, and height {Lb, Wb, H) of the building or obstacle, as
well as the characteristics of the approach flow. The mean velocity profiles are also shown in Figure 3b,
depicting the small counter-gradient flow in the cavity. The flow separation between the regions in the
cavity and farther downwind are called the near and far wake regions, respectively. The far downwind
wake is associated with enhanced turbulence intensity and negative vertical velocity as the flow begins to
recover from the encounter with the building. Wind tunnel studies have shown that the influence of
buildings and their wakes can extend to 10-20 building lengths downwind and beyond (Arya 2001). These
flow separation effects occur as the fluid flow of the air attempts to transfer from high to low pressure to
remain in equilibrium.
Figure 4 shows the influence of pollutant dispersion within an idealized symmetrical street canyon
based on the findings of Dabberdt et al. (1973). Due to pressure and turbulence effects, some of the mean
synoptically induced wind flow above the rooftops enters the street canyon and creates a localized vortex,
thereby restricting the recirculation of anything that is emitted within. However, the aspect ratio (the
average building height (H) divided by the street canyon width (Wc)) strongly influences street canyon
flow. Street canyons tend to have aspect ratios H/Wc ~ 1, while wide avenue canyons could be below 0.5,
with deep canyons > 2. Short, medium, and long length canyons can be characterized by their approximate
length (Lc) divided by height such as Lc/H ~ 3, 5, 7, respectively (Vardoulakis et al. 2003). Wide canyons
with HZWc ~ 0.3 (Figure 5a) create an isolated roughness flow that develops a cavity directly in lee of the
first upwind building and on the windward side of the second, but the center of the canyon may remain
22

-------
coupled with the mean flow. For wake interference flow when // 111 ~ 0.5 (Figure 5b), the disturbed air
does not have enough distance to modify its flow before encountering the next obstacle (Vardoulakis et
al. 2003, Oke 1988). Finally, H/Wc > 1 results in skimming flow in the street canyon (Figure 5c), allowing
the formation of a single vortex. Additional areas of low pressure and wind recirculation zones could occur
on street corners or intersections. Low wind speeds and/or deep canyons with H/Wc > 1 may create
recirculating cavities that leave the breathing level largely stagnant (Grimmond and Oke 1999). In
addition, the presence of a taller building among shorter buildings, as well as the distance and orientation
between them, can significantly modify the flow downstream (Arya 2001), as shown in wind tunnel
experiments with a tall tower among an array of shorter buildings (Heist et al. 2009).
(a)
Figure 3: Schematic of a) cavity and wake flow zones associated with a building or other square obstacle
in the mean flow and b) its relationship with the vertical wind profile, after Malitsky (1968).
, LANE .
M	W	
Figure 4: Schematic of streamlines when perpendicular flow encounters a street canyon, based on
Dabberdt et al. (1973).
23

-------
Iiuilnlnri rough nuns flow
Figure 5: Various perpendicular flows for urban canyons with a) isolated roughness, b) wake interference,
and c) skimming flow aspect ratios, based on Oke (1988).
4.4 Vertical Wind Profile
A vital input to most atmospheric dispersion models is an accurate representation of the vertical
wind velocity profile (the change in wind speed with height) just above the ground surface, as it will
dictate how fast and far a pollutant disperses. As a general principle, mean wind speed typically increases
with height because the effects of friction and surface roughness are lessened farther away from the ground
surface. Two commonly used wind profiles in dispersion models are the logarithmic wind profile and the
power law. The logarithmic wind profile or "Law of the Wall" is a semi-empirical relationship that depicts
wind speeds close to the surface under well developed, neutral boundary layer conditions:
	 w* z d	/i \
u = — In 	
where u is the wind speed at height z, u* is the friction velocity, k is the Von Karman constant (0.4), zo is
the surface roughness length, and d is the zero-plane displacement due to obstacles. Ten meters is
commonly used as the reference height to avoid surface-related effects. The logarithmic wind profile is
generally applicable in the lowest several hundred meters of the PBL (Stull 1988). A simple alternative to
the logarithmic wind profile is the power law expression:
u /z \r
uZ \zr/
(2)
where u is the mean wind speed at height r, and and zr are the reference wind speed at reference height,
respectively. The exponent m is based on the type of surface and is approximately 0.1 for smooth surfaces
and roughly 0.4 for urban areas (Arya 1999). The logarithmic wind profile (1) is commonly used in many
dispersion models and is considered more effective than the power-law wind profile, and (2) is in the
lowest 20 meters above the surface. Both methods provide a good estimation up to 100 m, but the power-
law is then more appropriate above 100 m and in the lowest part of the PBL (Cook 1985).
24

-------
The urban canopy boundary layer profile, another type of vertical wind profile used in dispersion
models, is used for situations in which there are buildings upwind of a release source, and the user wants
to account for urban drag on the inflow profile. The urban canopy boundary layer profile is not intended
for large downtown high-rise structures. (M. Brown, personal communication, 2018). The formula
represents vertical wind speed with height for two conditions: above and below H in the domain (where
H is the average height of the urban building canopy). The urban canopy profile accounts for lower wind
speeds below// as it is slowed by the drag from the buildings:
where u^ is the mean velocity measured at H and A is the attenuation coefficient representing the average
impact of the buildings or vegetation canopy on the flow (Nelson and Brown 2013). This equation (3) was
originally presented by Cionco (1965), who developed a wide range of empirical values for A (Cionco
1978). These values were introduced to modify flow based on vegetation canopies for a wide variety of
grasses, shrubs, and trees. The values represent an altered airflow response to vegetation roughness and
density, with values of approximately 1-2 for small trees, rice, and corn, and up to 4.5 for maple, fir, and
gum trees. Lower values represent rigid and sparsely arranged objects, while higher numbers indicate
dense and flexible obstructions (Cionco 1978). Buildings are undoubtedly dense as well as rigid, so there
is not a widely accepted value for A since the equation was not intended for dense urban use.
At heights higher than H, a transition to a modified logarithmic wind profile is introduced:
where L is the Obukhov length, which accounts for stability effects. Equation (4) is used to avoid
discontinuities in the velocity profile (Nelson et al. 2009). The premise behind the equations is discussed
in further detail in MacDonald (2000). Nelson et al. (2009) recommends setting zref equal to H (and
therefore wre/is the upwind velocity at H) to avoid the wind speed ballooning above the canopy height that
may introduce erroneous results.
u
(z H) =
(4)
25

-------
5.0	Types of Atmospheric Dispersion Models
Atmospheric dispersion modeling has evolved tremendously over the past 100 years while still
retaining many of the theoretical and mathematical representations of the fundamental dispersion
equations. Even though dispersion modeling has seen improvements throughout the decades, many of the
mathematical foundations are still based upon the original building blocks of Gaussian dispersion models
(i.e., Pasquill 1962, 1974), Lagrangian particle models (which track a particle or puff under a moving
frame of reference), stochastic (random walk) or other statistically based models, or Puff models. By the
1960s and 70s, computers began to be commonly used to solve Gaussian plume equations rapidly instead
of completing the computations by hand. As computers advanced in the 1980s, Lagrangian Puff models
and simple Eulerian models (with a fixed frame of reference as particles are free to move throughout the
domain) were introduced. The 1990s and 2000s saw an advancement of Eulerian 3D grid models as
algorithms and model resolution improved. Today, higher-order CFD models are commonly used in
research applications but are rarely used by emergency responders in the field. For this reason, classic and
relatively simple Gaussian-based dispersion models continue to be at the forefront of emergency response
due to their fast output times, although higher fidelity models are also used for emergency preparation
exercises.
Dispersion and diffusion models predict the distribution and concentration of a constituent
downwind and in a typical study, the local meteorological scenarios are defined in terms of a characteristic
wind speed, direction, and atmospheric stability. The atmospheric state and stability (i.e., stable, neutral,
or unstable/convective) may be represented by the discrete Pasquill stability class (Pasquill 1961), a
function of wind speed, solar radiation, and cloud cover, as turbulence measurements are rarely available
for the model domain of interest (Bowers et al. 1994). More recent advances in dispersion modeling
parameterize atmospheric stability using Monin-Obhukov similarity theory (e.g., Cimorelli 2005). Models
also require basic information on the amount of material released (source term) and removal mechanisms.
The process components of a dispersion model are shown in the flow chart in Figure 6. The rest of this
section provides a broad overview of the different types of dispersion models starting with the simplest
and progressing to the more advanced. A visual comparison of the models can be seen in Figure 7. The
strengths and limitations of different types of dispersion models are also provided in Table 1. This section
satisfies Objectives 5 and 6 as detailed in Section 2.3.
5.1	Box Models
These simple models assume that the domain is one large homogeneous volume, and any substance
entering this volume is uniformly and instantly mixed throughout the box (Arya 1999). Box models are
generally stationary (considering a city area or the transport between two regimes such as the troposphere
and stratosphere) and may consider fluxes or flow in and out of the box (Fin and iw), production (P),
chemical loss (Z), deposition (D), and emission (E) terms to understand production and destruction in
simplistic terms (5). Box models can often be solved on paper through the addition or subtraction of
production and loss terms and are best used for quick, simplistic transport calculations. The mass balance
of a box model is the sum of the sources minus sum of the sinks:
^ sources — ^ sinks = Fin + E + P — Fout — L — D	(^)
26

-------
Meteorological
Inputs
(Observation or
Forecast)
Source Term
(Location,
I Quantity, and Rate I
of Emissions)
Meteorological
Data Refinement
r	^
Transport and
Diffusion Source
Code
(Dispersion,
Plume Rise,
Transformation,
Advection, Urban,
and Complex
Terrain Effects)
Downwind
Concentrations
Deposition
Human Health
and Exposure
Effects
Environmental
Effects
Figure 6: Components of an atmospheric dispersion model, modified after OFCM (2002) and Turner
(1979).
5.2 Gaussian Plume Models
The Gaussian plume model is often used as the fundamental basis for many dispersion models and
is particularly useful for quick calculations that assume conditions to be horizontally homogeneous. Most
Gaussian plume models also imply that the pollutant source is released continuously and that the
concentration profile downwind of the release has a cross-section that resembles the classic Gaussian bell-
shaped distribution. The formulas are derived assuming steady-state conditions because their results
represent ensemble averages. Since the model assumes that meteorological conditions will remain
constant across the horizontal domain over time, a Gaussian plume model is best used for estimates within
20-50 km of the release location as long as the wind direction and speed are consistent. Therefore, it is
essential that the source term characteristics be accurate for the best model results. Gaussian plume models
are challenged in urban areas with complex building geometries because localized releases interact with
street canyons and building wakes, especially in the near-field region close to the release point (Belcher
et al. 2013). Comprehensive field, model, and laboratory understanding are generally lacking for near-
field dispersion effects (Coceal et al. 2014).
Early solutions to the diffusion equation implemented in basic Gaussian diffusion models were
based on turbulent transport of material through the gradient transport theory or the "K-Theory", in which
models assumed that turbulent transport was proportional to the gradient of the mean concentration, and
a proportionality factor, or diffusivity, was implemented to represent turbulent transport. Boundaries
usually play a large role in dispersion, including the presence of a temperature inversion or interaction of
the plume with the ground. Therefore, when the plume deflects off a surface, an "imaginary" source is
represented using the modified equation by Turner (1970). The Gaussian downwind concentration (C) at
downwind distance x, lateral distance^, and vertical distance z for a continuous point source is represented
as:
C (x, y, z) =
2,Tril(7y(7z
exp
\(y_
2 \<717
exp
1	tz - H\"	l/z + ti\
2	\ ~o7 / + exP ~ 2 ( j
1/z + H^
(6)
27

-------
where Q is the source strength, or emission rate of the released material, usually in grams or milligrams
per second; u is the mean transport wind speed in meters per second, which usually represents the wind
at source height or the layer containing the plume; and o (x, y, z) is the longitudinal (along-wind), lateral
(crosswind), and vertical dispersion coefficients in meters, which may be based on the classic Briggs
(1973) equations or the Pasquill-Gifford-Turner (PGT) stability classes and curves (Gifford 1961; Pasquill
1961; Turner 1970) and other formulations, and H accounts for the source release height. Arya (1999) and
Bowers et al. (1994) expand on these equations in much more detail, and a complete set of Gaussian plume
equations for various sources (such as line, point, or area sources that contain continuous releases) is
provided. Examples of Gaussian Plume models are AERMOD (see Section 8.4) or ALOHA (see Section
8.5).
5.3 Gaussian Puff Models
This model type (also known as a segmented plume) divides the emission into a series of
overlapping "puffs", allowing the release source no longer to be steady, if desired. The horizontal
meteorological conditions also do not need to be homogeneous. Puff models may treat releases as point,
line, or area sources where a pollutant is released as one instantaneous amount (as in an explosion) and
then tracked with the Eulerian or Lagrangian frame of reference. Puff models may also encompass a series
of successive near-instantaneous releases or "puffs" that are released and tracked discretely into the
ambient environment, since even continuous releases could be thought of as a series of overlapping puffs
(Arya 1999). Each individual puff is simulated by numerically integrating the 3D advection diffusion
equation as it diffuses into the air based on constant or time-varying wind conditions. According to the
statistical theory of diffusion, the mean concentration of a puff is Gaussian in all directions (unless affected
by an external barrier like a temperature inversion or obstacle), and the spread of the plume puffs is related
to statistical diffusion parameters (Arya 1999). Each puff of material is assumed to be horizontally
symmetrical and the average concentration (C) in the puff follows the Gaussian form from Slade (1968):
C (x, y, z) =
( 2 71) 2 (Jx (Jy Oz
¦exp
1/
x \2

1|
f y_\2
2 \
oV .
exp
2
Kay)
exp
i(L

(Jy
+ exp
i(-
2 \ o7
)'ll
(7)
The lateral plume dispersion parameter oy in the first denominator is assumed to be equal to ox as
a simplifying assumption in the integrated plume method. The Gaussian Puff approach leads to
calculations that could be more accurate since each successive puff responds to the overall wind conditions
and is tracked across multiple sampling periods. Puff models can be used for chemical or radioactive
releases and are better applied to continuous releases over a longer sampling period (instead of
instantaneous). However, the plume shape may vary based on sampling time versus puff travel time from
source to receptor (Hanna et al. 1982). Urban areas within the trajectory of a puff are often parameterized
as roughness elements, so the effects of single buildings, clusters, or full cities are not generally captured.
SCIPUFF is the main Lagrangian Gaussian puff dispersion code of the HP AC model commonly used
through IMAAC emergency response (DTRA 2004).
5.4 Lagrangian Stochastic Particle Models
A Lagrangian model divides emissions into small particles or parcels that are tracked individually
as they are stochastically transported and diffused downwind. At each time step in the model, the particles
28

-------
are moved by 1) the mean wind, 2) turbulent diffusion by random fluctuations in the horizontal and vertical
winds, and 3) molecular diffusion. The trajectories of particles in the 3D wind field are calculated by the
random walk method. Lagrangian dispersion models are typically employed for incidents involving
complex meteorology, strong wind shear, or complex wind schemes in urban areas (OFCM 2002). Since
the trajectories of thousands (or millions) of particles are tracked on each model time step and based on
the turbulent deviation of the wind from the previous time step, the simulation could be computationally
intensive. However, meteorological variables including wind fields are often run "offline" from the
dispersion code (i.e., not at the same time). Some researchers and operational modeling centers believe
Lagrangian models such as NARAC's LODI model used in IMAAC (Bradley 2005) can resolve point
sources and simulate turbulent diffusion in greater detail than Gaussian models. The QUIC model (Nelson
and Brown 2013) is another example of a Lagrangian stochastic model, as well as NOAA's Hybrid Single-
Particle Lagrangian Integrated Trajectory (HYSPLIT) model that is widely used for long-range trajectory
and dispersion calculations (Stein et al. 2015).
5.5	Eulerian Grid Models
In a Eulerian grid model, the domain is divided into a 3D array of rectangular grid cells where
within each cell, the mixing of a substance is uniform and instantaneous. This type of model is best used
to understand regional air quality issues such as ozone (O3), particulate matter (PM), and nitrogen oxides
(NOx), and is more commonly referred to as a regional air quality or mesoscale atmospheric model, rather
than a dispersion model, although they can be applied to regional dispersion. EPA's Community
Multiscale Air Quality (CMAQ) model (EPA 2019) is an example of a Eulerian Grid model. Since these
models are usually meant for larger spatial or temporal scales and regional air quality issues, they will not
be covered in this document since this work is more focused on near-field dispersion.
5.6	Higher Order Models
Dispersion in complex urban environments often involves complex phenomena that lower-order
dispersion models are unable to capture adequately. However, due to the dramatic increase in
computational power over the past few decades, CFD model advancements using LES or Reynolds-
averaged Navier-Stokes (RANS) methods can provide a more detailed description of the flow and
dispersion surrounding complicated urban obstacles with varying geometries (Tominaga and Stathopolous
2013). Traditionally, CFD models have been useful for research, case studies, or emergency planning
initiatives and are best applied to understand site-specific phenomena rather than for operational use due
to their complex input data preprocessing requirements and longer computational times. In this capacity,
reduced-order Gaussian dispersion models are still of prime significance due to their widespread use in
operational settings and timely simulation results (Philips et al. 2013). However, CFD and LES
simulations can provide denser evaluation datasets than field or laboratory studies and can be used to
improve algorithms in other dispersion models.
CFD models portray the advection and diffusion of pollutants in a fluid flow. The transport of a
contaminant is solved through the Navier-Stokes equations, which describe the mathematical basis of a
fluids flow. RANS equations are time-averaged expressions for the motion of a fluid and are essentially
derived through Reynolds decomposition where an instantaneous value is broken into its fluctuating and
time-averaged components. The atmospheric equations for continuity, motion, and thermodynamic energy
express the conservation of mass, momentum, and heat in a volume of fluid. However, the set of RANS
equations consists of one or more extra unknown values compared to the number of equations, thereby
not permitting a solution to exist for the highly non-linear system of equations in a turbulent flow (Arya
2001). This situation is regarded as the "fundamental problem of closure" and is one of the most
29

-------
challenging issues faced in simulating turbulent flows (Arya 1999). The closure method chosen defines
the speed of processing and the amount of detail in the simulation. Different methods of closure for
modeling the Reynold's stress terms, such as introducing eddy viscosity, mixing length, or other
turbulence "model" laws and concepts have been developed. These methods are beyond the scope of this
report and can be referenced in micrometeorology textbooks such as Arya (2001, 1999).
Another CFD modeling technique using the Navier-Stokes equations is direct numerical
simulation (DNS), where no turbulence closure models are needed in solving the Navier-Stokes equations.
Instead, the spatial and temporal scales of turbulence are resolved using a fine numerical grid to capture
all turbulence scales. This method of CFD modeling is extremely computationally intensive, even in low-
turbulence environments (Arya 2001).
CFD modeling through LES is a less resource-intensive approach than DNS that parameterizes the
smallest scales of a simulation through a Navier-Stokes equation-based filter (Arya 2001). The initial
concept for LES modeling was introduced in the early 1970s for weather and boundary layer meteorology
modeling (Deardorff 1970). The "low-pass filter" is a time and space-based method that removes small-
scale motions and employs a sub-grid scale model parameterization to address the most resource-intensive
processes of turbulence. As a result, CFD modeling with LES is a more advanced higher-order approach
that more appropriately captures some of the important physics compared to RANS (Castro et al. 2017).
Since LES requires less computational power than DNS, LES promises to be an important area of CFD
modeling research, especially for urban flows where LES has been shown to provide better results for
simulating concentration distributions around obstacles (Tominaga and Stathopolous 2013).
While CFD codes may appear to be a more accurate option for understanding dispersion, they have
their own set of challenges and limitations. The models are more likely to be used for research purposes
rather than for emergency response or operational use due to longer processing times, computational
requirements, and greater user-learning curve. The local area of simulation has also been shown to tend
to be somewhat "disconnected" from larger mesoscale phenomena because assumptions about the vertical
structure, surface energy fluxes, and wind patterns are parameterized for computational reasons (OFCM
2002). However, CFD modeling offers a realistic and detailed computational approach for detailing the
development and dispersion of a plume within the near-source canopy region (Philips et al. 2013). The
U.S. Naval Research Laboratory has introduced a hybrid plume dispersion model using LES called CT-
Analyst (Boris et al. 2003). The model produces near real-time contaminant transport predictions of CBR
agents in complex urban settings, although pathways for the release scenario and several databases,
including the wind fields, must be extensively set up and calculated in advance (Leitl et al. 2016).
5.7 Street Network Models
A newer generation of high-resolution urban dispersion models that simulate a network of
interconnected street canyons and intersections among rectangular buildings has been introduced by
Soulhac (2000) and further developed by Hamlyn et al. (2007) and Belcher et al. (2015). These street
network models require only a few basic inputs and somewhat resemble "modified" Gaussian dispersion
models (Ben Salem et al. 2015), as complex building structures are simplified and not explicitly defined.
SIRANE was the first operational street network dispersion model to simulate line and point sources
within the urban canopy (Soulhac et al. 2011). Instead of portraying a large plume over a domain, the
model domain is composed of a 3D network of interconnected streets surrounded by simplified cube
shaped buildings. The effects of the flow within the streets have their own parameterizations and are
decoupled from the overlying PBL above the urban canopy. A Gaussian plume approach above the canopy
accounts for the overall atmospheric dispersion throughout the UBL as pollutants are dispersed within the
street canyon based on building geometries. The model has been evaluated against field campaign data
30

-------
within the city of Lyon, France, and shown to perform reasonably well, although some errors existed
among spatial and temporal evolution of the source emissions in the simulation (Soulhac et al. 2012). A
more recent adaptation of SIRANE is the SIRANERISK dispersion model, which can simulate steady and
unsteady releases above and within the street network and has performed well when compared with wind
tunnel studies (Soulhac et al. 2016).
Another street network model called the University of Reading Street-Network Model (UoR-
SNM) has also been introduced but avoids velocity flow parameterizations that must be calculated
externally before the model can be run (Belcher et al. 2015). In a head-to-head comparison among UoR-
SNM, SIRANE, CFD models, and QUIC-generated flow fields, Hertwig et al. (2018) found that the
relatively simple street network models performed equally as well as or better than Lagrangian models
with 3D wind fields, while also saving computational time and cost. The goal was to study urban
dispersion on scales of interest for emergency response applications. SIRANE is currently the only
operational street-network model since the UoR-SNM, which is still under development. While these
models show promise for urban dispersion modeling, considerable testing and modification is required to
make the models frontrunners for urban dispersion applications. One of the biggest hurdles is to adapt the
urban morphology employed in these models for use in other locations with vastly different characteristics
(Hertwig et al. 2018). Most of the idealized testing was done in European cities that often do not resemble
the layout of cities in the United States.
31

-------
5.8 Comparisons, Strengths, and Limitations for Atmospheric Dispersion Models
Figure 7 and Table 1 provide a visual comparison and show strengths and limitations for the
different types of dispersion models.
Gaussian Plume
Models
Gaussian Puff
Models
Typical Computational Times:
Seconds	Minutes
Lagrangian
Dispersion Models
Minutes to Hour
CFD Models
Hours
Very fast
estimates with
minimal
calculations
Limited spatial or
temporal
variability of wind
Simplified urban
effects included
by parameterizing
plume spread
-Acimi #*c«
H -CHfctnw cm
-pcfcitar* ivImm
a*m*
Fast estimates
from a sequence
of puff emissions
Time-varying
winds
Winds are not
building-resolved
and lack spatial
variability

s»i
I Hi nzam s	V
-if* -i unr... V
£~t.l££5M,tsr2 v£'
Track individual
particles
downwind
Building-resolved
features and
parameterized
winds
Require
considerable
knowledge and
training
Detailed and
higher resolution
simulation
Based on
theoretical physics
Require expert
knowledge and
resources
Increased model complexity, accuracy, and computational requirements
Figure 7: Visual comparison of dispersion models that can be applied for homeland security, emergency
preparedness, and emergency response. As the models increase in complexity, so do their computational
and user requirements. Many have special applications for urban use. Figure adapted from presentation:
National Atmospheric Release Advisory Center's Urban Plume Dispersion Modeling Capability for
Radiological Sources by Gowardhan et al. (2018) with additional modifications.
32

-------
Table 1: Summary of strengths and limitations for different types of atmospheric dispersion models.
MODEL TYPE	STRENGTHS	LIMITATIONS
Box Model
•	Commonly used as a screening model
•	Fast and simple calculations with minimal input
requirements
•	Can account for simple photochemical production and
loss
•	Multiple "boxes" can interact with each other
Uniform distribution and single value arc calculated for the
entire domain
Poor representation for point sources in the near field
Simplest treatment of input (production) and removal (loss)
mechanisms and atmospheric conditions
Gaussian Plume
Model
Uses simple and well-tested, peer reviewed empirical
formulas that provide results quickly with relative
accuracy, but under simplified conditions
Provides reasonable representations of average or long-
term downwind concentration behavior
Consistent with the random nature of atmospheric
turbulence
Additional terms can be added or removed to modify the
Gaussian plume equation
Supported for regulatory use
Easy to use and implement
Core of many current dispersion models used for
emergency situations
Recommended for applications no greater than 30-50 km
from the source
Can be calculated only under a single wind observation
under limited spatial and temporal ranges
o Errors in input meteorology strongly impact the result,
especially in more complex and low wind situations
Best used for simple terrain without large fluctuations in
atmospheric stability
Misrepresentation of source term could significantly change
model result
Rudimentary representation of nocturnal boundary layer
May lack spatial concentration variability based on obstacles
Inherent uncertainty due to stochastic nature of turbulence
Gaussian Puff Model
Better than Gaussian plume for representation of t ime
varying meteorological effects ("curved" plume)
Strength similar to the Gaussian plume model but can
predict ensemble-averaged puff concentration as a
function of t ime as wind speed/direction changes.
More accurate in low wind speed situations.
Frequently used in emergency situations
Sometimes coupled to or implemented as preprocessor for
another model
Best applied to continuous releases over longer sampling
period to account for puff travel time
Similar limitations to the Gaussian plume model
Unable to capture strong changes in wind variations
Difficult to predict ground-level impacts based on puff
trajectory
Parameterized urban and complex topography (no individual
buildings seen)
May be less accurate than Gaussian plume model when wind
observations arc not representative for release location
33

-------
MODEL TYPE
STRENGTHS
Lagrangian Particle
Model
Predicts dispersion under lime varying meteorological
conditions with different stabilities
Continuous, short term, and instantaneous releases
Particles arc tracked on each time step on a random path
to allow greater representation of wind characteristics
Includes a type of parameterization for complex or urban
terrain
More flexibility in source release type
Some variations arc "CFD-likc" but run much faster
Eulerian Grid Model
Higher Order
(CFD and LES
Models)
Concentration values are calculated for each grid cell in
the 3D model for greater spatial coverage, detail, and
accuracy
Includes photochemical reactions
Detailed plume interaction from local meteorology and
source characteristics
Offers a realistic and detailed computational approach to
detail the development and dispersion of a plume within
the near-source canopy region
One of the best current representations of dispersion
around obstacles in complex environments
Provides a denser datasct to evaluate field or laboratory
studies
Flexibility with source type, domain geometry, and mesh
sizes
Street Network
Model
A newer generation of high-resolution urban dispersion
models that simulate a network of interconnected street
canyons and intersections among rectangular buildings
Requires only a few basic input conditions
Performs equally as well as or better than Lagrangian
dispersion models while saving on computational time
34
LIMITATIONS
•	More complicated to run and higher learning curve than
simpler models
•	May take a longer time to run depending on domain (<1
hour) so not always applicable to emergency response
•	More detailed meteorology and source term requirements
•	Complex to set up and run, precluding use for emergency
response
•	Need to interpolate meteorology using reanalysis data or run
"online" with the model
•	Mostly reserved for research applications
Solves timc-avcragcd RANS equations and users must often
outweigh tradeoffs between closure methodologies (i.e..
DNS vs. LES)
Computationally intensive and usually reserved for research-
grade applications rather than operational use
Domain of simulation tends to be somewhat "disconnected"
from the larger mcsoscalc systematic behavior
•	Research grade - considerable future testing and
modification are required to make the models operational.
•	Simplified building geometries that must be adapted to the
urban morphology in other locations
•	Most testing has been done in European cities that may not
resemble cities of the United States.

-------
6.0	Model Review Process
The remaining sections of this report address Objectives 7 and 8 to document and summarize
currently known atmospheric dispersion models into a concise reference table (Section 7.0). Then, a
selection of those models is screened for a more in-depth description, intended to serve as a quick reference
guide for federal, state, and local agencies and stakeholders requiring dispersion model guidance for
emergency preparation or response. The current section briefly describes the model review process, ready-
reference table, and methods for choosing a subset of models as part of a two- or three-page reference
guide in Section 8.0.
6.1	Quick Reference Table
Ninety-six dispersion models that simulate the fate, transport, and diffusion from an effluent source
have been identified in this report. The quick reference table includes dispersion models that can be used
for a wide variety of applications, including air pollution dispersion, particles dispersed within the wind
flow fields, as well as release agents specific to homeland security threats. Some models are designed for
more specific CBRNe applications involving gas, biological particles, and nuclear and explosive
dispersion. The purpose of this reference guide is not to recommend or endorse a specific model but to
provide users with a resource that documents the currently available models so that the user can make
informed decisions. Attempts have been made to keep the information concise and applicable to
emergency response officials. While this is a reasonably comprehensive list and assessment of current
dispersion models, it is not a completely exhaustive compilation of every known dispersion model both
past and present.
The starting point for this list was the early reports from NOAA and US DOE summarizing
consequence assessment models (OFCM 1999; Mazzola et al. 1995). These documents were used as the
foundation of this work and are described in greater detail in Section 2.3. However, many of the models
on these lists have not been updated recently, are no longer supported and are obsolete, are specific to
certain facility sites (such as particular nuclear power plants or national laboratories), absorbed into newer
model formulations, or even included as modules in more powerful and modernized models. One example
of this is SCIPUFF, which drives HP AC. In addition, the 3D meteorological fields generated from
CALPUFF can also drive MM5, WRF, CTDMPLUS, or ISCST3. Screening-level dispersion models are
also not included in this list, but the non-screening version of the model is described. For example,
SCREEN3 is the screening model for ISC3.
Literature reviews of peer-reviewed journals, technical reports, internet searches, and discussions
with emergency response personnel also led to the current dispersion model list. Models documented in
OFCM (1999) that could not be found with a relatively in-depth internet search were not included in the
model review and reference table. The internet search included model documentation or mention of the
model in any report or journal article. For example, limited information about the outdated "ARCHIE"
model is available on the internet and has been eliminated from review. The justification for this
elimination is that if an internet search cannot easily find information about a particular model, an
emergency response official may use something else. Emphasis has been placed on documenting
dispersion models developed or used within the United States, although some well-known and flexible
international models are featured in this document.
Section 7.0 provides a quick reference table for obtaining model summaries and other key
information. The list is numbered and sorted alphabetically by model name. The second column provides
the model name in its expanded and abbreviated form. The best reference source for the model is
hyperlinked on the model's short name or abbreviation. This link is likely to be a website with currently
35

-------
available information, a user's manual, or a description from the developer. For those models, a link is
provided to the next best source such as a journal article or other reference document, since some models
lack a good source of information online. For printed versions of this report, access to the hyperlinks is
available through the digital version. (Note: Some word processors may have trouble opening links to
PDF files. To access these websites and documents that are PDFs, copy and paste the link into your internet
browser or convert this document to a PDF and then try accessing the link.) The model developer is then
provided in the third column. Immediately after this, a one-sentence description outlines the model's core
function and purpose. The group of columns 5-10 consists of check boxes to identify the model type (e.g.,
Gaussian Plume, Puff, Lagrangian, Eulerian, or CFD). Columns 11-14 provide check boxes for the
model's intended CBRNe application or the best possible use scenario. There is some flexibility in the
identification of chemical, as general air pollution dispersion models technically simulate chemical
transport. The 15th column identifies the model's best emergency application aligned with the National
Planning Framework: either preparedness (including prevention, protection, and mitigation), response
(including recovery), or both, and is shaded in orange, purple, or blue, respectively. The five emergency
response frameworks are grouped into two to simplify the model's best application pre- or post-event.
Some models are better designed for planning or response purposes, and it is important for responders to
have this key information. The 16th column objectively defines the model's runtime speed and is therefore
directly related to its response classification. The yes (Y) or no (N) identification in columns 17-19
indicates whether the model simulates terrain or building effects and whether the model is proprietary.
The latter would be Y if the model is site-specific or does not provide the code to users outside its
developing entity.
The two rightmost columns in the quick reference table identify if the model was selected for
additional review later in this document. A three-tier classification system was applied to the model list
where the model was either: 1) excluded from the detailed list (pink), 2) possibly of use for emergency
preparation and response, but excluded from the detailed review (yellow), and 3) included in the detailed
review (green). If the model was not selected, a brief reason is provided in the final column for yellow
and pink classifications. An explanation of this process and number of identified models is outlined in
Table 2
6.2 Expanded Model Description
Out of the 96 dispersion models, 40 show viable use for emergency preparation and response
purposes, but only 16 were selected for expanded model descriptions. Of those 16 models, four were
identified to be best used for emergency planning purposes, and only one was best suited solely for
response. Twelve could be applied to pre- or post-release timeframes. The 16 models most viable for
emergency preparation, response, or both are expanded in a two- or three-page reference guide which
includes the information outlined in Table 3.
The information used for each model review was derived from several sources: 1) description,
documentation, manuals, and factsheets listed on the model's official website, 2) literature searches of the
model's name including peer-reviewed and gray literature, 3) a review of field, laboratory, and real case
studies of model applications obtained from internet searches, and 4) outreach to model developers, other
specialist users (including this report's author and research team), and emergency responders. Given the
broad range of models considered, direct testing of each model was beyond the scope of this review.
Ranking each model by a scoring system was also beyond the scope of this report since each model has
widely different capabilities and purposes.
36

-------
Table 2: Model classification criteria for inclusion or omission in detailed model review.
Classification
Color
Code
Number of
Models
Reasoning
Useful for
Emergency
Response
Applications

16
•	Developed specifically for emergency preparation,
response or for CBRNe applications
•	Widely used and referenced within literature with
significant support base
•	Fast and straightforward for most users
•	Free or minimal cost
•	Operational or research grade
•	Includes some application to urban environments
•	Developed and used for applications within the US
Possibly of
Use for
Emergency
Response
Applications

24
•	Simulates dispersion for some CBRNe applications
with some emergency response use
•	Use dependent on need or user's situation
•	Too specific for generic use
•	Not as easy or intuitive for use by non-modelers
•	Moderate to slow running
•	Significant self-research required
•	Mainly research and/or development grade
•	Incorporated within other dispersion models
•	Developed or primarily used for international
applications
Not Useful

56
•	Not applicable to emergency preparation or response
use
•	Site-specific for certain facilities (i.e., specific nuclear
power plants or national laboratories)
•	Not widely used or discontinued
•	Difficult or impossible to find information or
references
•	Not recently updated or replaced by more advanced
model(s)
•	Too expensive, proprietary, or not open source
37

-------
Table 3: Model criteria and explanation of information provided in the expanded model descriptions.
Model Criterion
Explanation
Model name and
Abbreviation
Short and expanded model name
Developer
Name of company, agency, or individual who developed the model
Type of Model
If the model is built upon a Gaussian Plume, Puff, Lagrangian, Eulerian Grid, or
CFD framework
Response Phase
Whether the model is best applied to emergency preparedness, response, or both
Original Application
Whether the model is meant mainly for urban, rural, or complex terrain or has
capabilities to simulate around buildings; additionally, the type of CBRNe
release(s) the model is best applied for
Model Description
1-2 paragraph description of the model framework, purpose, capabilities, and
recent studies, if available
Pros and Cons
Known model advantages, benefits, disadvantages, or shortcomings
Runtime
A general qualitative speed in setting up, running, and post-processing the model
results
Input Data Requirements
Typical information or data the user needs to initialize the model and difficulty
of preparing these datasets from publicly available information
Outputs
Nature and format of outputs
Data assembly
requirements during or
after emergency response
As above, but specifically considering the potential of rapidly setting up a model
to respond to an emergency
Code language
If known, the computing code foundation the model is developed on for potential
debugging
Public or Proprietary,
Cost
Model availability and price for government officials, researchers, or individuals
Ease of use
Qualitative measurement of simplicity for responders or researchers, including
any barriers to the widespread use in terms of training or specialized hardware or
software requirements
Ease of obtaining
information and
availability of technical
support
Ability to request external help, including a user support group, website help
pages, or technical support contacts
Source code availability
If the source code is available for dissemination for modification or debugging,
if needed
Installation requirements/
software
Hardware computing requirements or specialized technology needed
Maintenance Status
If the model is available for use, undergoing continuous development and
improvement, complete, or obsolete, including current version
Documentation
If information on model use and formation is documented in a user's guide or
website
Link to Website
Hyperlink to the best-known source of the model, as of Summer 2020
38

-------
7.0 Dispersion Models- Quick Reference Table




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
1
Q.M
C
.2
*
-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
6
Aeolus
NARAC. LLNL.
DOE
Research model to simulate high-resolution
flow and dispersion of hazardous material in
urban areas and complex terrain environments
for emergency planning guidance.




X

X
X
X
X
Preparedness
Mod-
erate
(
-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
12
AOPAC
Air Quality Package
Atmospheric
Environment
Service (AES);
Environment
Canada
Emergency response model for the prediction
of hazard zones from accidental puff or plume
chemical releases, included from a large
chemical database
X
X




X



Response
Fast
N
N
Y

No recent
updates,
proprietary and
no longer
supported
13
ARCON96
Atmospheric Relative
Concentrations
NRC, PNNL
Constant straight-line Gaussian dispersion
model used to calculate nuclear power plant
control room concentrations and habitability
from accidental releases of radionuclides
through air intakes
X







X

Mostly for
Preparedness
of accidental
releases
Fast
N
Y
N

Not an
emergency
response
model, specific
to NRC sites
14
AREA EVAC
Area Evacuation
Westinghouse
Savannah River
Company
Transport and dispersion code used alongside
2DPUF for the WINDS GUI to predict
radionuclide dispersion and best rally area
upon accidental release
X







X

Preparedness
Fast
N
N
Y

No recent
updates, site
specific to SRS
15
ASPEN
Assessment System for
Population Exposure
Nationwide
EPA
Alternative EPA Gaussian dispersion and
mapping tool to estimate toxic air pollutants
across a wide area of the LIS based on rate,
location of release, and meteorological
conditions, and removal processes for
calculating exposure by census tract
X





X



Preparedness
Fast
N
N
N

Mainly for air
pollution,
mostly replaced
by ISC3,
AERMOD, and
other exposure
models
16
AXAIRO
AXAOTHER XL
Westinghouse
Savannah River
Company
Dose assessment code used at the SRS to
predict hypothetical nearby and short-term
downwind radionuclide doses from inhalation,
plume, and ground shine. AXAIRQ considers
light to moderate winds while AXAOTHER
XL simulates high velocity winds and
tornadoes for safety-related documentation
X
X






X

Preparedness
Fast
Y
N
Y

Site specific to
SRS
17
BLP
Buoyant Line and Point
Source Model
Environmental
Research and
Technology, Inc.
(ERT)
Alternative EPA model designed to simulate
dispersion associated with stationary line and
point industrial sources, particularly aluminum
reduction plants, with buoyant plume rise and
down wash algorithms
X





X



Preparedness
Fast
N
Y
N

Limited
applications
(i.e., industrial
aluminum
plants); now in
AERMOD
18
BNLGPM
Brookliaven National
Laboratory Gaussian
Plume Model
Brookliaven
National
Laboratory (BNL)
Site specific dispersion code used to provide
real-time projection of downwind doses of
radionuclides released from BNL stacks based
on local on-site meteorology
X





X



Response
Fast
N
N
Y

Site specific to
BNL
41

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
19
B&M Workbook
Britter and McQuaid
Workbook
Britter and
McQuaid (1988)
A rapid non-computer-based screening
method based on a set of nomograms to
provide a hazard estimate of dense gas
dispersion and downwind, ground-level
concentration from continuous or
instantaneous releases
X
X




X



Reparedness
Fast
N
N
N

Not an
emergency
response
model; for
screening
purposes
20
CALINE4 and
CAL30HCR
California Line Source
Dispersion Model
California
Department of
Transportation
(Caltrans)
Steady state model for calculating pollution
concentrations at receptor locations downwind
of highway line sources to assess
transportation-related air quality impacts.
Replaced by AERMOD as one of EPA's
preferred and recommended models.
X





X



Reparedness
Fast
N
N
N

Not an
emergency
response mode;
replaced by
AERMOD
21
CALPUFF
California Riff Model
Originally Sigma
Research
Corporation (SRC);
now Exponent, Inc.
Multiple component, non-steady state Riff
model used to simulate buoyant, puff, or
continuous-release, long-range transport of
pollutants, emission and removal processes,
and sometimes used to drive other dispersion
models through high resolution meteorology.

X
X



X



Reparedness
Mod-
erate.
Y
Y
N


22
CANARY
Quest Consultants,
Inc.
Hazard assessment model used to model vapor
dispersion from pressurized, superheated, and
refrigerated liquids, pools, jets, fires, and
explosions for a database of many well-known
chemicals.
X





X


X
Both
Fast
N
N
Y

Requires
purchase from
consulting
company,
designed for
industry
applications
23
CAP88-PC
Clean Air Act
Assessment Package -
1988
DOE and EPA
A set of programs and packages for estimating
the dispersion, dose, and risk from
radionuclide emissions from up to six sources
at DOE facilities to ensure compliance with
the CAA
X







X

Reparedness
Fast
N
N
N

Not an
emergency
response model
24
CAPARS
Computer-Assi sted
Protective Action
Recommendations
System
AlphaTRAC
(Terrain
Responsive
Atmospheric Code)
A modernized version of the TRAC Risk
Assessment/Hazard Assessment (RAHA)
model used to produce real-time emergency
planning and response dispersion, deposition
plumes, and associated health impacts for
releases within complex terrain at DOE sites

X




X

X

Both
Fast
Y
N
Y

Designed for
use at DOE's
Rocky Flats
facility
42

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
25
CASRAM
Chemical Accident
Stochastic Risk
Assessment Model
ANL
A statistical analysis model that determines
the distribution of hypothetical outcomes of
affected populations associated with
hazardous chemical release materials stored or
transported through an area, using local
meteorology and Gaussian/dense gas plume
relationships for reporting in the ERG.
X





X



Response
Fast
N
N
N


26
CATS-JACE
Consequence Assessment
Tool Set/Joint
Assessment of
Catastrophic Events
DTRA; FEMA
Estimates the consequences of human and
natural disasters to the population,
infrastructure, and resources using underlying
dispersion models within a GUI and outputs
results in geographic information system
(GIS) formats for real-time response

X




X
X
X
X
Both
Fast
Y
N
Y

Capability
largely
encompassed
within HPAC
and HAZUS
suite; JACE
only available
to U.S. Federal
government
27
CT-Analyst
Contaminant Transport
Analyst
U.S. Naval
Research
Laboratory
An instantaneous, 3D LES model depiction of
CBRN releases within complex urban areas to
aid emergency responders in accidental or
intentional windbome contaminant transport
threats with fine-scale resolution




X

X
X
X
X
Both
Fast
(sees.)
Y
Y
N


28
CTDMPLUS
Complex Terrain
Dispersion Model Plus
Algorithms for Unstable
Situations
EPA
Refined elevated point-source, steady state
dispersion model for use in various
atmospheric stabilities and terrains, especially
for receptors on or near 3D terrain features,
and one of EPA" s preferred and recommended
models.
X





X



Preparedness
Fast
Y
N
N

Mostly for
complex
terrain-related
routine air
pollution
emissions, not
emergency
response
29
CUDM
Canadian Urban
Dispersion Model
Environment and
Climate Change
Canada
Semi-operational, building aware CBRN
dispersion modeling system similar to QUIC
and LODI with numerous features that
simulate complex urban flow and
concentrations from toxic releases at multiple
scales to be implemented into Canadian Reach
Back Services


X

X

X
X
X
X
Both
Mod-
erate
Y
Y
Y

Limited online
documentation,
model still in
improvement
stages, mainly
for Canadian
applications
43

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
30
D2-Puff
Innovative
Emergency
Management, Inc.
(IEM)
A puff/plume model originally designed in the
late 1980s as the D2PC model to estimate
downwind exposure values of toxic chemical
releases, especially those stored at LIS. Army
arsenals and DOD sites.
X
X




X


X
Both
Fast
N
N
Y

Now integrated
within the JEM
31
DEGADIS
Dense Gas Dispersion
Model
University of
Arkansas; EPA
An alternative EPA dense gas dispersion
model used to simulate the concentrations of
toxic chemical releases, especially for gases or
aerosols heavier than the ambient air, and
evaporating, upwardly, or zero-momentum
releases and jets over flat terrain.
X





X



Both
Fast
N
N
N


32
DELFIC/ FPTool
Defense Land Fallout
Interpretive Code/
Fallout Planning Tool
Oak Ridge National
Laboratory
(ORNL) and
Defense Nuclear
Agency
A nuclear fallout and cloud rise prediction and
consequence assessment software package,
built on SCIPLTFF dispersion model and
integrated within the Fallout Planning Tool,
used to predict radiological concentrations,
particle sizes, and dose rates resulting from
accidental radiological detonations

X






X

Preparedness
Fast
Y
N
Y/
N

One of the top
nuclear fallout
codes but hard
to find
information
from ORNL
33
DERMA
Danish Emergency
Response Model of the
Atmosphere
Danish
Meteorological
Institute
An operational emergency response, long-
range (20 km to global), 3D dispersion model
that incorporates hybrid stochastic (biological)
particle-puff diffusion that is integrated within
the Accident Reporting and Guidance System
(ARGOS), used primarily within Europe.


X



X
X
X

Both
Mod-
erate.
Y
N
N

LTsed primarily
within
Denmark and
Europe
34
DRIFT 3
Dispersion of Releases
Involving Flammables or
Toxics
LTK Health and
Safety Executive
(HSE)
Light and dense gas integral dispersion model
for simulating plumes from accidental instant
and continuous surface releases of toxic and
flammable substances
X




X
X



Preparedness
Fast
N
N
Y

Paid alternative
to DEGADIS
from UK
developers, but
extensively
peer reviewed
35
EPICode
Emergency Prediction
Information Code
Homann
Associates;
NARAC; LLNL
Software code that rapidly calculates source
terms based on material, height, duration, and
form, and neutrally buoyant downwind
concentrations of chemicals (gas, vapor, or
aerosol) released during hazardous industrial
and transportation accidents for use in DOE
applications
X
X




X
X


Both
Fast
N
N
Y/
N

Only for use in
" DOE
Emergency
Management
Issues Special
Interest Group
44

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
36
ESCAPE
Expert System for
Consequence Analysis
and Preparing for
Emergencies
Finnish
Meteorological
Institute
A simple Finnish internet browser-based
dispersion model and consequence analysis
tool used to rapidly estimate flammable and
hazardous continuous, instantaneous, and
ground-level gas and TIC plume releases to
inform emergency responders
X
X




X



Both
Fast
N
N
N

Developed for
the needs of the
Finnish
emergency
authorities
37
FEM3MP
Finite Element Model in
3-Dimensions and
Massively Parallelized
NARAC; DOE
A 3D, time-dependent, CFD-RANS, parallel
computing model used to investigate the
effects of turbulence, airflow, and dispersion
of chemical and biological agents released in a
complex urban environment under variable
winds




X

X
X


Preparedness
Slow
Y
Y
N

NARAC
integrated this
model within
another urban
dispersion
model
(AUDIM, now
Aeolus)
38
FLACS
FLame Acceleration
Simulator
Christian
Michelsen Institute
(CMI)
A CFD model used primarily within the oil
and gas industry to simulate the consequences
from fires, explosions, and toxic gas dispersal
out of industrial processing facilities




X

X


X
Preparedness
Mod-
erate.
N
N
Y

Not an
emergency
response
model, requires
costly purchase
39
Fluent
ANSYS. Inc.
Powerful physics-based, research-grade model
for a wide range of CFD applications (flow,
turbulence, heat transfer) of gases or particles
developed for a multitude of engineering uses




X

X



Preparedness
Slow
Y
Y
Y

Not
realistically
applicable for
emergency
response
40
FLEXPART
Flexible Particle
Dispersion Model
Institute of
Meteorology and
Climatology
(BOKU-Met).
Austria
A powerful and flexible long-range,
Lagrangian dispersion model used to simulate
forward or backward trajectories of particles,
gases, vapor, or radionuclides from source to
receptor (like HYSPLIT) and recently
incorporated into research grade weather
forecasting models.


X



X
X
X
X
Both
Mod-
erate
Y
N
N

Mainly for
research
purposes,
coupled to
models like
WRF
41
GENII V.2
Generalized
Environmental Radiation
Dosimetry Software
System - Flanford
Dosimetry System v.2
PNNL
A GUI package of radiological consequence
analysis software containing five independent
atmospheric, exposure, and dispersion models
to estimate chronic and acute dose and risk
from radionuclide releases in atmosphere or
water, developed for EPA exposure research
X
X






X

Preparedness
Fast to
Mod-
erate
N
Y
N

Not really an
emergency
response
model; used to
estimate risk
and exposure
from NRC sites
45

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
42
HASP
Hazard Assessment
Simulation and
Prediction Suite
UK Defence
Science and
Technology
Laboratory
fDSTLl Riskaware
Next-generation information management
suite of software tools and models to quickly
simulate CBRN dispersion in urban and rural
areas to permit emergency and military
personnel to more effectively respond and
contain hazardous releases for marine, cyber,
urban, and biological applications.
?





X
X
X
X
Both
Fast
(mins)
Y
Y
Y

Viable incident
modeling and
response
platform but
proprietary,
license needed,
and designed
forUK/EU
43
HGSYSTEM
Shell Research,
Ltd.
An alternative EPA dispersion modeling
system of several computer algorithms used to
simulate the source term and different types of
hazardous chemical and non-ideal gas
releases, especially dense gas (originally for
UF6). Includes HEGADIS model.'
X





X



Preparedness
Fast
(1-10
mins)
N
N
N

Like
DEGADIS in
many ways; no
recent updates;
limited
emergency.
response
applications
44
HIGRAD/ FIRETEC
High-Resolution Model
for Strong Gradient
Applications Fire
Behavior Model
LANL and LTnited
States Department
of Agriculture
(USDA) Forest
Service
Physics-based 3D code to simulate constantly
changing interactions between forest fires,
wind flows, fuels, and complex topography.




X

X



Preparedness
Slow
Y
N
Y

Research tool
only; main
application is
for forest fires
45
HOTMAC and
RAPTAD
Higher Order Turbulence
Model for Atmospheric
Circulation Random Puff
Transport and Diffusion
Yamada Science
and Art (YSA)
Corporation
An alternative EPA-preferred 3D Eulerian
weather model coupled with a puff dispersion
model to simulate pollutant flow and
dispersion throughout complex terrain and
simple urban areas

X




X
X


Preparedness
Fast
Y
Y
Y

No recent
updates or
model support;
use phased into
Atmosphere to
CDF (A2C)
Model
46
HotSDot
NARAC; LLNL
Fast running, field-portable dispersion
modeling tools developed for emergency
response personnel and planners to provide a
close-range (< 10 km), conservative estimate
of releases from radiological incidents.
X







X

Both
Fast
Y
N
N


46

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
47
HPAC
Hazard Prediction and
Assessment Capability
DTRA
A comprehensive, robust, operational, and
research-grade CBRN dispersion modeling
system built upon the SCIPUFF model
foundation that predicts the effects of
hazardous releases for civilian and military
populations by integrating high resolution
weather data and modifications for dense gas
and urban parameterizations (from Urban
Dispersion Model [UDM])

X




X
X
X
X
Both
Mod-
erate
Y
Y
Y/
N


48
HYROAD
Hybrid Roadway
Intersection Model
National
Cooperative
Highway Research
Program (NCHRP)
Hybrid roadway puff model that predicts
concentrations of carbon monoxide (CO) and
PM from vehicle emissions at receptors within
500 meters of roadway intersections.

X




X



Reparedness
Fast
N
N
N

Limited to no
emergency
response
application
49
HYSPLIT
Hybrid Single-Particle
Lagrangian Integrated
Trajectory
NOAA Air
Resources
Laboratory (ARL)
NOAA's robust dispersion modeling system
that calculates forward and backward air
parcel trajectories, pollutant transport,
chemical transformation, and deposition of
particles, gases, or aerosols that can be run
interactively through an internet browser or
downloaded to a computer. HYSPLIT uses
high-fidelity weather data for local or long-
range dispersion (>1000 miles) with
applications for emergency response.


X



X
X
X

Both
Fast
(sees)
Y
N
N


50
INPUFF
Gaussian Integrated Riff
Model
EPA
A simple, single stationary or moving source
Gaussian Riff model that calculates
downwind concentrations from deposition and
settling at up to 25 receptors from neutrally
buoyant gases released from stack or jet
sources.

X




X



Reparedness
Fast
N
N
N

Not recently
updated,
replaced by
newer models
like AERMOD
51
ISC3
Industrial Source
Complex Model 3
EPA
Alternative EPA steady state Gaussian model
used to assess pollutant concentrations from a
large number of industrial complex emission
sources, including deposition and downwash
from stacks.
X





X



Reparedness
Fast
N
Y
N

Replaced by
AERMOD; not
an emergency
response
dispersion
model
47

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
52
JEM
Joint Effects Model
DOD; Aeris, LLC.
A comprehensive, DOD-accredited, web-
based, operational dispersion modeling
software built upon SCIPLTFF used to simulate
accidental or intentional CBRN incidents and
weapon strikes across the LIS. Military with
advanced capacities for complex terrain, TICs,
human health indications, and urban
environments, encompassing many standalone
dispersion model codes

X




X
X
X
X
Both
Fast
Y
Y
Y


53
JOULES
Joint Outdoor-indoor
Urban Large-Eddy
Simulation
Aeris, LLC and
Lawrence Berkeley
National
Laboratory (LBNL)
An experimental physics-based LES modeling
system that produces high-fidelity simulations
of urban and indoor contaminant dispersion
for use in operational urban emergency
response tools such as the HP AC model,
where it is slated to identify performance
limitations.




X
X
X
X
X
X
Preparedness
Slow
(mins
to a
few
hrs)
N
Y
Y

Experimental
and currently
research grade
only
54
KBERT
Knowledge-Based-
system for Estimating
hazards of Radioactive
material release
Transients
Sandia National
Laboratory (SNL)
A risk analysis tool containing a basic
dispersion model, based on stability class,
used to estimate the risks and doses for in-
facility workers and the nearby public exposed
to accidental releases from chemical and
nuclear facilities
X





X

X

Response
Fast
N
N
N

Mostly for
NRC use, less
related to
emergency
responders; not
recently
updated
55
KDFOC4
"K" Division Defense
Nuclear Fallout Code
LLNL; NARAC
A nuclear fallout module now incorporated
within NARAC's assessment capability that
calculates the spread of gamma radiation
produced during above or below-ground
fission-source detonations by calculating time
and weather dependent plume rise
X







X
X
Both
Fast
N
N
Y

Not an
emergency
response
dispersion
model
56
LAPMOD
LAgrangian Particle
MODel
Enviroware, Italy
3D Lagrangian dispersion model used to
simulate dispersion and transport of gases,
odors, and inert or radioactive particles over
complex terrain from local meteorology.


X



X
X
X

Preparedness
Fast
Y
N
N

Mainly
research grade
with not much
emergency
response use
57
LPDM
Lagrangian Particle
Dispersion Model
National Center for
Atmospheric
Research (NCAR)
A research-grade model most recently
combined with NCAR's EUlerian
LAGrangian (EULAG) LES model used to
simulate realistic turbulent environments and
hazardous release scenarios based on
traditional Lagrangian particle dispersion.


X



X
X
X
X
Preparedness
Mod-
erate
Y
N
Y

Research grade
and largely
incorporated
within NCAR
EULAG model
48

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
58
MATHEW/ADPIC
Mass-Adjusted Three-
Dimensional Wind Field/
Atmospheric Diffusion
Particle-in-Cell
Atmospheric
Release Advisory
Capability
(ARAC). LLNL
Operational 3D wind model coupled with a
Lagrangian random walk dispersion model to
assess the impact of neutrally buoyant,
hazardous first order chemical and
radiological releases


X



X

X

Both
Fast
(mins)
Y
N
Y

Replaced by
newer model
(LODI) from
NARAC
59
MDIFF
NOAA Air
Resources
Laboratory Field
Research Division
Mesoscale emergency response puff model
used to calculate the transport and dispersion
of airborne material releases near Idaho
National Laboratory (INL), informed by local
weather Mesonet and an offspring of original
MESODIF model "

X




X



Response
Fast
N
N
N

Site specific for
use at INL, not
recently
updated
60
MELCOR and MACCS
MELCOR Accident
Consequence Code
System
SNL
MACCS is a comprehensive, straight-line
Gaussian plume model package used to
simulate the ecosystem and human dose and
exposure impacts of severe nuclear power
plant accidents, widely used across DOE
facilities from MELCOR model output.
X







X

Both
Fast
N
N
N


61
MIDAS-AT
Meteorological
Information Dispersion
and Assessment System
Anti-Terrorism
ABS Consulting;
PLG Inc.
An anti-terrorism puff dispersion modeling
system capable of simulating potential hazard
areas and aftereffects caused by a chemical or
biological agent attack inside a building or
urban area, including the spread of an agent
between floors and rooms of a building and
throughout the urban street canyon.

X




X
X
X

Both
Fast
Y
Y


Limited
information
available, must
purchase
62
MSS
PMSS
(Parallelized) Micro-
Swift Spray
Aria Technologies,
France and SAIC
A CFD-like 3D dispersion model coupled
with ARIA View designed to simulate
complex urban and industrial dispersion by
generating mass-constant streamlines and gas
or particle plumes around obstacles


X



X


X
Both
Mod-
erate
(min to
<1 h)
N
Y
Y

Mostly French
and EU
applications,
requires
purchase
63
NAME III
Numerical Atmospheric-
Dispersion Modeling
Environment
UK Met Office
A sophisticated 3D, random walk, short-to-
long range dispersion model used in research,
operational, and LTK emergency response
situations that employs flexible 3D
meteorological inputs, unlimited sources, and
forward/backward simulations, etc.


X



X
X
X

Both
Mod-
erate
N
N
N

Mainly used by
UK MetOffice
but available
for external
research use
with license
49

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
64
OBODM
Open Bum/Open
Detonation Dispersion
Model
U.S. Army,
Dugway Proving
Ground
Alternative EPA model that predicts
downwind transport, dispersion, and air
quality impacts using existing plume rise and
dispersion algorithms from open burning and
detonations of obsolete munitions and uses
algorithms from the Real-time Volume Source
Model (RTVSM)
X





X


X
Preparedness
Fast
Y
N
N

Limited
emergency
response
applications
65
OCD 5
Offshore and Coastal
Dispersion Model
EPA
Line, point, and area source dispersion model
to determine the impact of offshore emissions,
plume, and air quality near coastal regions,
and one of EPA" s preferred and recommended
models.
X





X



Preparedness
Fast
Y
N
N

Not updated in
many years;
most features
now in
AERMOD
66
OMEGA/ADM
Operational Multiscale
Environment Model with
Grid Adaptivity /
Atmospheric Dispersion
Model
SAIC
OMEGA is an operational multiscale
numerical weather prediction model
embedded with an atmospheric dispersion
model for use at adaptably large (Eulerian) to
small (Lagrangian) spatial scales with many
types of parameterizations to simulate gas and
particle transport


X
X


X
X


Preparedness
Slow
Y
N
Y

Not recently
updated/
replaced with
newer model;
Minimal
internet
presence
67
One
SAFER One
SAFER One FlazMat
Response
SAFER Systems
A suite of real-time, cloud-based emergency
modeling software used to monitor, simulate,
and mitigate chemical incidents by allowing
users to collaborate across platforms.
Designed for emergency responders, and in
many ways similar to HASP.

X




X



Both
Fast
Y
Y
Y

Viable real-
time response
platform, but
proprietary,
requires license
and doesn't
share methods
68
OSPM
Operational Street
Pollution Model
National
Environmental
Research Institute
of Denmark,
Aarhus University
An advanced Danish plume and box model
used to predict air quality (CO, PM, NOx)
inside urban street canyons from traffic
emissions from source to receptor by
considering building geometry, urban
turbulence, and chemical conversions.
X




X
X



Preparedness
Fast
N
Y
N

No CBRNe
applications
and not an
emergency
response model
69
PANACHE
Atmosphere Pollution
and Industrial Risk
Analysis
French Ministry
and Environmental
Agency (ADEME)
and Fluidyn/
Transoft
French proprietary suite of 3D finite fluid
mechanics modules for industrial, urban, and
complex terrain applications of hazardous
accidental or continuous releases




X

X



Preparedness
Mod-
erate
Y
Y
Y

Requires user
to pay
consultant from
Fluidyn for risk
analysis
50

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
70
PAVAN
Battelle; PNNL
Gaussian plume model used to calculate short-
term, ground-level, downwind radiological
concentrations from accidental, design flaw-
related nuclear power plant releases
X







X

Preparedness
Fast
N
N
N

Not an
emergency
response model
71
PHAST
Process Hazard Analysis
Software
DNV Software. UK
Process analysis and hazard consequence tool
for mainly industrial sites that examines the
behavior of an incident from an initial release
to far field dispersion of leaks, ruptures, spills,
and toxic clouds.
X
X




X



Preparedness
Fast
X
X
Y

Flazard analysis
software rather
than an
emergency
response model
72
PLUVUEII
Plume Visibility Model
EPA
Alternative EPA dispersion model that
calculates the visual range and atmospheric
discoloration (opacity) of plumes caused by
single S02 or NO combustion emission
sources in Class I (wilderness) areas
X





X



Preparedness
Fast
N
N
N

No emergency
response
application
73
PUFF-PLUME
PNNL
Emergency puff and continuous plume model
that predicts chemical pollution and
radionuclide transport, wet/dry deposition,
exposure pathways from an accidental release.
X
X






X

Response
Fast
N
N
Y

Site specific to
Savannah River
Site
74
PUMA
Riff Model for
Atmospheric Dispersion
Swedish Defence
Research Agency
(FOlf
A real-time puff model using Lagrangian
dispersion trajectories, with neutral and dense
gas chemical capabilities, designed for third
party integrations such as FOI's "Dispersion
Engine" software package

X




X



Preparedness
Fast
N
N
N

Mainly used in
EU; still
undergoing
development
and evaluation
75
OUIC
Quick Urban Industrial
Complex Model
LANL
Relatively fast-response model that computes
various CBRNe agent dispersals, including
dense gas, particles, jets, and explosions, on
the urban building-to-neighborhood scale with
the ability to track dispersion and flow fields
around buildings and structures.


X



X
X
X
X
Preparedness
Mod-
erate/
Fast
(sees -
hours)
Y
Y
N


76
RaDidAir
Ricardo Energy and
Environment, LTK
City-scale, Python-based dispersion modeling
system using AERMOD coupled with street
canyon model equations where model output
kernels are passed over roadway emissions
sources (NOx) to simulate urban air quality
X





X



Preparedness
Mod-
erate
(mins)
Y
Y
N

Mainly for
traffic
emissions, not
an emergency
response model
51

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
77
RASCAL 4.3.3
Radiological Assessment
System for Consequence
Analysis
U.S. NRC
Consequence assessment tool that uses the
RATCHET dispersion model for radiological
releases from nuclear facilities and
powerplants to determine source terms,
transport, dose, potential downwind effects,
and whether to evacuate or shelter in place.

X






X

Both
Fast
Y
N
N

Mainly used by
the Protective
Measures Team
of NRC for
power plants
and storage
facilities
78
RIMPUFF
Riso Mesoscale Riff
Model
Riso National
Laboratory
(Denmark)
An advanced, extensively tested, near real-
time mesoscale (<100 km) emergency
response model used primarily within Europe
to predict the transport and dispersion of
CBRN materials and is also incorporated
within European emergency centers and
response systems (i.e., ARGOS).

X




X
X
X
X
Both
Fast
Y
Y
N

Operationally
incorporated in
decision
support
systems but
primarily
within Europe
79
RLINE
Research Line-source
Dispersion Model
EPA
A research-grade, line-source dispersion
model used to evaluate chemically inert air
quality impacts in the near-road environment
from mobile sources along and nearby to
major roadways using AERMOD's
meteorology preprocessor.
X





X



Preparedness
Fast
N
N
N

For traffic
related
emissions, not
an emergency
response model
80
RSAC 7.2
Radiological Safety
Analysis Computer
Program
INL
Modified-Gaussian plume program that
calculates the dose, inhalation, ingestion, and
air immersion consequences from upwind
atmospheric radionuclide releases at nuclear
powerplant facilities from accidental or
sabotage scenarios on a personal computer
X






X


Both
Mod-
erate
N
Y
N

Mainly for use
at INL but can
be applied to
exposure of
fission products
elsewhere
81
RTDM3.2
Rough Terrain Diffusion
Model
ERT
A Gaussian model to estimate ground-level
concentrations of chemically stable pollutants
and buoyant plume behavior in areas of rough
or flat terrain in the nearby vicinity of one or
more collocated point sources.
X





X



Preparedness
Fast
Y
N
N

Not updated
since 80s;
limited CBRNe
application
52

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
82
SCIPUFF
SCICHEM
Second-order Closure
Integrated Riff Model
SCIPUFF with chemistry
Titan Corporation;
Sage Management
(Xator Corp.)
Electric Power
Research Institute
(EPRI)
An alternative EPA second-order closure puff
diffusion model used to simulate sequences of
3D, time-dependent puffs from a wide variety
of source geometries and types with flexible
meteorology inputs. The chemistry version
models the transport, dispersion, and chemical
reactions of gases and aerosol releases from
single or multiple sources. SCIPLTFF is the
transport and dispersion code of HP AC, JEM,
and is also integrated with other models.

X




X



Both
Mod-
erate
Y
Y
N

See HP AC or
JEM entry
83
SDM
Shoreline Dispersion
Model
EPA
An alternative EPA model used to determine
ground level concentration from tall stationary
point sources influenced by meteorological
phenomena near shoreline environments
affecting plume behavior and fumigation.
X





X



Preparedness
Fast
Y
N
N

Not emergency
response
related
84
SHARC/ERAD
Specialized Hazard
Assessment Response
Capability/Explosive
Release Atmospheric
Dispersion
SNL
A suite of five models (Nuke, AIRRAD,
Blast, and ERAD/PLTFF, MCK) that simulates
the release of radioactivity from nuclear
weapon explosions or detonations. The
Gaussian puff model, ERAD, is used to
predict the radiological detonation dispersion
and to assess time dependent, dynamic
explosive buoyant plume rise for exposure and
evacuation criteria.

X






X
X
Both
Fast
N
N
N


85
SIRANE
Atmosphere,
Impact & Risk
(AIR). Ecole
Centrale de Lyon,
France
The first and currently only fine-scale street-
network dispersion model designed to
simulate the flow and dispersion through a
network of interconnected streets with a
Gaussian approach to the adjacent urban
boundary layer above the street canopy.
X




X
X



Preparedness
Fast
N
Y
N

Currently in
development
stages; mainly
for European
city geometries
86
SLAB
LLNL
Alternative EPA dense gas model also
incorporated in ALOHA and ADAM Tool to
simulate jet, volume, evaporating pool, and
volume continuous or instant releases from
accidental or intentional episodes.
X
X




X



Both
Fast
(mins)
N
N
N

See ALOHA or
ADAM Tool
entries
53

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
87
STILT
Stochastic Time-Ill verted
Lagrangian Transport
Model
Harvard University,
MPI-Jena,
University of
Waterloo, and
Atmospheric &
Environmental
Research (AER)
A research-grade, Lagrangian particle
dispersion model used to derive upwind
source region concentrations and fluxes, such
as greenhouse and trace gas releases, based on
fixed downstream measurement receptors and
driven by high resolution weather prediction
models


X



X



Preparedness
Fast
Y
N
N

Generally for
air pollution
applications;
not an
emergency
response model
88
TAPM
The Air Pollution Model
Commonwealth
Scientific and
Industrial Research
Organisation
(CISRO), Australia
An advanced 3D model coupled with a
weather and Lagrangian particle model to
simulate the dispersion of emissions sources
in local-to-urban areas, including plume rise,
building wakes, and atmospheric chemistry.


X
X


X



Preparedness
Mod-
erate
Y
Y
N

For air
pollution; not
an emergency
response model
89
TRACE
Toxic Release Analysis
of Chemical Emissions
SAFER Systems
Consequence assessment chemical mass-
balance tool to simulate and visualize airborne
hazard material releases from chemical
incidents, including sprays and dense gas, to
update risk assessments and EPA RMP plans
at chemical sites.





X
X


X
Preparedness
Fast
N
N
Y

Risk
assessment; not
an emergency
response
model; See
One Model
90
UDM
Urban Dispersion Model
UK DSTL
The urban-based model currently incorporated
within HP AC that modifies a plume based on
street alignment and building density in urban
areas but does not resolve dispersion around
individual buildings.

X




X
X


Preparedness
Mod-
erate
Y
Y


See HPAC
entry
91
UoR-SNM
University of Reading
Street Network Model
University of
Reading, UK
A research-grade, street network urban
dispersion model similar to the operational
SIRANE model, but without flow parameters,
that represents particle flow within an urban
area as a system of connected boxes at
intersections.





X
X



Preparedness
Slow
N
Y
Y

Requires LES
flow fields;
research grade
so not realistic
for emergency
response
92
VALLEY
EPA
An EPA alternative steady-state screening tool
for rural and complex terrain to estimate 24-h
average pollutant concentrations for point or
area sources (stacks or industrial areas),
related to predecessor VALDRIFT model.
X





X



Preparedness
Fast
Y
N
N

Screening
dispersion
model only
93
VAPO
Vulnerability Analysis
and Protection Option
DOD. DTRA,
Applied Research
Associates (ARA)
A 3D vulnerability and risk assessment
software tool (rather than a dispersion model)
that predicts effects of structural damage,
injury, and human risk from terrorist related
CBRNe blasts at building sites in urban areas.





X



X
Both
Fast
(mins)
Y
Y
Y

Assesses risk
and structural
impacts from a
blast rather
than dispersion
54

-------




Model Type
CBRNe Type






4)
#
Model
Full Name and
Link to Best
Source
Developer
Description
Gaussian Plume
Gaussian Puff
Lagrangian
Stochastic Particle
Eulerian Grid
CFD
Other
Chemical
Biological
Radiological/Nuclear
Explosive
Emergency
Response
Stage
Speed
Terrain Effects?
Building Effects?
Proprietary?
Classification
Criteria
Reasons for not
including in detai
review
94
VENTSAR XL
VENTSAR-Excel
Westinghouse
Savannah River
Company
An Excel-based Gaussian dispersion model
that incorporates plume rise and building
effects, used to determine downwind doses
and risk from exhaust effluent.
X





X

X

Preparedness
Fast
N
Y
Y

Not recently
updated; not
emergency
response;
designed for
SRNL
95
VLSTRACK
Vapor, Liquid, and Solid
Tracking
LIS. Naval Surface
Warfare Center
Hazard prediction model used by DOD to
provide downwind hazard predictions for a
range of chemical and biological warfare
agent attacks, including munitions.

X




X
X


Preparedness
Mod-
erate
N
N
Y

Incorporated
within JEM
model
96
XOODOO
PNNL
Gaussian dispersion model used to calculate
long-term, routine, intermittent, or expected
release concentrations and depositions at
radial distances up to 50 miles out from
nuclear reactor site.
X







X

Preparedness
Fast
N
N
N

Not an
emergency
response
model, retired
55

-------
8.0 Expanded Model Descriptions
8.1 ADAM Tool
Accident Damage Analysis Module (ADAM) Tool
Developer
Joint Research Centre (JRC) of the European Commission (EC), Major
Accident Hazards Bureau (MAHB)
Type of Model
Gaussian Puff and Plume Dispersion Model
Response Stage
Emergency Preparedness
Original
Application
Chemical and explosive releases from hazardous industrial accidents

The ADAM Tool is a software package developed by the EU's JRC to assess
the consequences and damages associated with an accidental, hazardous
industrial chemical release. ADAM is designed to be a comprehensive
consequence assessment tool to simulate toxic airborne concentrations and
exposures from chemical fires, explosions, and gaseous cloud releases from
industrial facilities for prevention and preparedness. The model can support
industrial risk management, land use and emergency planning, enforcement of
regulations, inspection and monitoring, and identify weak areas for site
improvement (Fabbri and Wood 2019). It contains an extensive database of
substances, their physical properties, and exposure effects (i.e., LD50 and
IDLH). The ADAM Tool can calculate the physical hazard situations and
human health impacts that may arise from thermal radiation, over-
pressurization of tanks, flammable releases, explosions, and loss of
containment of a toxic chemical. The model contains a GIS mapping tool to
assess spatial risk of the affected area.
Model Description
ADAM contains three modules that track the dangerous substance from loss of
containment to impact on affected populations. The first module requires the
source term, including the amount released, flow rate, and thermodynamic
state of the released agent. The second module estimates the physical effects
from the release (i.e., fires, explosions, toxic clouds) and its local dispersion.
The vulnerability is calculated in the third module to inform the potential level
of harm to exposed individuals based on intensity, dose, and exposed duration
for the specific release to initiate protective action and lifesaving measures.
The dispersion modeling component is built upon the existing and well
verified SLAB Gaussian Puff/Plume model developed by LLNL. SLAB is
commonly applied to dense gas scenarios, although it can simulate neutrally
buoyant and lighter than air releases. ADAM can model continuous, finite, and
instantaneous releases from source types including ground-level evaporating
pools (area releases), horizontal or vertical jets, and stacks or elevated releases.
All effluent can be gases, aerosols, or a combination of liquids and gases. The
SLAB code was rewritten and streamlined into the ADAM Tool. A
comprehensive model evaluation was performed by Fabbri and Wood (2019)
by conducting a series of relevant release scenarios and benchmarking the
results with similar software and experimental field datasets. The ADAM Tool
was found to simulate various release scenarios well using the default model
56

-------

options. The most recent evaluation has been done with the Jack Rabbit II
(JRII) chlorine field study dataset (Fabbri et al. 2020).
Pros
Modern, evaluated modeling tool built upon a well-used dispersion modeling
platform; calculates vulnerability and physical health effects
Cons
Does not consider environmental consequences; primarily used for emergency
preparation within EU nations; requires detailed information about the release
Runtime
Fast
Input Data
Requirements
General knowledge of meteorological conditions; detailed specifics about the
release mechanism and agent
Outputs
Dispersion plume of effluent and hazard area contour maps; vulnerability and
physical harm regions for exposed individuals; graphs of relevant parameters
from the release; lethality curves
Data assembly
requirements
during or after
emergency
response
Knowledge of meteorological conditions; release mechanism and agent
information
Code language
C++
Public or
Proprietary, Cost
The model is primarily an EU tool used to support implementation of the
Seveso Directive (control of major hazardous accidents). It is available to EU
countries or other regulators associated with chemical safety and security.
However, it is also available to Organisation for Economic Co-operation and
Development (OECD) countries (the U.S. is an OECD country). Distribution
is made on request to interested government users (and some non-commercial
research users) that fit these criteria. It is not available to consultants.
Ease of use
Not known
Ease of obtaining
information and
availability of
technical support
General queries can be sent to: IRC- fof®ec.eurooa.eu
Source code
availability
No
Installation
requirements/
software
Not known
Maintenance Status
The model was launched in 2019 and is available to interested counties and
government organizations (Fabbri and Wood 2019)
Documentation
The technical guidance document is available at:
httos://oublications.irc.ec.eurooa.eu/reoositorv/bitstrea /kina2873

2enn.pdf
Link to Website
https://adam.irc.ec. europa.eu/ em/adam/content
57

-------
8.2 ADAPT/LODI
Atmospheric Data Assimilation and Parameterization Tool (ADAPT)/ Lasransian
Operational Dispersion Integrator (LODI)
Developer
National Atmospheric Release Advisory Center (NARAC), Lawrence
Livermore National Laboratory (LLNL), DOE
Type of Model
Lagrangian Particle Dispersion Model
Response Stage
Both Emergency Preparedness and Response
Original
Application
Various CBRNe releases for operational use throughout urban or rural areas
Model Description
ADAPT/LODI is a 3-D, Lagrangian, operational transport and diffusion model
that calculates possible trajectories, concentrations, and depositions of fluid
"particles" in a turbulent flow. The particles are intended to represent various
types of hazardous CBRNe releases, ranging from thermal or momentum
driven releases from stacks or fires, to detonations from chemical explosives or
nuclear sources. The model is NARAC's chief operational emergency
response resource for IMAAC's plume generation service. The system
contains two models: 1) ADAPT, which is used to construct 3D meteorology
fields for use in 2) LODI, the Lagrangian dispersion model. ADAPT develops
key meteorological parameters including winds, temperature, pressure,
humidity, and precipitation. These variables are obtained from the most recent
NWS observations (such as airport sites, weather balloons, and weather
networks) when results are needed instantaneously. For extended or ongoing
atmospheric releases, gridded model datasets or other weather models, such as
WRF, may be used (Nasstrom et al. 2007). ADAPT creates wind fields using
the finite element method (a method of solving equations over a large area
divided into smaller and simpler parts), which is also beneficial over
nonhomogeneous and complex terrain. ADAPT can produce input for LODI
within one minute (Bradley 2005).
The LODI model employs a Lagrangian stochastic Monte Carlo approach
(which calculates an average based on a nearly Gaussian distribution of
atmospheric turbulence) and then solves the 3D advection-diffusion equations.
The model can produce a time series of instantaneous and time-integrated
effluent concentrations and depositions, as well as a detailed plume within 5-
15 minutes. The model can simulate dispersion for a variety of spatial and
temporal scales, including dispersion over regional to local scales. LODI can
integrate multiple point, line, area, spherical, or moving sources, including
variable emissions rates. Particle size distributions, radiological decay, wet and
dry deposition, and resuspension algorithms are also incorporated within the
model. Results can be output to GIS mapping tools where spatial analyses can
inform responders of protective action zones, exposure guidelines, and regions
where doses exceed safe levels. ADAPT/LODI has evolved from the
MATHEW-ADPIC and ARAC2 models since the 1990s.
2 See information about ARAC at: https://narac.Hnl.gov/content/mods/piiblications/op-model-description-evatiiation/UCRL-
JC-125034.pdf
58

-------
Pros
Produces results rapidly; proven to be an effective operational model;
Cons
Not fine scale enough to predict dispersion at the street or neighborhood level
within urban areas
Runtime
Generally fast, within 5-15 minutes depending on the domain
Input Data
Requirements
Location of the release and source characteristics
Outputs
Processed outputs result in maps of air or ground contamination, dose, and
health effects resulting from the release, including protective action zones
Data assembly
requirements
during or after
emergency
response
Location of the release, local meteorology, and source characteristics
Code language
Unknown
Public or
Proprietary, Cost
Proprietary, but use may be granted for some research and development
applications
Ease of use
Moderate
Ease of obtaining
information and
availability of
technical support
Ouestions can be directed to owner-narac-web-SDtfoHistserv.llnl.gov who will
forward the request to the appropriate individual
Source code
availability
No
Installation
requirements/
software
Unknown
Maintenance Status
Currently used as an operational model within NARAC for IMAAC
Documentation

Link to Website
https://narac.llnl.80v/tools/operational-modelin8/dispersion-model-lodi
59

-------
8.3 Aeolus
Aeolus
Developer
NARAC (National Atmospheric Release Advisory Center), Lawrence
Livermore National Laboratory (LLNL); DOE
Type of Model
Computation Fluid Dynamics (CFD) Model
Response Stage
Emergency Preparedness
Original
Application
CBRNe CFD model for complex terrain and urban research applications

Aeolus is NARAC's primary research and development model that simulates
high resolution flow and dispersion of hazardous material through urban areas
and complex terrain environments. The model, which is generally used for
emergency planning guidance, is a physics-based and building-resolving CFD
code based on the finite volume method (solving equations on the small
volume surrounding a point on the computational mesh/grid). Aeolus is used
within NARAC's operational emergency response applications alongside
ADAPT/LODI, but mainly for emergency planning guidance. Even though
Aeolus is still a research-grade model, it is being phased into operational use
for the generation of urban products for state and local agencies though
IMAAC (Gowardhan et al. 2018). The model can simulate releases from
nuclear power plant accidents, detonations, toxic industrial chemical spills,
RDDs, and biological and chemical agents.
Model Description
Aeolus can be run under a fast, operational mode using a RANS solver for
potential operational use or when many simulations are needed. Alternatively,
it can be run at high resolution through the more detailed LES method for
research and planning. The operational mode can produce results within 5-10
minutes on a laptop, but the LES simulation takes several hours. As with other
RANS models, Aeolus solves the incompressible Navier-Stokes equations on a
staggered Cartesian grid. Aeolus RANS consists of a solver to produce the
steady state wind and turbulence fields as well as a Lagrangian dispersion
model to predict the contaminant dispersion throughout the urban morphology.
Radiological source terms and half-life behaviors have also been integrated
into Aeolus based on explosive plume rise. The model can also simulate
buoyant and dense gases and particles. To facilitate faster model setup times in
urban areas, building profile domains have been generated and stored within
NARAC's geographical database for over 130 cities across the US. Terrain
data are also available on a 10 m grid across the US. Meteorology can be input
through forecast model data (i.e., HRRR, NAM, GFS), or through a wind
profile specified by the user. Aeolus has evolved from the FEM3MP model to
AUDIM over the past several years. It has been extensively evaluated against
the Joint Urban 2003 field study and shown to produce good agreement (Lucas
et al. 2016).
Pros
RANS model generally has fast runtimes; resolves building profiles for urban
dispersion; Evaluated against field data and showed good agreement
60

-------
Cons
Highest resolution simulation could take hours; mainly for research and
development purposes
Runtime
Variable depending on simulation choice; RANS simulation about 5-10
minutes, high resolution LES takes several hours on laptop
Input Data
Requirements
Latitude and longitude of the release, domain size, resolution, period of
simulation, and details about the source, meteorology
Outputs
Time evolving spatial plots (exportable to GIS mapping software) of the
dispersion of particles downwind of release; 3D deposition on surfaces,
effective dose and hazard zones near release
Data assembly
requirements
during or after
emergency
response
Location of the release, source characteristics, and meteorology (expected to
take only about 2 minutes)
Code language
Unknown
Public or
Proprietary, Cost
Proprietary, but use may be granted for some research and development
applications
Ease of use
Moderate
Ease of obtaining
information and
availability of
technical support
Ouestions can be directed to owner-narac-web-SDtfoHistserv.llnl.gov who will
forward the request to the appropriate individual
Source code
availability
No
Installation
requirements/
software
Unknown
Maintenance Status
Currently being used and developed by NARAC
Documentation
See website for more information
Link to Website
httos://narac.Hnl.gov/research-and-develoDment/urban-disDersion-modeling
61

-------
8.4 AERMOD
American Meteorological Society/Environmental Protection Agency Regulatory
Model (AERMOD)
Developer
U.S. EPA and American Meteorological Society (AMS); Developed by
AERMIC (American Meteorological Society/EPA Regulatory Model
Improvement Committee)
Type of Model
Gaussian Plume Dispersion Model
Response Stage
Both Emergency Preparedness and Response
Original
Application
Gaussian plume model to determine regulatory, source permitting, and
downwind concentrations from source to receptor in steady state conditions
Model Description
AERMOD is EPA's preferred and recommended Gaussian dispersion model to
simulate the concentration of gases and particles at downwind receptors from
surface and elevated stationary sources (Cimorelli et al. 2005). It is a steady-
state model for use in various atmospheric stability conditions based on the
PBL structure. The model incorporates a well-established boundary layer,
scaling, and turbulence concepts and parameterizations. Under stable
atmospheres and low turbulence conditions, the model applies a Gaussian
approach. During unstable, convective periods, it uses a non-Gaussian method
for the vertical component of the plume. AERMOD includes special treatment
for single or multiple point, area, and volume sources. It accounts for plume
rise, the effects of building downwash, complex terrain for point sources,
limited interactions within urban areas, and wet and dry deposition. The model
produces concentrations for an array of downwind receptors. The user can
specify the quantity and density of the receptor sites for the most appropriate
dispersion representation. AERMOD is EPA's primary regulatory dispersion
model to assess concentration fields at emission sites. It is specifically used for
New Source Review (to issue emission source permits, such as at industrial
locations), to develop State Implementation Plans, formulate mitigation plans
for non-attainment areas using NAAQS, and to generally evaluate the effects
and behavior of downwind effluent dispersion.
AERMOD simulations are set up with the use of two data input preprocessors.
AERMET is the meteorological preprocessor that defines the meteorological
state of the PBL. AERMAP is a terrain data preprocessor that implements U.S.
Geological Survey (USGS) Digital Elevation Data and has algorithms that
determine the terrain features used by AERMOD. Other preprocessors may
optionally be used. AERSCREEN can rapidly run the AERMOD algorithms
with pre-selected meteorology as a screening tool to decide if a full simulation
is needed. AERSURFACE accounts for land-use and land-cover to develop the
surface characteristics (friction velocity, Bowen ratio, and albedo), and
BPIPPRM incorporates multiple building dimensions near the source to
provide an effective building for building downwash calculations. AERMOD
simulates the effects of single buildings adjacent to the source, but generally
lacks robust urban flow field capabilities. Its development was strongly
influenced by micrometeorological theory as well as research and development
from field and wind tunnel studies. The model has also been extensively
62

-------

evaluated through field tests. AERMOD has replaced or incorporated many
older models such as BLP and 0CD5. It was originally promulgated as a
replacement to ISCST33.
Pros
Fast runtimes; widely supported by EPA as the preferred regulatory model for
source permitting, SIP analysis, and traffic conformity studies; free; theoretical
concepts supported by field and laboratory studies
Cons
Susceptible to all limitations of Gaussian plume models; may underpredict
concentrations in some situations; model setup may be somewhat challenging
for some users; limited to downwind receptor distances of about 20-50 km;
does not account for different types of CBRNe releases
Runtime
Fast; within seconds, but depends on number of sources, receptors, and
simulation periods
Input Data
Requirements
Meteorological state of the PBL (e.g. wind, temperature, stability), surface and
terrain characteristics, source location and release characterization, location of
the downwind receptors
Outputs
Concentrations at downwind receptors
Data assembly
requirements
during or after
emergency
response
Local wind speed and direction near the source to construct a vertical wind
profile, effluent source characteristics
Code language
FORTRAN
Public or
Proprietary, Cost
Free through EPA's SCRAM website; companies such as Lakes
Environmental (httos://www. weblakes.com/oroducts/aermod/index.htmn.
Breeze Software, and Enviroware offer AERMOD within more user-friendly
GUI windows, but the cost is not insignificant (over $1,600 for AERMOD
View by Lakes). Some companies offer free accounts for EPA or government
employees
Ease of use
Moderate, runs from a Windows command line prompt. The model is easier to
run if used through paid GUIs
Ease of obtaining
information and
availability of
technical support
Support for the EPA SCRAM website can be obtained by contacting George
Bridgers: bridgers.george®,eoa.gov. The SCRAM website posts a wide range
of support documents, test cases, and evaluation reports. Many companies also
provide consulting services
Source code
availability
Yes, available on EPA's SCRAM website along with executables
Installation
requirements/
software
32- or 64-bit Windows PC
3 See: https://www.epa.gOv/scram/air~qnalitv-dispersion~modeling~prererred~aiMl~recommetMled~models#aermod
63

-------
Maintenance Status
Continuously updated and improved by EPA, most recent version as of mid-
2020 is AERMOD vl9191
Documentation
A comprehensive user guide is available at:
httos://www3.eoa.gov/ttn/scram/models/aermod/aermod userguide.odf.

Several quick reference guides are also available on the SCRAM website
Link to Website
httDs://www.eoa.gov/scram/air-aualitv-disoersion-modeling-Dreferred-and-
recom m en ded-m odel s#aerm od
64

-------
8.5 ALOHA (CAMEO)
Computer-Aided Management of Emergency Overations/Areal Locations ofHazardous
Atmospheres (CAMEO/ALOHA)
Developer
U.S. EPA and the NOAA Office of Response and Restoration
Type of Model
Gaussian Plume Dispersion Model
Response Stage
Both Emergency Preparedness and Response
Original
Application
Local dispersion and threat zone estimation during accidental chemical releases

CAMEO/ALOHA is a simple hazard modeling package designed for emergency
responders. The software can help decision-makers rapidly plan and respond to
numerous types of chemical gas clouds, jets, fires, and dense gas releases within
a range of 100-10,000 meters of the release4. The software determines threat
zones which provide an estimate of downwind distance where proactive
measures should be taken. If NOAA or EPA is activated by IMAAC reach back
support, CAMEO/ALOHA may be used. The software package contains four
distinct entities: 1) CAMEO Chemicals, 2) CAMEO/tw, 3) ALOHA, and 4)
MARPLOT. CAMEO Chemicals is a comprehensive, proprietary database of
hazardous chemical datasheets and chemical physical properties that provides
information similar to that in the classic orange US DOT ERG. CAMEO
Chemicals rapidly displays descriptive properties of the chemical of interest.
CAMEO//?? is a database used to develop planning guidance about chemicals
within a local community such as details about a specific facility, chemical
transportation routes, and emergency response procedures. The plotting software
in CAMEO is MARPLOT.
Model Description
ALOHA is CAMEO's simple Gaussian plume dispersion model that simulates
the approximate spatial extent of a release hazard zone (Jones et al. 2013). It can
be used directly at the scene since results are generated within seconds from only
a few details about the chemical release and current meteorology. Although
simplified, ALOHA can account for variations in atmospheric stabilities based
on day- or nighttime releases, dispersion parameters that account for terrain, air
and chemical temperatures, and liquid evaporation rates (Jones et al. 2013).
Modules for fires, explosive releases, ruptures from pressurized tanks, and mists
or pools of evaporating chemicals have also been added to the most recent
version. ALOHA assesses the rate at which chemicals are released and
vaporized from their containment device to calculate the source strength. Non-
neutrally buoyant, dense gas releases have also been incorporated into the
model. These simplified algorithms are based on the DEGADIS model (Spicer
and Havens 1989). ALOHA was first developed by EPA and NOAA in the late
1980s specifically for the use by EPA's Environmental Response Team (ERT).
It may also be used to perform RMP guidance for chemical storage sites. The
model largely replaces the legacy Automated Resource for Chemical Hazard
Incident Evaluation (ARCHIE) model developed by the U.S. DOT.
4 For more information, see: https://www.epa.gov/eameo/wfaat~eameo-software-suite
65

-------
Pros
Simple, easy to use model for first responders; free and widely distributed;
comprehensive database of chemicals; fast model result
Cons
Lacks some simple additions like plume rise and certain custom user inputs;
susceptible to all limitations of Gaussian plume models; results best used for
informative guidelines
Runtime
Fast, seconds
Input Data
Requirements
Local atmospheric conditions, identity of the chemical, and details about the
spill scenario
Outputs
Threat zone estimates within a grid in ALOHA can be plotted on maps in
MARPLOT, GIS software, or Google Earth
Data assembly
requirements
during or after
emergency
response
General idea of local weather conditions (wind speed and direction), chemical
type released
Code language
C and some Python
Public or
Proprietary, Cost
Freelv available through EPA's website: https://www.epa.gov/cameo/aloha-
software
Ease of use
Easy, software used through the CAMEO GUI
Ease of obtaining
information and
availability of
technical support
Questions, comments, suggestions, and software issues can be addressed by
emailing the RMP Reporting Center: RMPRCfS.eDacdx.net NOAA's Office of
Response and Restoration: orr.cameo@noaa.gov, or bv calling the CAMEO help
desk at (703) 227-7650. Training can be found through:
https://response.restoration.noaa.gov/training-and-
education/training/workshops/cam eo-training.html
Source code
availability
Yes, but since the chemical database is a proprietary component, a user license
must be set up with the American Institute of Chemical Engineers at the cost of
$3,400 per vear through www.aiche.org/dippr. The source code itself is free, but
a license is still required.
Installation
requirements/
software
Most Windows PC or Mac operating systems, with capability as far back as
Windows 7 and iOS Mountain Lion (10.8); portable versions on smartphones are
also available
Maintenance Status
Regular updates to the chemical library, user interface, program functionality,
and help documentation. Most recent version as of mid-2020 is Version 5.4.7,
last updated in September 2016
Documentation
ALOHA Technical Documentation for v5.4.4 is found at:
https://response.restoration.noaa.gov/sites/default/files/ALOH h Doc.pdf
Link to Website
https://www.epa.gov/cameo/aloha-software and
https://response.restoration.noaa.gov/oil-and-chemical-spills/chemical-
spills/aloha
66

-------
8.6 CALPUFF
California Puff Model (CALPUFF)
Developer
Sigma Research Corporation (SRC), now Exponent, Inc.
Type of Model
Lagrangian, Gaussian Puff Dispersion Model
Response Stage
Emergency Preparedness
Original
Application
Moderate to long-range transport of gaseous substances through even and
complex terrain

CALPUFF is a non-steady state Lagrangian Puff model used to simulate
buoyant, instantaneous, or continuous-release, long-range transport of airborne
contaminants (i.e., PM, SOx, NOx, or inert particles) (Scire et al. 2000). As
opposed to a steady-state model, CALPUFF can simulate multiple emission and
removal processes at various rates by not necessarily maintaining equilibrium.
The model is listed as one of EPA's alternative dispersion models for assessing
long range transport of pollutants and its impacts on human health and the
environment. It has the capability of simulating time-varying point and area
sources, domains as small as few hundred meters to a large as hundreds of
kilometers, simulation times from one-hour to one-year, chemical conversion
and removal mechanisms, and special treatments for complex terrain (Scire et al.
2000). It consists of wet and dry deposition, building downwash, and fumigation
algorithms. The model can also account for low wind speeds, near-field impacts
from source to receptor, and regulatory air quality applications (such as
attainment areas, visibility, and criteria pollutants).
Model Description
CALPUFF includes three main modules that aid in pre- and post-processing.
CALMET is a 3D meteorological model to develop hourly wind and temperature
fields for the gridded domain. Specifications of the PBL and local topography
(including terrain blocking flows or bodies of water) are also included.
CALPUFF is the transport and dispersion model that advects puffs of effluent
released from emission sources. The model uses the meteorology generated from
CALMET to predict the downwind dispersion and puff behavior. Non-gridded,
simplified wind profile data may also be used if CALMET is not run. CALPUFF
then produces hourly concentration and deposition values at user-specified
receptor locations downwind of the release. CALPUFF tracks the puffs using a
Lagrangian frame of reference. The final component called CALPOST processes
the model output to summarize the results into average and maximum
concentrations at the receptors. Additional modules aid in quality control checks
and flexibility for reading in meteorological or terrain data. To enhance the
functionality, each component of the model can be run through an optional GUI
window to prepare, configure, and run the model. CALPUFF also interfaces
with other meteorological models such as MM5 and WRF to allow greater
support for localized meteorological processes.
Pros
Continuously updated, well-tested, and listed as an EPA alternative dispersion
model; permits long run times at distances as great as 200 km downwind of
source
67

-------
Cons
Limited emergency response use, mainly used to perform analyses that help
address regulatory and air quality issues; could have a large learning curve
Runtime
Depends on number of sources, receptors, and length of simulation; could be
seconds to hours
Input Data
Requirements
At a minimum, a wind and temperature profile; source type, emission rate, and
locations of receptors
Outputs
Average and maximum concentrations at the user-specified downwind receptors;
indication of atmospheric visibility and regulatory air quality attainment at each
receptor
Data assembly
requirements
during or after
emergency
response
Wind and temperature profiles, emission source specifics
Code language
FORTRAN
Public or
Proprietary, Cost
Freely available to anyone through Exponent, Inc.'s website; a more user-
friendly version is also available with a streamlined GUI and postprocessing
system by Lakes Environmental or Breeze Software: httos://www.breeze-
software.com/software/calDiifT although the price is $3,595. Private consultants
will also run the model for a cost.
Ease of use
Moderate, when used with a GUI window.
Ease of obtaining
information and
availability of
technical support
The GUI windows contain an extensive help system. Training can be obtained
from the Exponent developers. EPA provides some reference guides on SCRAM
website.
Source code
availability
Yes
Installation
requirements/
software
Windows PC
Maintenance Status
As of mid-2020, the standard, stable distribution version is CALPUFF v7.2.1.
CALPUFF v7.3.1 is also available as a beta release. V5.8.8 is EPA's approved
alternative regulatory version of the model.
Documentation
User's guide for CALPUFF v6 can be downloaded at:
http://www.src.corn/calpuff/download/CAL"i I I , ^rsiob 1 rlnstructions.od
£ with an addendum for v7 at:
htto://w ww.src.com/calouff/download/download.htm
Link to Website
http://yvww.src.com/
68

-------
8.7 C AS RA M
Chemical Accident Statistical Risk Assessment Model (CASRAM)
Developer
Argonne National Laboratory (ANL)
Type of Model
Statistical Analysis tool incorporating a Gaussian Plume Dispersion Model
Response Stage
Emergency Response
Original
Application
Straight line Gaussian plume model for chemical releases over simplified even
terrain
Model Description
CASRAM is a statistical analysis model that determines the distribution of
hypothetical outcomes of affected populations associated with hazardous
chemical releases of materials stored or transported through an area. Using
chemical shipment profiles, routes, and meteorology inputs, the model runs
tens of thousands of incidents for rail and highway chemical accidents (Brown
et al. 2017). The statistical plume results are then reported in the U.S. DOT's
ERG for protective action distances and routing-based risk assessments
(Brown et al. 2001). Most recently, CASRAM was run for technical guidance
in the 2016 ERG book (Brown et al. 2017) with a forthcoming report for the
2020 ERG. The model predicts hazard zone distributions to identify the
threshold chemical concentration where local populations could be affected. It
employs a Monte Carlo statistical analysis framework, which sets it apart from
other Gaussian models like ALOHA or SCIPUFF. CASRAM determines the
distribution of possible outcomes to provide a probability for each specific
release consequence. EPA and OSHA health exposure guidelines and
associated consequences are also estimated.
The model simulates both the physical and thermodynamic-related effects of a
hazardous chemical release by computing fixed or time-varying release rates
from tanks in liquefied, compressed, evaporated, or flashed chemical states. A
dense gas algorithm was added after the 2000 ERG using empirical
entrainment parameterizations from the DEGADIS model formulation (Brown
et al. 2017). Chemical reactivity, deposition, and various empirical surface
types and atmospheric stabilities are also incorporated within the model. A
weather and climate database for over 200 cities customizes the statistical
analyses based on region and state.
Pros
Theoretical atmospheric dispersion framework built upon existing and sound
principles; model results published and updated in each ERG version for
practical emergency responder use
Cons
Model is not generally available for use outside ANL but used to inform
resources used by responders
Runtime
Fast
Input Data
Requirements
Chemical release rate, type, and amount; general meteorological conditions
(wind and atmospheric stability)
Outputs
Statistical analysis of hazard zones following accidental container releases
69

-------
Data assembly
requirements
during or after
emergency
response
Chemical release rate, type, and amount; general meteorological conditions
Code language
Unknown
Public or
Proprietary, Cost
Proprietary; the model is not publicly distributed outside ANL
Ease of use
Unknown
Ease of obtaining
information and
availability of
technical support
The best point of contact is one of CASRAM's main developers, David F.
Brown at ANL: dbrown@anl.gov ("https://www.anl.gov/profile/david-f-
brown).
Source code
availability
No
Installation
requirements/
software
Unknown
Maintenance Status
Still used as of mid-2020. The code is updated and maintained every 2-3 years,
as per communication with David Brown.
Documentation
Information about the model can be found inside this 2017 technical
document, although there is no official publicly available manual:
https://www.phmsa.dot.gov/sites/phmsa.dot.gov/files/docs/training/hazmat/erg
IN6/201 'M ^-technical-document.pdf
Link to Website
See documentation above
70

-------
8.8 CT-Analyst
Contaminant Transport Analyst (CT-Analyst)
Developer
U.S. Naval Research Laboratory
Type of Model
LES CFD Model
Response Stage
Both Emergency Preparedness and Response
Original
Application
Flexible for all types of CBRN releases within complex urban areas
Model Description
CT-Analyst is a hybrid plume dispersion model that provides an instantaneous,
3D, CFD LES model depiction of CBRN releases within complex urban areas
to aid emergency responders in accidental or intentional airborne contaminant
transport threats. The model simulates plume dispersion and propagation
within the urban canopy at fine-scale resolution. Normally, LES simulations
require lengthy processing and computational times, but CT-Analyst can
produce dispersion results within seconds. Before a potential accidental release
scenario, velocity fields are pre-computed for numerous meteorological
conditions using NRL's high resolution LES transport model FAST3D-CT.
The simulated database structure (called "dispersion nomografs") is processed
into an efficient form used by CT-Analyst. It has also been shown to produce
more detailed dispersion information with better results than more common
Gaussian puff and plume models (Boris et al. 2003). The model was designed
after 9/11 as a fast-response dispersion resource that can run with limited
information about the source type.
CT-Analyst can incorporate inputs from fixed and mobile sensors or inform
the optimal locations for placing monitoring sites for model evaluation. The
model uses principles of fluid dynamics and turbulence to simulate urban
dispersion. Even though the steering wind direction and velocity magnitude
influences the direction of plume spread, the specific urban morphologies and
orientation of structures and streets control localized concentrations.
Specifically, the model has high enough resolution to simulate building vortex
shedding, recirculation zones, solar heating variations, and surface roughness
(Boris et al. 2003). The model aims to better predict hazardous dispersion to
avoid additional fatalities, exposures, and to plan the best course of evacuation.
Pros
Rapid results, which are ideal for emergency response use; has been evaluated
through field studies and published conference proceedings
Cons
Preprocessing velocity fields is a lengthy process and may be difficult for
responders; requires that FAST3D-CT be run for the specific case
Runtime
Simulation results are near-instantaneous and can be produced within seconds,
but computational fields must be prepared ahead, which can take hours
Input Data
Requirements
Measurements from isolated sensors (for model verification), general
meteorological conditions (wind speed and direction)
71

-------
Outputs
Dispersion plume that can be output to mapping services and rapidly
disseminated
Data assembly
requirements
during or after
emergency
response
Limited information is needed, including a general sense of the release
location, type, and local meteorology
Code language
Much of the source code and modules are written in Fortran
Public or
Proprietary, Cost
The model can be downloaded bv request at: httos://www.nrl.navv.mil/lco/ct-
an al v st/download
Ease of use
CT-Analyst has an easy-to-use interface that is simple to run once transport
fields are generated through FAST3D-CT
Ease of obtaining
information and
availability of
technical support
Questions or comments can be directed through the contact form at:
httos://www.nrl.navv.mil/lco/ct-analvst/contact
Source code
availability
No
Installation
requirements/
software
Windows PC
Maintenance Status
Model is still used and supported by NRL
Documentation
See: Boris J.P., G. Patnaik, T. Young, Jr., 2003: CT-Analyst: Verification and
Validation, NRL Report 4-1226-3377.
Link to Website
httos://www.nrl.navv.mil/lco/ct-analvst
72

-------
8.9 DEGADIS
Dense Gas Dispersion Model (DEGADIS)
Developer
University of Arkansas and the U.S. Environmental Protection Agency
Type of Model
Gaussian Plume Dispersion Model
Response Stage
Both Emergency Preparedness and Response
Original
Application
Dense chemical gas releases over even terrain
Model Description
DEGADIS is a dense gas dispersion model used to simulate the concentrations
of toxic chemical releases, especially for gases or aerosols heavier than the
ambient air. The model can simulate evaporating pools and upward-directing
or zero-momentum releases and jets, primarily over flat, level terrain. As one
of EPA's alternative models, it can also predict the dispersion processes
accompanying the gravity-driven flow and entrainment of the dense gas into
the atmospheric boundary layer (Spicer and Havens 1989). DEGADIS is
designed for zero-momentum, ground-level, area sources released from gas or
aerosol clouds. The model can predict the downwind dispersion as a stably
stratified plume or gas cloud. It has also been modified to simulate the vertical
plume or cross section using the Pasquill-Gifford parameters to represent
turbulent entrainment within the gas cloud. Although the model is primarily
designed for ground-level sources, it can simulate the plume centerline and
maximum concentration as a jet or plume lofts and then slumps back towards
the surface due to gravity. The model can simulate continuous, finite (a
constant rate over a short period of time), or transient (time-varying) release
durations.
DEGADIS, which is like the HGSYSTEM model in many ways, is freely
available, evaluated, and recommended as an alternative model by EPA.
DEGADIS has been tested and evaluated against some dense gas field and
laboratory studies, although robust opportunities for these tests and evaluations
are somewhat limited. Specifically, DEGADIS was evaluated using eight field
experiments in Hanna et al. (1993) with a more recent evaluation against
chlorine measurements from the JRII field study that is forthcoming.
Pros
Quick and accurate estimations of dense gas releases; model formulated on
peer-reviewed dispersion principles (such as PGT stability classes, boundary
layer similarity theories, and dense gas behavior); other models use DEGADIS
formulations for their core dense gas dispersion
Cons
The free version of the model is run on a command line; otherwise, a paid GUI
is available
Runtime
Fast
Input Data
Requirements
General meteorological and boundary layer conditions; specifics about the
release agent, duration, amount, and method
73

-------
Outputs
Prediction of the downwind concentrations at various heights
Data assembly
requirements
during or after
emergency
response
General meteorological conditions; specifics about the release agent, duration,
amount, and method
Code language
The source code is written in Fortran 77
Public or
Proprietary, Cost
Available for free download through EPA's SCRAM website, or through the
Breeze Software platform: httos://www.breeze-software.com/Software/LFG-
Fire-Risk/Product-Tour/DEGADIS-Model/. However, the Breeze GUI is not
free
Ease of use
Moderate; runs from a Windows command line prompt. Versions using the
GUI window make operation more straightforward
Ease of obtaining
information and
availability of
technical support
Support for the EPA SCRAM website can be obtained by contacting George
Bridgers: bridgers.george®,eoa.gov. Specific model support or questions can
be directed to one of the develooers. Dr. Tom Soicer: tosfS.uark.edu
Source code
availability
Yes, on EPA's SCRAM website
Installation
requirements/
software
Windows PC
Maintenance Status
Minor changes that do not change the model computations were introduced
into DEGADIS v2.1 in September 2012
Documentation
A user's guide is available online at:
httos://www3.eoa.gov/ttn/scram/userg/other/degadis2.odf
Link to Website
httDs://www.eoa.gov/scram/air-aualitv-disDersion-modeling-alternative-
models#degadis
74

-------
8.10 HotSpot
HotSvot
Developer
National Atmospheric Release Advisory Center (NARAC), Lawrence
Livermore National Laboratory (LLNL)
Type of Model
Gaussian Plume Dispersion Model
Response Stage
Both Emergency Preparedness and Response
Original
Application
Radiological releases in simple terrain regions
Model Description
HotSpot is a simplified Gaussian Plume model that provides emergency
planners and responders a fast, field-portable set of software tools for
evaluating radioactive release incidents. The model is designed for near-
surface releases under short dispersion ranges and durations (less than 10 km
and 24 hours). The model produces the best results under open terrain and
simple meteorological conditions. Due to these limiting factors, HotSpot
provides a fast but somewhat conservative means of approximating the effects
of an accidental or intentional radioactive release. It can estimate continuous or
instantaneous releases from explosions, fuel fires, and wide-area
contamination events. The core dispersion model is built upon the general
Gaussian Plume equation and accounts for various atmospheric stabilities,
surface types and roughness, deposition, and plume rise (Homann and Aluzzi
2014). An additional tool estimates the effect of nuclear weapons, including
neutron and gamma, blast, and thermal effects. The software also computes a
first-order approximation of radiation and inhalation dose effects associated
with explosions and facilities that handle nuclear materials. The model
contains an extensive source term database and can simulate the dispersion of
plutonium, uranium, tritium, and other radionuclides through plume,
explosion, fire, and resuspension modeling methods.
First released in 1985, HotSpot has added plotting and contour plotting
capabilities, and results can be exported to Google Earth or other GIS plotting
software. The fast, yet conservative estimation of the radioactive release is
designed so emergency responders can get a general sense of the episode (for
example, ionizing radiation from the deposition of particles is ignored (Hill
2003)). Effective doses are estimated for the immediate and acute radiological
impact on internal organs. The code also can estimate the potential fallout and
arrival time, dose rate, and propagation of the fallout radioactivity after the
release and as far as several weeks post-event.
Pros
Simple and fast, reasonable dose, exposure, and dispersion predication to
inform emergency responders
Cons
Not for use during incidents with complex terrain or variable weather
conditions; may underestimate some effects and provide a conservative
prediction; susceptible to all limitations of Gaussian dispersion models
75

-------
Runtime
Fast, 15-30 seconds or less
Input Data
Requirements
Isotope release type, mass, and general meteorological conditions
Outputs
Hazard zones and dose estimates from release plume
Data assembly
requirements
during or after
emergency
response
Information about the release type and amount, general weather conditions
Code language
Visual Basic, Microsoft .NET Framework
Public or
Proprietary, Cost
The latest version can be freely downloaded by filling out by registering as a
HotSpot user at: https://naracweb.llnl.8ov/web/hotspot/re8isterUser.html
without having to have a NARAC account.
Ease of use
Very easy for most users, simplified GUI
Ease of obtaining
information and
availability of
technical support
While a public help forum does not exist, questions or problems can be
directed to hotspot@lM.80v
Source code
availability
No
Installation
requirements/
software
Windows PC
Maintenance Status
Currently operated and updated by LLNL to incorporate the most current
radiological dose conversion methodologies; Current Version 3.1.2 as of mid-
2020
Documentation
The user's manual can be downloaded at:
https://narac.llnl.80v/content/assets/docs/HotSpot-UserGuide- ?
Link to Website
https://narac.llnl.80v/hotspot
76

-------
8.11 HPAC
Hazard Prediction Assessment Capability (HPAC) Model
Developer
U.S. Department of Defense (DOD), Defense Threat Reduction Agency
(DTRA), and Applied Research Associates, Inc. (ARA)
Type of Model
Gaussian Puff Dispersion Model
Response Stage
Both Emergency Preparedness and Response
Original
Application
Various CBRN releases in complex terrain and urban areas

The HPAC model is a comprehensive, operational, and research-grade CBRN
dispersion modeling system that integrates high resolution meteorological data
(DTRA 2004). It can be used for hazardous release-agent planning purposes
(i.e., "forward deployable") or through reach back service for civilian and
military populations. The model can be applied to a wide variety of defense,
industrial, or transportation-related accidents. HPAC is the primary model
used by DTRA for IMAAC emergency response plumes and can typically be
delivered to customers within 20-30 minutes after the initial request. The
model can be activated quickly because DTRA automatically pulls in real-time
NWS weather data and archives it on their meteorological data servers. These
databases also store worldwide NWP products and climate reanalysis data.
Historical weather for numerous locations can also be accessed. The model has
been in use since 1995 and is managed by DTRA out of Ft. Belvoir, VA.
HPAC is used extensively with the DOD and has been evaluated for several
urban field experiments (Chang et al. 2005), and most recently by Miner et al.
(2019).
Model Description
HPAC's primary transport and dispersion model is built up on the SCIPUFF
Gaussian puff model (Sykes et al. 2007) that has been extensively tested and
developed since the 1980s. SCIPUFF, which has also been incorporated within
many other dispersion models, permits fast computational times (within
minutes) and many advanced capabilities, including atmospheric transport and
dispersion plume estimations, urban parameterizations, deposition, dose, and
human effects-hazards. The source term can be identified by a particle size
distribution and can incorporate continuous, instantaneous, and finite duration
releases. NCAR's Hazardous Material Source Term Estimation tool5 is also
being used and developed within HPAC to streamline the input process.
SCIPUFF uses the detailed NWS meteorology to simulate time and space-
varying puffs from the effluent source that travel downwind and disperse,
resulting in an accurate representation of the atmosphere at the time and
location of the release, including splitting puffs when they grow too large due
to wind shear and turbulence (Miner et al. 2019). Recent additions to
SCIPUFF simulate the effects of potential radioactive releases from nuclear
weapons or power plant reactor accidents and modifications for dense gas and
simple chemistry and aerosols. HPAC/SCIPUFF also uses urban canopy
modifications to account for changes in the wind speed profile (Cionco 1978)
5Visit this link for more information: littps://nar.ucar.edn/20.1.8/nit/tiazardous-materia 1-sonrce-tenn-estimation
77

-------

as well as urban parameterizations from DSTL's UDM (Hall et al. 2002).
While UDM does not resolve dispersion around individual buildings, it
modifies a plume based on street alignment and building density in urban
areas. SCIPUFF can also account for variations in the terrain and land
surfaces, which tends to have a large influence on the plume transport. Digital
terrain elevation files are used to develop mass consistent wind and turbulence
within the model through natural obstacles. Many additional capabilities are
also built into HP AC, all of which is run through a GUI window.
Pros
Fast access to real-time weather data through meteorological data servers;
extensively evaluated with field data and shown to have good performance;
used operationally by many government entities
Cons
May be complicated to use without knowledge of the software; not a large
online support base (but HP AC instructional classes exist)
Runtime
Moderately fast (within 15-30 minutes or less)
Input Data
Requirements
Time and location of the release, information about the source term
Outputs
Dispersion plume with estimated hazard zones downwind of the source
Data assembly
requirements
during or after
emergency
response
Time and location of the release, information about the source term; HP AC is
also used when resources are requested through IMAAC.
Code language
The core SCIPUFF code is written in FORTRAN 90 but operation of HP AC is
streamlined through a GUI window
Public or
Proprietary, Cost
Available for free to US Government employees and contractors, other
government-related uses, and to academia by emailing the software
distribution officer: Bonnie.a.cassano.ctr@mail.mil or the first email address
under the technical support box below. An application is required and will be
submitted to DTRA for approval.
Ease of use
Moderate, due to input options
Ease of obtaining
information and
availability of
technical support
User support and assistance can be obtained by emailing:
dtra.belvoir.rd.mbx.Reachback-Software-Distribiition@mail.mil. Help
regarding the meteorological data server and archived weather can be directed
to: dtra.belvoir.rd.list.meteorologicakdata-services@mail.mil
Source code
availability
No
Installation
requirements/
software
Windows PC with at least 20-25 GB of free hard drive space if the entire
archived meteorological data is desired
78

-------
Maintenance Status
Continuously updated and improved by DTRA and its contractor ARA. A recent
stable release was HP AC v6.5 (mid-2018)
Documentation
The HP AC v4.04 user's guide available in PDF at:
ftp://fto.atdd.noaa.gov/pub/gunter/hpac 404 users manual.odf for online
viewing; the newest HP AC model releases include the documentation within
the root directory on the CD shipped from DTRA
Link to Website
HP AC is currently (as of mid-2020) not posted on DTRA's Research and
development website: https://www.dtra.mil/Mission/Mission-
Directorates/Research-and-Development/ but information about the model can
be obtained through
httos://www.aca.osd.mil/ncbdD/nm/narD/Radiation Data/Soecialized Radiolo

gical.htm and the following papers: Miner et al. (2019), Chang et al. (2005),
and several others.
79

-------
8.12 HYSPLIT
Hybrid Single Particle Lasransian Integrated Trajectory Model (HYSPLIT)
Developer
NOAA Air Resources Laboratory (ARL)
Type of Model
Lagrangian Stochastic Particle Dispersion Model
Response Stage
Both Emergency Preparedness and Response
Original
Application
Local-to-regional forward and backward trajectory of particles or air parcels
Model Description
HYSPLIT is NOAA's Lagrangian dispersion model that calculates simple
forward and backward air parcel trajectories, contaminant transport, chemical
transformation, and deposition of particles, gases, or aerosols over regional
(mesoscale) or long-ranges (synoptic; >1000 miles) (Stein et al. 2015). Due to
its high fidelity of operation, it is one of the most used transport and
dispersion models in the world. The model can generate trajectories using
archived, gridded meteorological model data for past episodes or gridded
model simulations for future predictions (Draxler et al. 2020). HYSPLIT can
be run rather quickly through an internet browser on NOAA's ARL READY
website for archived episodes, or through reanalysis data from as far back as
1949. Additionally, it can be downloaded and run locally on a Windows PC,
Mac, or LINUX workstation through a GUI or script. The latter is used
mainly for research purposes and can be driven with weather forecast data
generated from models like WRF. There are at least 15 options for gridded
3D meteorology inputs using global or North American datasets. The gridded
wind fields on the READY website contain horizontal resolutions ranging
from 3 km to 1° and various vertical resolutions using pressure- and
elevation-related coordinate systems. Many users run a backward trajectory
analysis at a receptor site to determine the origin of an air mass or
contaminant source. The model can calculate the dispersion of an unknown
material point source (instantaneous or long duration), where it calculates the
forward trajectory of generic particles. It can also simulate prescribed burns,
wildfire smoke, and volcanic eruptions.
HYSPLIT calculates the advection and diffusion using a Lagrangian moving
frame of reference as the trajectory of particles or parcels move from their
original location. The model simulates transport interactions at and between
multiple levels above the earth's surface. Pollutant dispersion is calculated
through a series of puffs as they advect downwind. Many options such as wet
and dry deposition, radioactive decay, resuspension, the addition of more than
one contaminant source and rates, and various turbulence parameters have
been incorporated. As such, HYSPLIT has been used to assist with
emergency dispersion analyses if NOAA is called to provide reach back
service as part of IMAAC. HYSPLIT has been under continuous development
since the late 1980s and continues to undergo routine improvements. It has
replaced NOAA's Volcanic Ash Forecast Transport and Dispersion Model
(VAFTAD) and the TRIAD model from 1970s-80s.
Pros
Fast, free, and well-documented with large support base; model has been
extensively evaluated and used in the atmospheric sciences field and is used
80

-------

operationally within NOAA; provides an accurate representation of plume
due to time-varying and high-resolution meteorology
Cons
May not provide as much detail on local dispersion-related effects as some
emergency responders require; needs complex gridded wind input datasets
Runtime
Fast (seconds through the READY website platform)
Input Data
Requirements
3D gridded meteorological model data, basic information about release source
Outputs
Dispersion plume or (forward or backward) spatial trajectory of source or
receptor, which can be plotted or output to Google Earth
Data assembly
requirements
during or after
emergency response
Meteorological model data from the most recent NOAA model runs (i.e.,
NAM, GFS, or HRRR) can be gathered through the READY website; basic
information about release source
Code language
Most of the source code is written in Fortran
Public or
Proprietary, Cost
Freely available through NOAA ARL's website
Ease of use
Web-based use is very simple, and results can be generated within minutes
with limited prior use. Other versions (i.e., through the LINUX command line
and using forecast model data) may be more complicated for some users
Ease of obtaining
information and
availability of
technical support
A large support forum to communicate questions, improvements, problems,
and ideas is available through: httos://hvsolitbbs.arl.noaa.gov/. Various
technical tutorials are also available for self-paced training:
httos:/Av ww.readv.noaa.gov/HYSPL ©rials.oho
Source code
availability
Source code repository is available for non-commercial use only to a limited
number of registered users. Interested parties can send a request via email to
arl.webmasterfS.noaa.gov, but granting the request is subiect to the discretion
of HYSPLIT developers. Modifications or improvements to the source code
are expected to be shared with the HYPLIT user community
Installation
requirements/
software
Model can be run through a browser, or locally on a 64-Bit Windows PC,
Mac, or through the command line on LINUX systems. Users do not need to
register to use the web-based trajectory or dispersion software using archived
meteorological data. Registration is only required to use forecast data or to
download the LINUX or registered versions for PC or Mac computers.
Registration is permitted for government, commercial, educational/academia,
or non-profit users.
Maintenance Status
Continuously updated and improved. The most recent version as of mid-2020
is HYSPLIT v5.0.0 released in April 2020. Status updates are posted on:
httos://w ww.arl.noaa.gov/hvsolit/hvsolit-model-uodates/
Documentation
A complete web-based user's guide is available at:
httos://www.readv.noaa.gov/hvsolitusersguide/ or through NOAA Technical
Memo ERL-ARL 230 (Draxler 1999) along with other self-paced resources
Link to Website
httos://www.readv.noaa.eov/HYSPLIT.Dho
81

-------
8.13 JEM
Joint Effects Model (JEM)
Developer
Department of Defense (DOD); Aeris, LLC.
Type of Model
Gaussian Puff Dispersion Model
Response Stage
Both Emergency Preparedness and Response
Original
Application
For most CBRNe releases in remote or urban areas
Model Description
JEM is a comprehensive and operational dispersion modeling software application
used to simulate accidental or intentional CBRNe incidents and weapon strikes. The
model is widely used within the U.S. military with advanced capacities for complex
terrain, TICs, human health indications, and urban environments. The model is
browser-based and runs through an internet application to allow for portable and
near real-time simulations following various types of releases or strikes, although it
can also be run on a stand-alone system. JEM is primarily supported, maintained,
and used by the US Army and DOD. It is currently the only accredited tool to
effectively model impacts from hazardous releases by assisting warfighters in
planning for and mitigating the effects of WMD. The model simulates the impact of
downwind dispersion based on various weather conditions (wind speed, direction,
and atmospheric stability), terrain, local structures, and release material
interactions. The DOD can generate JEM results on a 24/7 basis through its internet
and telephone reach back service. JEM can also be implemented for strategic or
tactical use within the U.S. or overseas.
JEM's core dispersion model is built upon the SCIPUFF Gaussian Puff Model
(Sykes et al. 2007) to simulate time and space-varying puffs from the effluent
source that travel downwind and disperse. This is the same model that drives the
dispersion component in HP AC. For more information about SCIPUFF, see the
entry for HP AC in Section 8.11. JEM contains two options for urban dispersion: an
urban canopy parameterization based on wind and turbulence profiles from
SCIPUFF, and the UDM model from DSTL (Hall et al. 2004). UDM was also
provided in the quick reference table. A comprehensive evaluation of JEM from
four different urban field studies has been reported in Chang and Hanna (2010) and
Hanna and Chang (2012). JEM versions 1 and 2 are currently in use, although JEM
1 is being phased out to support more modern computer technologies. JEM replaces
and/or incorporates the DOD's VSLTRACK and D2Puff dispersion models for
chemical releases. As of FY20, work is underway to better align JEM 2 and HP AC
6.5 for time and cost considerations since the model framework, individual
components, and user interfaces are somewhat similar.
Pros
The only DOD-accredited tool to simulate CBRNe dispersion for warfighting and
tactical purposes; uses well-documented and evaluated SCIPUFF model framework
Cons
Mainly for military use, although some external research use is possible
Runtime
Fast
82

-------
Input Data
Requirements
Wind direction and speed, release specifics
Outputs
Spatial estimation of plume from release
Data assembly
requirements
during or after
emergency
response
Meteorology (wind speed, wind direction, and a general indication of weather
conditions and atmospheric stability) near the release location, method and type of
release and rate of emission
Code language
The core of SCIPUFF is written in FORTRAN 90 but operation of JEM is
streamlined through a GUI for most uses
Public or
Proprietary, Cost
Mainly for use by DOD, contractors, and some foreign militaries (e.g., Spain and
Canada)
Ease of use
Web-based GUI simplifies operation
Ease of obtaining
information and
availability of
technical support
Not known
Source code
availability
No
Installation
requirements/
software
Windows PC-based; also deployed on UNIX systems and is integrated into
Command and Control C2 systems across the DOD. The model is available in a
stand-alone version or through a networked or web platform
Maintenance Status
The latest version is JEM 2 as of mid-2020 with planned continual development,
integration, and deployment, and additional cloud-based capabilities to at least
FY23. JEM 2 is still being evaluated and improved with field study data
Documentation
A comprehensive technical document for the JEM model does not exist (Hanna and
Chang 2012). The transport and dispersion model specifics can be found in the
SCIPUFF documentation bv starting at: http://www.sciputt.or8/
Link to Website
https://asc.annv.mil/web/portfolio-item/ioint-effects-model-iem/
83

-------
8.14 MELCOR and MACCS
MELCOR Accident Consequence Code System (MACCS)
Developer
Sandia National Laboratory (SNL)
Type of Model
Gaussian Plume Dispersion Model
Response Stage
Both Emergency Preparedness and Response
Original
Application
Around and adjacent to nuclear power plant sites; radiological/nuclear releases
Model Description
MACCS is a comprehensive straight-line Gaussian plume model package used
to develop a probabilistic risk assessment from a severe accidental atmospheric
release of radioactive material from light water nuclear power plants (Chanin
et al. 1998). Mainly developed and used for and by the NRC since 1990,
MACCS simulates ecosystem and human dose and exposure impacts within
and adjacent to nuclear power plants. It can diagnose potential land
contamination levels, exposure and risk to susceptible populations based on
recommended response actions and economic losses resulting from an
accident. MACCS incorporates wind and atmospheric turbulence from time-
varying meteorology, plume rise, wet and dry deposition, inhalation, cloud and
ground shine, ingestion, and shielding. The suite of codes contains
MelMACCS (the preprocessor code that interfaces MELCOR, the process
analysis code simulating the chain of events during a meltdown, with
MACCS), WinMACCS (a graphical user interface), SecPop (a program to
generate consequence calculations based on population, land use, and
economic databases), and COMIDA2 (a food pathway model to estimate doses
of radionuclides from consumption). MACCS provides a comparative
assessment of various dose-threshold models to more objectively quantify the
uncertainty of various inputs. The model also incorporates a road network
model to suggest the best evacuation routes to limit exposure from the
radioactive release and corresponding plume.
MACCS is used to inform emergency preparation and response guidance
around reactor sites. The NRC requires all nuclear power plants applying for
or renewing operating licenses to perform cost-benefit analyses using
MACCS. The software suite is currently the only code used by the NRC to
inform Level 3 critical nuclear episode risk assessments post-release.
Additionally, the DOE uses MACCS to assess safety by demonstrating
emissions at powerplant boundaries remain below regulatory limits. The NRC
may also use MACCS for modeling and risk analysis support for IMAAC.
Recent uncertainty analyses called the State-of-the-Art Reactor Consequence
Analyses (SOARCA) project have documented best modeling practices from
numerous studies and current knowledge on severe nuclear accidents (Chang
et al. 2012). SNL also provided severe accident modeling support during the
Fukushima Power Plant disaster.
Pros
No other U.S. publicly available dispersion modeling and consequence
analysis code currently offers all MACCS's capabilities; currently supported
by the NRC for risk analyses, planning, and power plant licensing;
continuously enhanced and improved
84

-------
Cons
Mainly used by staff at DOE facilities; model is susceptible to all inherent
limitations and simplifying principles of Gaussian plume models; MELCOR is
complex to use if the user is not familiar with power plant controls
Runtime
Model runs rapidly (within seconds to minutes) once configured
Input Data
Requirements
Local meteorology, nuclear powerplant information. WinMACCS contains
databases of local populations, economic situations, and land use.
Outputs
Series of risk and consequence analyses
Data assembly
requirements
during or after
emergency
response
Local, time-varying meteorology (wind speed, wind direction, and a general
indication of weather conditions), method of accident and potential rate of
emission
Code language
FORTRAN
Public or
Proprietary, Cost
Public distribution to domestic utilities, vendors, academic institutions,
commercial enterprises, and some international organizations by filling out a
non-disclosure agreement. The software is free to academic institutions, NRC
contractors, U.S. federal government, and some international government
organizations although no technical assistance is provided. There is a $2,500
one-time fee for shipping, handling, and installation service for commercial
organizations only.
Ease of use
Running, setting up, and producing the output from MELCOR is potentially
time-consuming. However, a program called MelMACCS has been developed
to streamline the source integration and the dispersion component to the
consequence analysis software, which acts as a preprocessor interface.
Ease of obtaining
information and
availability of
technical support
No technical assistance is provided, but certain program assistance and
questions can be provided through wg-maccs-entity@sandia.gov or for a fee.
Source code
availability
No
Installation
requirements/
software
The software runs through an application call WinMACCS 4.0, a graphical
user interface that streamlines model setup and results
Maintenance Status
The latest version of MACCS is 4.0 as of mid-2020. The software is currently
implemented across DOE platforms with continuous maintenance for current
and future reactor designs. Code modernization is underway. The MACCS
Development Team is working to couple the model with HYSPLIT as an
alternate and improved dispersion model. Additional features to evaluate
economic impacts on gross domestic product from power plant accidents is
also underway.
85

-------
Documentation
An extensive description of the MACCS dispersion model is available at:
https://maccs.sandia.gov/docs/MACCS factsheets/ 'S%20Model%20De
scription.pdf.
MACCS User's guide:
https://maccs.sandia.80v/docs/MACCS factsheets/Code%20Manual%20for%
20M ACC S2%20Vol%201 .pdf
MELCOR also has its own set of code manuals:
https://www.nrc.gOv/docs/ML1704/T 40A429.pdf

Link to Website
https://maccs.sandia.gov/maccs.aspx
86

-------
8.15 QUIC
Quick Urban and Industrial Complex (QUIC) Model
Developer
Los Alamos National Laboratory (LANL), DOE (PI: Michael Brown)
Type of Model
Lagrangian particle, random walk urban dispersion model
Response Stage
Emergency Preparedness
Original
Application
Various CBRNe releases within urban areas
Model Description
QUIC is a relatively fast Lagrangian dispersion model that can compute
pollutant dispersal on the building-to-neighborhood scale (Nelson and Brown
2013). It is "CFD-like" in the sense that it simulates detailed wind flow and
pressure fields around obstacles but runs relatively quickly on a laptop
depending on the domain size and release specifications. The model contains
algorithms that calculate flow fields around building profiles and through
street canyons (Brown 2014) based on the work of Rockle (1990), with
improvements from Nelson et al. (2008, 2009) and others. The addition of
buildings could produce more realistic results around obstacles and through
street intersections since certain neighborhoods could receive higher pollutant
concentrations while others remain relatively unaffected due to building wake
and cavity effects. When setting up the QUIC simulation through its GUI, a
shapefile can be imported with building dimensions, or the user can develop
their own domain using CityBuilder. The model can be run using an inner and
outer domain to expedite the simulation and to develop the appropriate
turbulence fields as the wind encounters the inner focus area. QUIC includes a
3D wind field model called QUIC-URB that generates the flow conditions
around the urban obstacles. The local meteorology (wind speed and direction)
is added via the MetGenerator tool. A module calculates the vertical wind
profile based on theoretical boundary layer scaling equations, or the user can
import their own profile. Various wind profiles can also be implemented
within the domain as a function of time.
The placement and specification of the release parameters are defined in the
transport and dispersion model called QUIC-PLUME. This is a Lagrangian
random walk dispersion model that calculates concentration and deposition
fields from the flow generated in QUIC-URB. QUIC can account for a variety
of point, area, and line CBRNe releases with more advanced properties,
including dense gas, evaporation, and buoyant dispersion effects. Custom
properties related to the release type such as specifications of a toxic gas or
particle size distribution, amount, location, and more specific thermodynamic
details can be defined. QUIC can also track individual inert particles
downwind from the source. Experimental building infiltration, exposure, and
re-aerosolization algorithms have also been implemented, but these options
require more testing. The resulting plume can be plotted within the model's
GUI or exported to other plotting software and GIS maps for additional
analysis.
87

-------
Pros
Relatively fast running and accounts for building and street canyon effects in a
realistic way; evaluated against field and laboratory data (Brown et al. 2013),
many flexible input options
Cons
The learning curve can be high depending on application; lack of model
support, lack of output formats for postprocessing
Runtime
Seconds to < 1 hour
Input Data
Requirements
Buildings must be in shapefile format or constructed within QUIC-GUI,
accurate details about source terms and meteorology
Outputs
2D and 3D spatial plots of contaminant deposition or concentration
Data assembly
requirements
during or after
emergency
response
Moderate-high due to complex source term classifications
Code language
FORTRAN in executables, run through MATLAB
Public or
Proprietary, Cost
Free for researchers, government employees, contractors, and academia
Ease of use
Moderate; requires user to be familiar with user's manual, but run through
GUI window
Ease of obtaining
information and
availability of
technical support
No formal user support group but users can email developer for questions or
assistance: mbrownf2Hanl.gov
Source code
availability
No, unless working with developer to improve model
Installation
requirements/
software
None; works within MATLAB, but an executable is provided to run the model
as a standalone version. Runs on 32- and 64-bit Windows and Mac computers.
Maintenance Status
Continuously improved, currently v6.26 in 2018
Documentation
Well documented user's manual:
httos://www.lanl.gov/oroiects/ciuic/oDen files/OUI artGuide.odf
Several peer reviewed publications and conference proceedings.
Link to Website
https://www.lanl. gov/proi ects/quic/
88

-------
8.16 SHARC/ERAD
Specialized Hazard Assessment Response Cavabilitv/Exvlosive Release Atmospheric
Dispersion (SHARC/ERAD)
Developer
Sandia National Laboratory (SNL)
Type of Model
Gaussian Puff Dispersion Model
Response Stage
Both Emergency Preparedness and Response
Original
Application
Radiological dispersal devices (RDDs), radiological/nuclear and explosive
release types over flat terrain
Model Description
SHARC/ERAD is a suite of five models that simulates the release of
radioactivity from nuclear weapon explosions, RDDs, or other accidental or
intentional detonations. The software assesses the time-dependent dynamic
explosive buoyant plume rise and then estimates the associated human
exposure and evacuation decisions through an integrated geographic
information system and population databases. The SHARC model package
contains Nuke 2.0 to predict the immediate nuclear and radiation effects,
AIRRAD 2.0 for nuclear fallout estimations, Blast 2.0 to provide effects from
the explosion, ERAD 7.0 for the dispersal of radioactivity by explosively
driven plume rise or through conventional non-buoyant releases, and MCK,
which is the Monte Carlo Gaussian puff model for plume dispersion6. The 3D
puff dispersion model was first developed in the 1980's to assess the time-
dependent buoyant rise and atmospheric transport for different meteorological
conditions. It incorporates various surface roughness lengths for even terrain
types and models vertical diffusion using a Monte Carlo method, a random and
probabilistic approach adapted for turbulent dispersion.
SHARC can simulate multiple scenarios based on the time and location of the
event. It can predict nuclear fallout patterns and provide guidance for short-
term relocations and long-term evacuations. Fatalities, casualties, and exposure
estimates are determined though U.S. Census and Landscan data with the
ability to export graphical products to Google Earth or other GIS platforms.
The software produces automated graphical displays of areas affected by the
radiation and can develop integrated reports on the incident for responders.
With advance notice of a nuclear or RDD threat, SHARC contains a decision
support tool to determine the best way to move the device out of the area. In
addition, SHARC is incorporated within the Turbo FRMAC (Federal
Radiological Monitoring and Assessment Center) software to calculate official
federal response guidance and inform the proper actions needed between
federal, regional, and local emergency responders and planners.
Pros
Moderate input data requirements, vertical variation in meteorology, fast
computational time (which allows for multiple scenarios to be run), population
databases for a wide variety of locations
6 For more information, see: https://www.osti.gov/servlets/purl/1124469
89

-------
Cons
Able to simulate only under flat and open terrain, software restricted to
explosive and RDD releases
Runtime
Fast, approximately 2-5 minutes execution time
Input Data
Requirements
1-D vertical wind and temperature profiles for a single time period
Outputs
Fallout pattern, casualties, sheltering, evacuation guides, containment and
mitigation effects
Data assembly
requirements
during or after
emergency
response
Local ID meteorology, source term and knowledge of the release
Code language
Not known
Public or
Proprietary, Cost
Available to the international emergency response community upon request
through SNL's Nuclear Incident Response Program at: httos://mitt».sandia.gov
Ease of use
GUI simplifies operation
Ease of obtaining
information and
availability of
technical support
Ouestions can be directed to: nirp-support®,sandia.gov or nirp-
fogbugzfS.sandia.gov regarding specific inquiries about the software
Source code
availability
No
Installation
requirements/
software
Windows PC or Workstation; runs through GUI and combined with SNL's
Turbo FRMAC
Maintenance Status
Continuously maintained; current version is SHARC 2019 ERAD v7.0 as of
mid-2020
Documentation
User's manual available once software and account are requested and created
on SNL's website
Link to Website
https://nirD.sandia.gov/Software/SHARC/
90

-------
9.0 Concluding Remarks
Atmospheric dispersion models have evolved substantially since their introduction into the
emergency response community. The foundations of atmospheric transport and dispersion theory were
developed almost 100 years ago, but simple dispersion models that calculate downwind plume
concentrations were not introduced until the 1960s. These models, developed extensively over the
following decades, were used mainly to assess pollution levels from point emission sources, such as single
effluent stacks and industrial sites. The terrorism on US soil in 2001, and subsequent fear post-9/11
encouraged the growth and improvement of higher fidelity dispersion models, especially for urban areas
with greater threats for human exposure. Other homeland security threats, such as powerplant accidents,
biological releases, or chemical spills also demonstrate the critical need for dispersion modeling to be
continuously tested, developed, and improved. Dispersion modeling offers a crucial insight for emergency
preparation or planning scenarios so responders can be well-equipped and make knowledgeable decisions.
Dispersion modeling has also proved to be a critical component during the emergency response (Leitl et
al. 2016) and post-response stages to inform evacuation of affected communities, sample for
contamination, decontaminate surfaces, and manage waste generated from the recovery process.
The goal of this report is to briefly explain the fundamental concepts of atmospheric transport and
dispersion and provide a comprehensive database of dispersion models that can be used for emergency
preparation and response to facilitate discussion between public, private, academic, and/or government
sectors that use them. The abundance of model options often creates confusion and results in challenging
decisions on the type of model to use for a specific scenario. A comprehensive model review of this
magnitude has not recently occurred. This report also provides background information and a literature
review on previous model review efforts. Much of those databases laid the foundation for this work, with
modifications and additions for the current state of dispersion modeling. The report also provides
introductory concepts on boundary layer meteorology and the various types of dispersion models. A basic
understanding of the physical processes governing how dispersion models work is crucial when
interpreting the results. This work is intended to provide a quick reference for those new to dispersion
modeling or for those seeking to expand their knowledge base. It is not meant to replace primary literature
sources such as textbooks.
This report outlines and alphabetically sorts dispersion models with potential applications for
CBRNe risks. An extensive quick reference table for 96 different dispersion models is provided in Section
7.0. Sixteen of those models were selected for a more detailed, two-page review in Section 8.0 due to their
potential applicability and usefulness for emergency response. Twenty-four models were also identified
that could be potentially useful, but additional research is needed by the user to decide if that model is a
viable fit. The model review was not meant to recommend or endorse a specific model but to provide users
with a resource of options that document the currently available models so they can make their own
informed decisions. Additionally, this resource is not meant to take the place of other Federal dispersion
modeling options and reach back services such as IMAAC and NARAC.
Similar to the results presented in Mikelonis et al. (2018) and EPA (2018), no one single model is
found to have all the requirements conceived as beneficial during consequence management of a wide
area response. Out of the 16 dispersion models selected for detailed analysis, six were Gaussian Plume
models, four were Gaussian Puff models, four were Lagrangian particle models, and two were CFD or
LES model. A few models incorporated both Gaussian Puff and Plume relationships. The model review
results in the following observations:
1) Most dispersion models are developed and maintained by Federal government agencies or
National Laboratories
91

-------
2)	Most CBRNe models are not widely distributed to the public and require certain criteria to
obtain the model, such as being a government employee, contractor, or within academia
3)	Most emergency response models are Gaussian plume or puff models that run quickly, while
emergency preparation models are Lagrangian or CDF-based
4)	Model runtime is usually related to the model's complexity
5)	Finding model support may be difficult for many models, and some are not widely described
or documented online
6)	Model complexity depends mainly on the user's expertise, available hardware, the overarching
model framework, and the input data requirements
7)	Input data, especially to the detail most models require, may not be available immediately post-
release. This is a crucial time when the model is needed for emergency response guidance.
8)	A quantification of model uncertainty is usually not depicted directly by the model, but may
be accomplished by rerunning the model with different input options or tuning the parameters
9)	Most dispersion models are built upon the same underlying framework and mathematical
equations based on fundamental dispersion and turbulence theories
10)	There is extensive overlap in the capabilities of dispersion models, with additional elements
built for specific purposes related to the agency's mission
1 l)Many models are site-specific or research grade
12)	Source codes are generally not released by the developers
13)	More user-friendly and mobile model options are needed, particularly with intuitive user
interfaces though laptops, smartphone apps, or through remote cloud-computing.
Based on the review presented here, an acceptable balance of speed, model performance, ease of
use, and purpose of application should clearly be established when choosing a dispersion model. The
review shows that complex Lagrangian or CFD models have greater data requirements and runtimes that
may not be readily available in an emergency scenario. As noted by Mikelonis et al. (2018), the choice of
modeling software may also be influenced by existing software and personnel expertise in an affected
location. Among EPA-developed models, users have a variety of choices in dispersion models since many
have been developed for research or regulatory use. Most of the model codes are open source and easily
obtainable. User support and troubleshooting is available, and models are well documented. As a result,
EPA's AERMOD dispersion model is a viable option to consider for future development for contaminant
fate and transport following a hazardous release event because of the availability of the source code and
robust model framework that has been extensively designed by, and evaluated with, field and laboratory
data. In particular, AERMOD could benefit from additional beta (test) options to better account for wind
profiles influenced by buildings and street canyons. Urban modifications or parameterizations may be
some of the most important developments for future versions of dispersion models used in emergency
preparation and response.
92

-------
lO.OReferences
Arya, S. P., 2001: Introduction to Micrometeorology, 2nd Ed. Academic Press, International Geophysics
Series Volume 79, 420 pp.
Arya, S. P., 1999: Air Pollution Meteorology and Dispersion. Oxford University Press, 310 pp.
ADMLC (Atmospheric Dispersion Modeling Liaison Committee), 2013: ADMLC-R7, January 2013,
Annex text available online at: https://admlc.com/publications/.
Barlow, J. F. and O. Coceal, 2009: A review of urban roughness sublayer turbulence, Met. Office
Technical Report No. 527.
Belcher, S. E., O. Coceal, E. V. Goulart, A. C. Rudd, and A. G. Robins, 2015: Processes controlling
atmospheric dispersion through city centres, J FluidMech, 763, 51-81.
Belcher, S. E., O. Coceal, J. R. C. Hunt, D. J. Carruthers, and A. G. Robins, 2013: A review of urban
dispersion modelling, Version 2. Technical Report prepared by the ADMLC, 5th December 2012,
96 pp., https://admlc.files.wordpress.com/2014/05/admlc-r7-2012-2-text.pdf, accessed January
10, 2018.
Ben Salem, N., V. Garbero, P. Salizzoni, G. Lamaison, and L. Soulhac, 2015: Modelling pollutant
dispersion in a street network, Bound. Layer. Meteorol., 155, 157-187.
Beychok, M. R., 1979: The Fundamentals of Stack Gas Dispersion, Milton R. Beychok, 201 pp.
Boe, T., P. Lemieux, D. Schultheisz, T. Peake, and C. Hayes, 2013: A Planning Tool for Estimating Waste
Generated by a Radiological Incident and Subsequent Decontamination Efforts-13569, WM2013
Conference, February 24-28, 2013, Phoenix, AZ.
Boris J. P., G. Patnaik, and T. Young, Jr., 2003: CT-Analyst: Verification and Validation, NRL Report 4-
1226-3377.
Bowers, J. F., C. A. Biltoft, and E. G. Astling, 1994: Meteorological Source Book for Diffusion Model
Applications, Chemical/Biological (CB) Technical Data Source Book, Volume XVII, DPG/JCP-
94-013, U.S. Army Dugway Proving Ground, Meteorology Division.
Bradley, M. M., 2005: NARAC: An Emergency Response Resource for Predicting the Atmospheric
Dispersion and Assessing the Consequences of Airborne Radionuclides, Conference Proceedings
from 2nd International Conference on Radiation in the Environment, Nice, France, UCRL-CONF-
214781.
Briggs, G. A., 1973: Diffusion estimation for small emissions, ATDL Report No. 79, NOAA, ARL, Oak
Ridge, TN.
Britter, R. E. and S. R. Hanna, 2003: Flow and Dispersion in Urban Areas, Annual Review of Fluid
Mechanics, 35, 469-496.
Britter, R. E. and J. McQuaid, 1988: Workbook on the Dispersion of Dense Gases, Health and Safety
Executive Contract Research Report No. 17/1988, Health and Safety Directorate, Sheffield, UK.
93

-------
Broughton, E., 2005: The Bhopal disaster and its aftermath: a review, Environmental Health, 4, 1-6.
Brown, D. F., W. A. Freeman, and W. D. Haney, 2017, Development of the Table of Initial Isolation
Distances and Protective Action Distances for the 2016 Emergency Response Guidebook,
ANL/GSS-17/1, August 2017, Argonne National Laboratory, Lemont, IL, 188 pp.
Brown, D. F., W. E. Dunn, and A. J. Policastro, 2001: A National Risk Assessment for Selected Materials
in Transportation, ANL/DIS-01-1, Argonne National Laboratory, Argonne, IL.
Brown, J. T., 2011: Presidential Policy Directive 8 and the National Preparedness System: Background
and Issues for Congress, CRS Report for Congress, Congressional Research Service, Report No.
R42073, 29 pp.
Brown, M. J., 2014: QUIC: A Fast, High-Resolution 3D Building-Aware Urban Transport and Dispersion
Modeling System, AWMA Environmental Manager, 4, 28-31.
Brown, M. J., A. Gowardhan, M. Nelson, M. D. Williams, and E. Pardyjak, 2013: QUIC transport and
dispersion modelling of two releases from the Joint Urban 2003 field experiment. International
Journal of Environment and Pollution, 52, 263-287.
Calfee, M. W, L. J. Rose, S. Morse, D. Mattorano, M. Clayton, A. Touati, N. Griffin-Gatchalian, C. Slone,
andN. McSweeney, 2013: Comparative Evaluation of Vacuum-Based Surface Sampling Methods
for Collection of Bacillus Spores, Journal of Microbiological Methods, 95, 389-96.
Castro, I. P., Z-T Xie. V. Fuka, A. G. Robins, M. Carpentieri, P. Hayden, D. Hertwig, and O. Coceal,
2017: Measurements and Computations of Flow in an Urban Street System, Boundary Layer
Meteorology, 162, 207-230.
Chang, R., J. Schaperow, T. Ghosh, J. Barr, C. Tinkler, and M. Stutzke, 2012: State-of-the-Art Reactor
Consequence Analyses (SOARCA) Report, Office of Nuclear Regulatory Research, United States
Nuclear Regulatory Commission, NUREG-1935, 200 pp.
Chang, J. C. and S. R. Hanna, 2010: Independent results validation of JEM for urban applications. Final
report prepared for JEM IV&V program office, Available from Chang Consultants, Fairfax, VA.
Chang, J. C., S. R. Hanna, Z. Boybeyi, and P. Franzese, 2005: Use of Salt Lake City URBAN 2000 Field
Data to Evaluate the Urban Hazard Prediction Assessment Capability (HPAC) Dispersion Model.
Journal of Applied Meteorology, 44, 485-501.
Chanin, D., M. L. Young, J. Randall, and K. Jamali, 1998: Code Manual for MACCS2: Volume 1, User's
Guide, Sandia National Laboratory, NUREG/CR-6613, SAND97-0594, Vol. 1, 325 pp.
Cimorelli, A. J., S. G. Perry, A. Venkatram, J. C. Weil, R. Paine, R. B. Wilson, R. F. Lee, W. D. Peters,
and R. W. Brode, 2005: AERMOD: A Dispersion Model for Industrial Source Applications. Part
I: General Model Formulation and Boundary Layer Characterization, J. Appl. Meteor., 44, 682-
693.
Cionco, R. M., 1965: A mathematical model for air flow in a vegetative canopy, J. Appl. Meteor., 4, 517-
522.
94

-------
Cionco, R. M., 1978: Analysis of canopy index values for various canopy densities, B. -Layer Meteor., 15,
81-93.
Coceal, O., E. V. Goulart, S. Branford, T. G. Thomas, and S. E. Belcher, 2014: Flow structure and near-
field dispersion in arrays of building-like obstacles, Journal of Wind Engineering and Industrial
Aerodynamics, 125, 52-68.
Cook, N. J., 1985: The designer's guide to wind loading of building structures: Part 7, Butterworths
Publishers, London.
Dabberdt, W. F., F. L. Ludwig, W. B. Johnson, 1973: Validation and applications of an urban diffusion
model for vehicular pollutants, Atmospheric Environment, 7, 603-618.
Dadosky, E. J., 2010: Interagency Modeling Atmospheric Assessment Center Local Jurisdiction: IMAAC
Operations Framework, M. S. Thesis, Naval Postgraduate School, Monterey, CA, 124 pp.
Deardorff, J. W., 1970: A numerical study of three-dimensional turbulent channel flow at large Reynolds
numbers, J Fluid Mech, 41, 453-480
DTRA (Defense Threat Reduction Agency), 2004: Hazard Prediction and Assessment Capability (HPAC)
4.04 Help Manual. Defense Threat Reduction Agency (DTRA), 6801 Telegraph Rd., Alexandria,
VA.
DePaul, F. T. and C. M. Sheih, 1986: Measurements of wind velocities in a street canyon, Atmospheric
Environment, 20, 455-459.
Draxler, R. R., B. Stunder, G. Rolph, A. Stein, and A. Taylor, 2020: HYSPLIT USER'S GUIDE, Version
5, Last Revision: April 2020, 249 pp, Available online at:
https://www.arl.noaa.eov/dociiments/reports/hysplit user euide.pdf.
Draxler, R. R., 1999: HYSPLIT 4 User's Guide, NOAA Technical Memorandum ERL ARL-230, June
1999, 35 pp.
EPA (Environmental Protection Agency), 2020: Homeland Security, Strategic Research Action Plan
(STRAP) 2019-2022, March 2020, EPA 601/K-20/002, 43 pp.
EPA, 2019: CMAQ (Version 5.3.1) [Software], Available from: https://doi.ore/10.5281/zenodo.3585898.
EPA, 2018: Survey and Assessment of Fate and Transport Models for Use Following a Wide-Area Urban
Release to Inform Mapping, Characterization, and Site Clearance, Report EPA/600/R-18/282,
September 2018, 65 pp., Research Triangle Park, NC.
EPA, 2017: The Emergency Planning and Community Right-to-Know Act Fact Sheet, Office of Land and
Emergency Management, November 2017, 7pp.
EPA, 2009a: Fact Sheet: Clean Air Act Section 112(r): Accidental Release Prevention / Risk Management
Plan Rule, EPA 550-R-09-002, March 2009, US EPA, Office of Solid Waste and Emergency
Response, 3pp.
EPA, 2009b: General RMP Guidance - Chapter 9: Risk Management Plan, March 2009, 10pp.
95

-------
EPA, 2009c: Risk Management Program Guidance for Offsite Consequence Analysis, EPA 550-B-99-
009, March 2009, US EPA, Office of Solid Waste and Emergency Response, 134pp.
EPA, 2004: General RMP Guidance - Chapter 4: Offsite Consequence Analysis, April 2004, 27pp.
EPA, FEMA, and US DOT, 1987: Technical Guidance for Hazards Analysis: Emergency Planning for
Extremely Hazardous Substances, December 1987, 194 pp.
Fabbri, L., M. H. Wood, I. Azzini, and R. Rosati, 2020: Global sensitivity analysis of the ADAM
dispersion module: Jack Rabbit II test case, Atmospheric Environment, in press.
Fabbri, L. and M. H. Wood, 2019: Accident Damage Analysis Module (ADAM): Novel European
Commission tool for consequence assessment—Scientific evaluation of performance, Process
Safety and Environmental Protection, 129, 249-263
FEMA (Federal Emergency Management Agency), 2010: Developing and Maintaining Emergency
Operations Plans, Comprehensive Preparedness Guide (CPG) 101, version 2, November 2010,124
pp.
FEMA, 1998: IS-010, Animals in Disaster: Module A, Awareness and Preparedness, Emergency
Management Institute, 185 pp.
Garbero, V., 2008: Pollutant dispersion in urban canopy study of the plume behavior through an obstacle
array, Ph.D. dissertation, DIASP, Politecnico di Torino - LMFA, Ecole Centrale de Lyon, Italy.
Gifford, F. A., 1961: Use of routine meteorological observations for estimating atmospheric dispersion,
Nucl. Safety, 2, 47-51.
Gowardhan, A., B. Pobanz, S. Neuscamman, J. Donetti, H. Walker, R. Belles, B. Eme, S. Homann, M.
Simpson and J. Nasstrom, 2018: National Atmospheric Release Advisory Center's Urban Plume
Dispersion Modeling Capability for Radiological Sources, June 19, 2018, Lawrence Livermore
National Laboratory, LLNL-PRES-751301.
Grimmond, C. S. B. and T. R. Oke, 1999: Aerodynamic properties of urban areas derived from analysis
of surface form, J. Appl. Meteorol., 38, 1261-1292.
Halitsky, J.,1968: Gas diffusion near buildings, In "Meteorology and Atomic Energy" (D. H. Slade, ed.),
pp. 221-255, Tech. Info. Cent., U.S. Department of Energy, Oak Ridge, TN.
Hamlyn, D., T. Hilderman, and R. Britter, 2007: A simple network approach to modelling dispersion
among large groups of obstacles, Atmos Environ, 41, 5848-5862
Hall D. J., A. M. Spanton, M. Hargrave, S. Walker, I. H. Griffiths, and D. R. Brooks, 2004: The Urban
Dispersion Model (UDM): Version 2.9 Technical Documentation. Defence Science and
Technology Laboratory, Physical Sciences Department, Porton Down, Salisbury, UK.
Hall, D. J., A. M. Spanton, I. H. Griffiths, M. Hargrave, and S. Walker, 2002: The Urban Dispersion
Model (UDM): Version 2.2, Defence Science and Technology Laboratory Tech. Doc.
DSTL/TR04774, 106 pp.
Hanna, S. R., and J. C. Chang, 2012: Acceptance criteria for urban dispersion model evaluation,
Meteorology and Atmospheric Physics, 116, 133-146.
96

-------
Hanna, S. R., J. C. Chang, and D. G. Strimaitis, 1993: Hazardous gas model evaluation with field
observations, Atmospheric Environment. Part A. General Topics, 27, 2265-2285.
Hanna S. R., G. A Briggs, R. P. Hosker, 1982: Handbook on Atmospheric Diffusion, Atmospheric
Turbulence and Diffusion Laboratory, National Oceanic and Atmospheric Administration,
DOE/TIC-11223, 102 pp.
Heist, D. K., L. A. Brixey, J. Richmond-Bryant, G. E. Bowker, S. G. Perry, and R. W. Wiener, 2009: The
effect of a tall tower on flow and dispersion through a model urban neighborhood, Part 1. Flow
characteristics, J. Environ. Monitoring, 11, 2163-2170.
Hendricks, C. G. S., and M. J. Hall, 2007a: The history and science of CBRNE agents, part I, Chemist,
84, 1-18.
Hendricks, C. G. S., and M. J. Hall, 2007b: The history and science of CBRNE agents, part II, Chemist,
84, 1-18.
Hertwig, D., L. Soulhac, V. Fuka, T. Auerswald, M. Carpentieri, P. Hayden, A. Robins, Z.-T. Xie, and O.
Coceal, 2018: Evaluation of fast atmospheric dispersion models in a regular street network,
Environmental Fluid Mechanics, 18, 1007-1044.
Hill, A., 2003: Using the Hazard Prediction and Assessment Capability (HPAC) Program for Radiological
Scenarios Relevant to the Australian Defence Force, Defence Science and Technology, CBRN
Defence Centre, Platforms Sciences Laboratory, DSTO-CR-0294, 28 pp.
Holmes, N. S. and L. Morawska, 2006: A review of dispersion modelling and its application to the
dispersion of particles: An overview of different dispersion models available, Atmospheric
Environment, 40, 5902-5928.
Homann, S. G. and F. Aluzzi, 2014: HotSpot Health Physics Codes, Version 3.0 User's Guide, National
Atmospheric Release Advisory Center, Lawrence Livermore National Laboratory, Livermore,
CA, Report No. LLNL-SM-636474, August 27, 2014, 198 pp.
Jones, R., W. Lehr, D. Simecek-Beatty, and R. M. Reynolds, 2013: ALOHA® (Areal Locations of
Hazardous Atmospheres) 5.4.4 Technical Documentation, NOAA Technical Memorandum NOS
OR&R 43, National Oceanic and Atmospheric Administration, Office of Response and
Restoration, Emergency Response Division, Seattle, WA.
Leitl, B., F. Harms, E. Berbekar, J. Boris, G. Patnaik, K. Obenschain, and S. Fischer, 2016: Local-scale
hazmat dispersion modelling for first responders based on high-resolution computational fluid
dynamics-an overview of CT-Analyst Hamburg, Chemical Engineering Transactions, 48, 43-48.
Lindsay, B. R., 2012: Federal Emergency Management: A Brief Introduction, CSR Report for Congress,
Congressional Research Service, Report No. R42845, November 30, 2012, 32 pp.
Lucas, D. D., A. Gowardhan, P. Cameron-Smith, and R. L. Baskett, 2016: Impact of meteorological inflow
uncertainty on tracer transport and source estimation in urban atmospheres, Atmos. Environ., 143,
120-132.
MacDonald, R. W., 2000: Modeling the mean velocity profile in the urban canopy layer, B. -Layer Meteor.,
97, 25-45.
97

-------
Mazzola, C. A., R. P. Addis, and Emergency Management Laboratory Oak Ridge Institute for Science
and Education, 1995: Atmospheric Dispersion Modeling Resources, 2nd Edition, Prepared for the
United States Department of Energy, Emergency Management Advisory Committee,
Subcommittee on Consequence Assessment and Protective Actions, March 1995, 278 pp,
Available online at: http://www6.uniovi.es/ema/admr.pdf.
Mikelonis, A., T. Boe, W. Calfee, and S. D. Lee, 2018: Urban Fate and Transport Modeling of
Contaminants: The Unique Needs of Emergency Response and the Potential for Adapting Existing
Models, Journal of Water Management Modeling, 26, 1-7.
Miner, S., T. Mazzola, S. Herring, R. Fry, and R. Meris, 2019: Performance of Hazard Prediction and
Assessment Capability Urban Models for the UDINEE Project, Boundary-Layer Meteorol., 171,
423-437.
Monbureau, E. M., D. K. Heist, S. G. Perry, L. H. Brouwer, H. Foroutan, W. Tang, 2018: Enhancements
to AERMOD's building downwash algorithms based on wind-tunnel and Embedded-LES
modeling, Atmospheric Environment, 179, 321-330.
Mongeon, A., 2018: IMAAC - Butterflies to Fukushima, presentation at the 22nd Annual George Mason
University Conference on Atmospheric Transport and Dispersion Modeling, June 2018.
Nasstrom, J. S., G. Sugiyama, R. L. Baskett, S. C. Larsen, and M. M. Bradley, 2007: The National
Atmospheric Release Advisory Center (NARAC) Modeling and Decision Support System for
Radiological and Nuclear Emergency Preparedness and Response, Int. J. Emergency Management,
4, 1-32.
NAS (National Academies of Sciences, Engineering, and Medicine), 2018: Strategies for Effective
Improvements to the BioWatch System: Proceedings of a Workshop, Planning Committee for a
Workshop on Strategies for Effective Biological Detection Systems Board on Life Sciences;
Division on Earth and Life Studies, The National Academies Press, Washington, DC, 222 pp
National Research Council, 2003: Tracking and Predicting the Atmospheric Dispersion of Hazardous
Material Releases: Implications for Homeland Security, The National Academies Press,
Washington, DC, 93 pp.
Nelson, M. and M. Brown, 2013: The QUIC Start Guide (v6.01) The Quick Urban & Industrial Complex
(QUIC) Dispersion Modeling System. Report No. LA-UR-13-27291, Prepared by Los Alamos
National Laboratory, Los Alamos, NM.
Nelson, M. A., M. D. Williams, D. Zajic, E. R. Pardyjak, and M. J. Brown, 2009: Evaluation of an Urban
Vegetative Canopy Scheme and Impact on Plume Dispersion, AMS 8th Symp. Urban Env., Report
No. LA-UR-09-00068, Phoenix AZ.
Nelson, M., B. Addepalli, F. Hornsby, A. Gowardhan, E. Pardyjak, and M. Brown, 2008: Improvements
to fast-response urban wind model, 15th AMS/A WMA Met. Aspects Air Poll., Report No. LA-UR-
08-0206, 6 pp.
Nicholson, S. E., 1975: A pollution model for street-level air, Atmospheric Environment, 9, 19-3.
98

-------
OFCM (Office of the Federal Coordinator for Meteorology), 2004: Federal Research and Development
Needs and Priorities for Atmospheric Transport and Diffusion Modeling, September 2004, FCM-
R23-2004, Silver Spring, MD.
OFCM, 2002: Atmospheric Modeling of Releases from Weapons of Mass Destruction: Response by
Federal Agencies in Support of Homeland Security, August 2002, FCM-R17-2002, Silver Spring,
MD.
OFCM, 1999: Directory of Atmospheric Transport and Diffusion Consequence Assessment Models,
FCM-I3-1999, Silver Spring, MD.
OFCM, 1993: Directory of Atmospheric Transport and Diffusion Models, Equipment, and Projects, FCM-
13-1993, Silver Spring, MD.
Oke, T.R., 1988: Street design and urban canopy layer climate, Energy and Buildings, 11, 103-113.
Pasquill, F. and F. B. Smith, 1983: Atmospheric Diffusion, 3rd ed., Ellis Horwood Ltd Publishers,
Chichester, UK, 437 pp.
Pasquill, F., 1974: Atmospheric Diffusion, 2nd ed. Halstead Press.
Pasquill, F., 1962: Atmospheric Diffusion, D. Van Nostrand Co. Ltd, New York.
Pasquill, F., 1961: The Estimation of the Dispersion of Windblown Material, Meteorological Magazine,
90, 33-49.
Perry, R. D. and J. D. Fetherston, 1997: Yersinia pestis—Etiologic Agent of Plague, Clinical Microbiology
Reviews, 1, 35-66.
Philips, D. A., R. Rossi, and G. Iaccarino, 2013: Large-eddy simulation of passive scalar dispersion in an
urban-like canopy. Journal ofFluid Mechanics,723, 404-428.
Radzikowski, P., 2008: 'Reach-Back'—A New Approach To Asymmetrical Warfare Intelligence, Army
Magazine, 58(12), December 2008, 2 pp.
Reynolds, O., 1894: On the dynamical theory of incompressible viscous fluids and the determination of
the criterion, Phil. Trans. Roy. Soc. London, A97, 354-373.
Richardson, L. F., 1922: Weather Prediction by Numerical Process, Cambridge University Press.
Rockle, R., 1990: Bestimmung der stomungsverhaltnisse im bereich komplexer bebauungsstrukturen,
Ph.D. dissertation, Fachbereich Mechanik, der Technischen Hochschule, Darmstadt, Germany.
Runyon, S., 2017: IMAAC as a Decision-making Support Capability for CBRNE Incidents, Defense Threat
Reduction Agency Technical Reachback Division, Plenary Presentation at the 2017 ESRI National
Security and Public Safety Summit, San Diego, CA.
Ryan, S., W. Calfee, J. Wood, B. Attwood, and F. Schaefer, 2010: Research to Support the
Decontamination of Surfaces and Buildings Contaminated with Biothr eat Agents, In Disinfection,
Sterilization, and Antisepsis: Principles, Practices, Current Issues, New Research, and New
99

-------
Technologies, edited by William Anthony Rutala. Arlington, VA: APIC (Association for
Professionals in Infection Control and Epidemiology).
Schmidtgoessling, R. D., 2009: Shelter-In-Place: Indoor Exposure Assessment During an Airborne
Chemical, Biological, Radiological, And Nuclear (CBRN) Event. M.S. thesis,
AFIT/GIH/ENV/09-M02, Department of the Air Force, Air University, Air Force Institute of
Technology, Wright-Patterson Air Force Base, Ohio.
Scire, J. S., D. G. Strimaitis, R. J. Yamartino, 2000: A User's Guide for the CALPUFF Dispersion Model
(Version 5), Earth Tech, Inc., Concord, MA, 521 pp.
Slade, D. H., 1968: Meteorology and Atomic Energy 1968, TID-24190, U.S. Atomic Energy Agency,
Washington, D.C.
Soulhac, L., G. Lamaison, F. X. Cierco, N. B. Salem, P. Salizzoni, P. Mejean, P. Armand, L. Patryl, 2016:
SIRANERISK: modelling dispersion of steady and unsteady pollutant releases in the urban
canopy, Atmos Environ, 140, 242-260.
Soulhac, L., P. Salizzoni, P. Mejean, D. Didier, I. Rios, 2012: Themodel SIRANE for atmospheric urban
pollutant dispersion; part II, validation of the model on a real case study, Atmos Environ, 49, 320-
337.
Soulhac, L., P. Salizzoni, F. X. Cierco, R. Perkins, 2011: The model SIRANE for atmospheric urban
pollutant dispersion; part I, presentation of the model. Atmos Environ, 45, 7379-7395.
Soulhac, L., 2000: Modelisation de la dispersion atmospherique a l'interieur de la canopee urbaine, PhD
Thesis, Ecole Centrale de Lyon, 351 pp.
Spicer, T. and J. Havens, 1989: User's Guide for the DEGADIS 2.1 Dense Gas Dispersion Model, EPA-
450/4-89-019, US Environmental Protection Agency, Research Triangle Park, NC, 419 pp.
Sykes R. I., S. Parker, D. Henn, and B. Chowdhury, 2007: SCIPUFF Version 2.3 Technical
Documentation, L-3 Titan Corp, Princeton, NJ.
Stein, A. F., R. R. Draxler, G. D. Rolph, B. J. Stunder, M. D. Cohen, and F. Ngan, 2015: NOAA's
HYSPLIT Atmospheric Transport and Dispersion Modeling System, Bull. Amer. Meteor. Soc., 96,
2059-2077.
Stewart, S., 2014: The Biggest Threat Dirty Bombs Pose is Panic, Forbes, September 11, 2014, retrieved
on August 16, 2018 from: https://www.forbes.com/sites/workday/2018/08/08/ taking-the-next-
step s-for-tom orro ws-tal etrt/#23 a95 3 5 064d 1.
Stull, R. B., 1988: An Introduction to Boundary Layer Meteorology, Kluwer Acad. Publ., Dordrecht,
Boston, London, 666 pp.
Taylor, G. I., 1921: Diffusion by continuous movements, Proc. London Math Soc, Ser. 2., 20, 196-202.
Tominaga, Y. and T. Stathopoulos, 2013: CFD simulations of near-field pollutant dispersion in the urban
environment: a review of current modelling techniques, Atmospheric Environment, 79, 716-730.
100

-------
Turner, D. B., 1979: Atmospheric dispersion modeling, a critical review, J. Air Pollut. Control Assoc., 29,
502-519.
Turner, D. B., 1970: Workbook of Atmospheric Dispersion Estimates, US EPA, ASRL, Research Triangle
Park, NC.
U.S. GAO (US Government Accountability Office), 2008: First Responders' Ability to Detect and Model
Hazardous Releases in Urban Areas Is Significantly Limited. Report No. GAO-08-180, Prepared
by US GAO, June 27, 2008 in Washington DC., https://www.gao.gov/products/GAO-08-180,
accessed December 15, 2017.
U.S. Department of Homeland Security (DHS), 2016: National Response Framework, edited by US DHS.
U.S. NRC (U.S. Nuclear Regulatory Commission), 2018: Backgrounder on Dirty Bombs, fact sheet
updated May 25, 2018 and available online at: https://www.nrc.gov/reading-rm/doc-
collections/fact-sheets/fs-dirtv-bombs.html, retrieved August 16, 2018.
Van de Walle, B. and M. Turoff, 2008: Decision support for emergency situations, Information Systems
and e-Business Management, 6, 295-316.
Van Leuken, J. P. G., A. N. Swart, A .H. Havelaar, A.Van Pul, W. Van der Hoek, D.Heederik, 2016:
Atmospheric dispersion modelling of bioaerosols that are pathogenic to humans and livestock - A
review to inform risk assessment studies, Microbial Risk Analysis, 1, 19-39.
Vardoulakis, S., B. Fisher, K. Pericleous, and N. Gonzalez-Flesca, 2003: Modelling air quality in street
canyons: a review. Atmospheric Environment, 37, 155-182.
101

-------
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
United States
Environmental Protection
Agency
Office of Research and
Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300

-------