EPA/600/R-05/149
November 2005
TMDL Model Evaluation and Research
Needs
By
Leslie Shoemaker, Ting Dai, and Jessica Koenig
Tetra Tech, Inc.
Fairfax, Virginia 22030
Contract 68-C-04-007
Mohamed Hantush
Land Remediation and Pollution Control Division
National Risk Management Research Laboratory
Cincinnati, Ohio 45268
National Risk Management Research Laboratory
Office Of Research And Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
Abstract
The report was submitted in fulfillment of contract number 68-C-04-007 by Tetra Tech, Inc., under the sponsorship
of the United States Environmental Protection Agency. This review examines the modeling research needs to
support environmental decision-making for the 303(d) requirements for development of total maximum daily loads
(TMDLs) and related programs such as 319 Nonpoint Source Program activities, watershed management,
stormwater permits, and National Pollutant Discharge Elimination System (NPDES) discharge evaluations. By
examining the currently available models and considering the needs for TMDLs and related watershed programs, a
comprehensive list of modeling research needs can be developed.
More than 65 currently available models were evaluated for their capabilities and applicability to TMDL
development and related watershed management activities. Evaluation tables were developed to facilitate
comparison of models and inventory the potential gaps in model capabilities, and fact sheets were developed for
models to provide more detailed information on the capabilities of each model. Existing integrated models systems
were also evaluated and compared, based on data processing, modeling tools, and model linkages supported. The
review of available models demonstrates that many of the dominant pollutant types and waterbodies can be
simulated using available technologies. However, many specific technical gaps remain, especially in linkages
between air, surface water, groundwater and receiving water models.
The model reviews and emerging trends in technology were considered in developing a comprehensive list of
research needs that encompass a variety of sources, processes, waterbodies, data, systems, and integration needs.
This diversity of needs is consistent with the current development of TMDLs across the country. Initially, TMDL
development focused on dominant source and pollutant types, but more recently, emphasis has shifted to completing
TMDLs under a variety of site-specific conditions and supporting more detailed implementation planning. Because
of the specialized and diverse characteristics of the needs, an equitable prioritization of specific needs cannot be
defined. Key recommended research areas that could benefit multiple applications include: integrated best
management practice (BMP) modeling systems, more physically based representation of watersheds, and support for
linkage of watershed and receiving water models.
The review recommends that this diverse set of technical needs should be supported by new and more flexible
modeling systems and tools. Development of integrated modeling systems can provide the commonly needed tools
and support adoption of new solution techniques, source representation, and algorithms. Providing integrated
system platforms, ideally Internet-based, can help minimize duplication of effort (shared on line data management,
data display, shared resources), while maximizing resources for more fundamental development and research of key
components. The use of Internet-based technologies has now emerged as a viable and practical medium for
management of data, analysis techniques and tools to support TMDL and more generalized watershed analyses.
Development of a standardized Internet-based framework could provide significant cost saving for the management
and application of models. In addition, a standardized and open framework, with clearly defined linkage
capabilities, could encourage research and continuous testing and update of new components.
Future development of models and the supporting infrastructure of data and guidance can support informed
environmental decision-making, improve understanding of the physical systems in our world, and ultimately provide
information to support the effective restoration and protection of the nation's waters.
-------
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 National Risk Management Research Laboratory (NRMRL) is the Agency's center for investigation of
technological and management approaches for preventing and reducing risks from pollution that threaten human
health and the environment. The focus of the Laboratory's research program is on methods and their cost-
effectiveness for prevention and control of pollution to air, land, water, and subsurface resources; protection of
water quality in public water systems; remediation of contaminated sites, sediments and ground water; prevention
and control of indoor air pollution; and restoration of ecosystems. NRMRL collaborates with both public and
private sector partners to foster technologies that reduce the cost of compliance and to anticipate emerging
problems. NRMRL's research provides solutions to environmental problems by: developing and promoting
technologies that protect and improve the environment; advancing scientific and engineering information to
support regulatory and policy decisions; and providing the technical support and information transfer to ensure
implementation of environmental regulations and strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan. It is published
and made available by EPA's Office of Research and Development to assist the user community and to link
researchers with their clients.
Sally Gutierrez, Director
National Risk Management Research Laboratory
in
-------
IV
-------
Contents
Abstract ii
Foreword iii
Contents v
Figures vii
Tables viii
Acronyms and Abbreviations ix
Acknowledgments xiii
Chapter 1 Introduction 1
Chapter 2 Modeling Needs for TMDL Development 3
TMDL Modeling Requirements 4
Analysis Categories 6
Model Selection Considerations 11
Chapters What is a Model? 15
Model Complexity 15
Alternatives Analysis 16
Model Development 16
Integrated Modeling Systems and Linked Models 17
Trends in Model Development 18
Chapter 4 Available Models 21
Chapters Applicability of Models 35
Application Criteria 35
Capabilities and Limitations of Currently Available Models 52
Integrated Modeling Systems 54
Chapter 6 Case Studies 59
Development of Mercury TMDLs in Arivaca Lake and Pefia Blanca Lake, Arizona 59
Background and Problem Identification 59
Source Assessment 60
Model Selection 62
Model Setup 67
Model Evaluation 69
Model Application 72
Development of a Nutrient TMDL in the Cahaba Paver, Alabama 74
Background and Problem Identification 74
Source Assessment 77
Model Selection 79
Model Setup 80
Model Evaluation 84
-------
Model Application 86
Chapter? Research Needs 91
Methodology for Identifying Research Needs 91
Model Capabilities 92
Sources 93
Hydrology 94
Sediment Loading 96
Pathogens 97
Nutrient Loading Simulation 98
Mercury 99
Other Pollutants and Toxics 99
Chloride and Selenium 100
Management Practice Simulation 100
Hydrodynamics 101
Sediment Transport 102
Nutrients and Eutrophication 103
Ecological/Habitat 104
Data 104
Model Defensibility 105
Systems Development and Supporting Tools 107
Integrated Modeling Systems 107
Conclusions 108
References Ill
Appendix: Model Fact Sheets 129
VI
-------
Figures
Figure 2-1. Ten major analysis categories 6
Figure 3-1. Steps for performing allocation analysis 16
Figure 6-1. Location of Upper Cahaba River watershed, Alabama 75
Figure 6-2. Locations of sites utilized for ecoregion reference-stream analysis 76
Figure 6-3. Locations of major (>1.0 MOD) NPDES-permitted point source discharges in the
Upper Cahaba River watershed 77
Figure 6-4. Water quality sampling sites used to assess nonpoint source concentrations of TP and other nutrients.. 78
Figure 6-5. Nonpoint source concentrations of TP as a function of percent urban area 79
Figure 6-6. MRLC land use aggregation calculated by LSPC subbasin delineation 81
Figure 6-7. Precipitation sites with 3 years of hourly or daily data used in the LSPC watershed model 81
Figure 6-8. Example of LSPC GIS interface for selecting headwater subbasins 82
Figure 6-9. Examples of EPDRIV1 cross-sections derived from FEMA survey data 83
Figure 6-10. Schematic of the functional relationship of data inputs in the Cahaba Spreadsheet Model 84
Figure 6-11. Example of hydrologic model calibration: model and observed streamflow 85
Figure 6-12. Example of monthly streamflow predictions compared to USGS data at seven sites 85
Figure 6-13. Estimated monthly median TP concentrations in the Cahaba River in September 1999, from
Trussville (upstream at left) to Centreville (downstream at right). Gray lines indicate maximum and
minimum monthly predicted TP concentrations based on streamflow variation 86
Figure 6-14. Cahaba River nutrient TMDL evaluation points upstream of critical reaches 87
Figure 6-15. TMDL scenario, growing-season median TP for 1999-2001 and three=year average 88
VII
-------
Tables
Table 2-1. Top 10 303(d) List Impairments in the United States 4
Table 2-2. Summary of Analysis Sequences for Analysis Categories 8
Table 4-1. Summary of Available Models 22
Table 4-2. Summary of Receiving Water Simulation Capabilities 27
Table 4-3. Summary of Watershed Simulation Capabilities 29
Table 4-4. Summary of Management Practice Simulation Capabilities 31
Table 4-5. Overview of Models-Review Categories 33
Table 5-1. TMDL Endpoints Supported 37
Table 5-2. General Land and Water Features Supported 40
Table 5-3. Special Land Features Supported 43
Table 5-4. Special Water Features Supported 47
Table 5-5. Application Considerations 50
Table 5-6. Capabilities of Integrated Modeling Systems 56
Table 6-1. Cross-Sectional Comparison of Studied Lakes 65
Table 6-2. Comparison of Summer Hypolimnetic Water Chemistry between Studied Lakes 66
Table 6-3. Sensitivity Analysis on the Ruby Dump Sediment Load Multiplier 70
Table 6-4. Sensitivity Analysis on the Ball Mill Site Sediment Load Multiplier 70
Table 6-5. D-MCM Calibration for Pena Blanca, Arivaca, and Patagonia Lakes 72
Table 6-6. TMDL Allocations for Pena Blanca and Arivaca Lakes 73
Table 6-7. Summary of Median Nutrient Concentrations for April-October at the ADEM Reference Stations in
Ecoregion67 76
Table 6-8. Sites Within the Cahaba River Watershed Not Impacted by Point Sources 78
Table 6-9. TMDL Summary for Cahaba River Phased TP Reductions 88
Table 6-10. Existing and Predicted Instream Growing Season Median TP in the Cahaba River 89
Vlll
-------
Acronyms and Abbreviations
ADEM
AGNPS
AGWA
AnnAGNPS
ARS
BASINS
BMP
BOD
CAEDYM
CFR
CH3D-IMS
CH3D-SED
CONCEPTS
CREM
CWA
DCIA
DDT
DEM
DIAS/IDLMAS
DMR
DOC
DWSM
ECOMSED
ED AS
EFDC
ELM
EMC
EPA
EPIC
FEMA
FMS
CIS
GISPLM
GLEAMS
GLLVHT
Alabama Department of Environmental Management
Agricultural Nonpoint Source Pollution Model
Automated Geospatial Watershed Assessment
Annualized Agricultural Nonpoint Source Pollution Model
Agricultural Research Service (USDA)
Better Assessment Science Integrating Point and Nonpoint Sources
Best management practice
Biochemical oxygen demand
Computational Aquatic Ecosystem Dynamics Model
Code of Federal Regulations
Curvilinear-grid Hydrodynamics SD-Integrated Modeling System
Curvilinear Hydrodynamics 3D-Sediment Transport
Conservational Channel Evolution and Pollutant Transport System
Council on Regulatory Environmental Modeling
Clean Water Act
Directly connected impervious areas
Dichloro-Diphenyl-Trichloroethane
Digital elevation model
Dynamic Information Architecture System/Integrated Dynamic Landscape Analysis
and Modeling System
Discharge monitoring report
Dissolved organic carbon
Dynamic Watershed Simulation Model
Estuary and Coastal Ocean Model with Sediment Transport
Ecological Data Application System
Environmental Fluid Dynamics Code
Everglades Landscape Model
Event mean concentration
U.S. Environmental Protection Agency
Erosion Productivity Impact Calculator
Federal Emergency Management Agency
Flexible Modeling System
Geographic information system
GIS-Based Phosphorus Loading Model
Groundwater Loading Effects of Agricultural Management Systems
Generalized, Longitudinal-Lateral-Vertical Hydrodynamic and Transport
IX
-------
CMS
GSSHA
GWLF
HEC-6
HEC-6T
HEC-HMS
HEC-RAS
HSCTM-2D
HSPF
HUC
KINEROS2
LA
LEAM
LSPC
MCM
Mercury Loading Model
MOD
MINTEQA2
MMS
MODIS
MOS
MRLC
MS4
MUSLE
NEXRAD
NHD
NPDES
P8-UCM
PCB
PCSWMM
PGC-BMP
ppb
PRMS
QUAL2E
REMM
RF3
scs
SED3D
SLAMM
SME
SNTEMP
Groundwater Modeling System
Gridded Surface Subsurface Hydrologic Analysis
Generalized Watershed Loading Functions
Scour and Deposition in Rivers and Reservoirs
Sedimentation in Stream Networks
Hydraulic Engineering Center-Hydrologic Modeling System
Hydrologic Engineering Center-River Analysis System
Hydrodynamic, Sediment, and Contaminant Transport Model
Hydrologic Simulation Program—FORTRAN
Hydrologic Unit Code
Kinematic Runoff and Erosion Model, v2
Load allocation
Land-Use Evolution and Impact Assessment Model
Loading Simulation Program in C++
Mercury Cycling Model
Watershed Characterization System—Mercury Loading Model
Million gallons per day
Metal Speciation Equilibrium Model for Surface and Ground Water
Modular Modeling System
Moderate Resolution Imaging Spectroradiometer
Margin of safety
Multi-Resolution Land Characteristics
Municipal separate storm sewer system
Modified Universal Soil Loss Equation
Next Generation Weather Radar
National Hydrography Dataset
National Pollutant Discharge Elimination System
Program for Predicting Polluting Particle Passage through Pits, Puddles, and Ponds—
Urban Catchment Model
Polychlorinated biphenyl
Stormwater Management Model
Prince George's County Best Management Practice Module
Parts per billion
Precipitation-Runoff Modeling System
Enhanced Stream Water Quality Model
Riparian Ecosystem Management Model
Reach File, version 3
Soil Conservation Service
Three-Dimensional Numerical Model of Hydrodynamics and Sediment Transport in
Lakes and Estuaries
Source Loading and Management Model
Spatial Modeling Environment
Stream Network Temperature Model
-------
SPARROW
SSTEMP
STORM
SWAT
SWMM
TMDL
TN
TP
TSS
USAGE
USDA
USFWS
USGS
USLE
WAMView
WARMF
WASP
WEPP
WinHSPF
WLA
WMS
WQBEL
WRDB
WWTP
XP-SWMM
SPAtially Referenced Regression On Watershed Attributes
Stream Segment Temperature Model
Storage, Treatment, Overflow, Runoff Model
Soil and Water Assessment Tool
Storm Water Management Model
Total maximum daily load
Total nitrogen
Total phosphorus
Total suspended solids
U.S. Army Corps of Engineers
U.S. Department of Agriculture
U.S. Fish and Wildlife Service
U.S. Geological Survey
Universal Soil Loss Equation
Watershed Assessment Model with an Arc View Interface
Watershed Analysis Risk Management Framework
Water Quality Analysis Simulation Program
Water Erosion Prediction Project
Interactive Windows Interface to HSPF
Wasteload allocation
Watershed Modeling System
Water quality-based effluent limits
Water Resources Database
Wastewater treatment plant
Stormwater and Wastewater Management Model
XI
-------
Xll
-------
Acknowledgments
The authors would like to thank Mohamed Hantush, Project Officer, and technical reviewers, Dr. Zhonglong
Zhang, U.S. Army Corps of Engineers, Environmental Laboratory, and Dr. Deva Borah, Illinois State Water
Survey, for their insightful and thorough comments. The authors would also like to thank Dr. John Hamrick
and Dr. Jenny X. Zhen of Tetra Tech, Inc., who provided significant contributions to this document, including
conducting multiple model reviews to develop model evaluation tables and fact sheets and identify research
needs. In addition, the authors would like to acknowledge their fellow members of the Tetra Tech TMDL and
Modeling Center whose experience in developing thousands of TMDLs was integral to the development of this
review and who supported this project with model reviews, examples, and excellent suggestions for future
needs. Specifically, the following people developed model fact sheets and case studies to support this
document:
William Anderson
Khalid AM
Clint Boschen
SenBai
Jon Butcher
Steve Carter
Mustafa Faizullabhoy
Yoichi Matsuzuru
Sabu Paul
John Riverson
Eric Thirolle
Brian Watson
The authors would also like to acknowledge the editorial review staff for their excellent review capabilities.
Xlll
-------
XIV
-------
Chapter 1 Introduction
Models are used to answer questions, support decision-making, and assess alternatives. Models are used as a tool to
describe and understand the dynamics of physical systems including watersheds and receiving waters such as lakes,
rivers, estuaries, and coastal areas. Analysts use models to answer questions such as:
How do inputs from human sources and land management activities affect the conditions of our receiving
waters?
How should these inputs be changed to benefit the condition of our receiving waters?
Exploration of these cause and effect relationships drives the need to develop and apply models. However, the
development of models that can reliably represent watersheds and receiving waters is also a tremendous challenge.
The analyst must consider how to represent the system with sufficient accuracy to have confidence in the result.
Practical and technical constraints require that the representation of the system is consistent with available
information, time, resources and scientific understanding. Each modeling application needs to address this
challenge. Each modeling application must address the challenge of balancing the needs of the particular study at an
appropriate level of accuracy and reliability.
Current environmental protection programs rely on modeling to evaluate and select various control strategies.
Historically, models have been used to derive water quality-based effluent limits (WQBELs) for point source
discharges. Large-scale national estuary program activities (e.g., Chesapeake Bay, Tampa Bay), which represent
some of our most significant resources, have used models to determine allowable loading of nutrients and restoration
needs. More recently, the Total Maximum Daily Load (TMDL) program has required the determination of loading
allocations that will result in restoring waters designated as impaired on state 303 (d) lists. In many cases, models
are used during TMDL development to evaluate the relationship between load reduction and compliance with water
quality standards. Models provide a "linkage" between loads and receiving water conditions. To address the TMDL
development needs for a highly diverse group of impaired waters and pollutant sources, a wide range of model
techniques is required. Although models are available and have been used successfully in environmental and water
quality management since the 1970s, the diversity of the pollutants, sources, and receiving water conditions to be
evaluated under the TMDL program have placed new challenges on the use of modeling to support environmental
decision-making.
Watershed management planning also increasingly relies on modeling to develop restoration goals and identify load
reduction needs. U.S. Environmental Protection Agency (EPA) released Nonpoint Source Program and Grants
Guidelines for States and Territories (October 23, 2003) for grants appropriated by Congress in Fiscal Year 2004
and in subsequent years. The guidelines, available online at http://www.epa.gov/fedrgstr/EPA-
WATER/2003/October/Day-23/w26755.htm, identify the following nine minimum elements that must be included
in watershed plans funded by Section 319:
1. Causes and sources
2. Pollutant load reduction estimates
3. Management measures needed
4. Technical and financial assistance required to implement
5. Information and education activities
6. Implementation schedule
-------
7. Interim measurable milestones
8. Indicators to evaluate progress toward load reductions and water quality standards
9. Monitoring program
Addressing these guidelines, especially elements 2 and 3, places increased emphasis on developing watershed plans
that result in meeting water quality standards and demonstrate a linkage between pollutant sources and water quality.
The specific limitations of models for TMDL support and watershed management are not well documented. States
and EPA continue to face new challenges as they address more complex waterbodies, impairments, and sources, for
which there is limited experience. As models and supporting systems evolve, research should be targeted to those
areas where our analysis capabilities need to be improved. As technology and various Internet-based applications
and mapping systems continue to improve, new modeling systems (e.g., one or more linked models) and supporting
analysis tools (e.g., data preparation, output display, optimization, uncertainty analysis) will clearly be needed.
This review will examine the modeling research needs to support environmental decision-making and programs
such as 303 (d)-related development of TMDLs, implementation of 319 Nonpoint Source Programs, watershed
management, stormwater permits, and National Pollutant Discharge Elimination System (NPDES) discharge
evaluations. By examining the currently available models and considering the needs for TMDLs and related
watershed programs, a comprehensive list of modeling needs can be developed.
The first task of the review process was to identify and evaluate currently available model capabilities. The second
task evaluates the ability of models to simulate the TMDL- and watershed-related load reduction needs and
management related alternatives. The evaluation also considers model performance, simulation capabilities, level of
effort, training, and user interfaces. Supporting this evaluation are case studies of two successful TMDL modeling
applications for nutrients and mercury. Finally, the review provides recommendations for key areas of research to
address the modeling needs and fill the gaps identified in the review of available models. The evaluation also
considers the need for various software and supporting tools, including linkages between existing models, and
further integration of new and emerging modeling resources to address multiple media (e.g., air, surface waters,
groundwater).
This report is organized as follows:
• Chapter 2 discusses modeling needs for TMDL development, including programmatic and technical issues
that affect model selection and application for TMDL development.
• Chapter 3 provides background on models, including what a model is, processes and levels of complexity
represented in available models, and recent trends in model development.
• Chapter 4 discusses the types of models available and identifies the models included in this review.
• Chapter 5 evaluates the applicability of the reviewed models and integrated modeling systems to TMDL
development and watershed management applications and provides an evaluation of their capabilities.
• Chapter 6 includes two case study applications: Case Study #1 - Development of Mercury TMDLs in
Arivaca Lake and Pena Blanca Lake, Arizona; Case Study #2 - Linked Model Development for Nutrient
TMDL in Cahaba River, Alabama
• Chapter 7 discusses modeling needs and recommendations for future research.
• Appendix includes detailed fact sheets on each of the models included in the review.
-------
Chapter 2 Modeling Needs for TMDL Development
Section 303(d) of the Clean Water Act (CWA) and the EPA's Water Quality Planning and Management Regulations
(40 Code of Federal Regulations [CFR] Part 130) require states to identify and list those waters within their
boundaries that are water quality-limited, to prioritize them, and to develop TMDLs for the pollutants of concern.
States develop and submit the 303 (d) list that defines water quality-limited waters to EPA for review and approval.
Water quality-limited waters are waterbodies that do not meet applicable water quality standards or are not expected
to meet applicable standards after application of technology-based effluent limitations for point sources.
A TMDL is the allowable load of a specific pollutant that
can be discharged into a waterbody and meet water
quality standards. TMDLs consist of wasteload
allocations (WLAs) for point sources, load allocations
(LAs) for nonpoint sources, and a margin of safety
(MOS). A TMDL is based on the relationship between
pollutant sources and water quality and provides the
scientific basis for a state to establish water quality-based
controls to reduce pollutant loads from both point and
nonpoint sources to restore and maintain the quality of
the state's water resources. The load expressed is not
necessarily a daily load but is expressed in the
appropriate averaging period that protects water quality
standards. TMDLs, once implemented should result in
meeting the states' water quality standards.
Across the country, thousands of waters are listed as
impaired by a wide variety of pollutants. Based on most
recent state 303(d) lists, there are approximately 34,000
impaired waters in the United States and more than
59,000 associated impairments (National Section 303(d)
List Fact Sheet, http ://www. epa. gov/o wo w/tmdl/.
accessed August 1, 2005). Metals, pathogens, nutrients
and sediment are the most common pollutants included
on state lists, and the top 10 listed impairments account
for over 75 percent of the total listings in the nation (Table 2-1). Since January 1, 1996, EPA has approved almost
15,000 TMDLs, accounting for approximately 25 percent of the nationwide listings.
Loading capacity. The greatest amount of loading that a
water can receive without violating water quality standards. (40
CFR130.2(f))
Load allocation (LA). The portion of a receiving water's
loading capacity that is attributed either to one of its existing or
future nonpoint sources of pollution or to natural background
sources. (40 CFR 130.2(g))
Wasteload allocation (WLA). The portion of a receiving
water's loading capacity that is allocated to one of its existing
or future point sources of pollution. (40 CFR 130.2(h))
Total maximum daily load (TMDL). The sum of the individual
WLAs for point sources and LAs for nonpoint sources and
natural background. TMDLs can be expressed in terms of
either mass per time, toxicity, or other appropriate measure.
(40 CFR 130.2(i)) TMDLs must be calculated with seasonal
variations and a margin of safety and must take into account
critical conditions for stream flow, loading, and water quality
parameters. (40 CRF 130(c)(1))
Margin of safety (MOS). TMDLs must be established with a
margin of safety that takes into account any lack of knowledge
concerning the relationship between effluent limitations and
water quality. (40 CFR 130.7(c)) EPA guidance explains that
the MOS may be implicit (i.e., incorporated into the TMDL
through conservative assumptions in the analysis) or explicit
(i.e., expressed in the TMDL as a portion of the loading
capacity).
-------
Table 2-1. Top 10 303(d) List Impairments in the United States
General Impairment Name1 Number Reported Percent Reported Cumulative Percent
Metals 11,526 19.2 19.2
Pathogens 7,896 13.2 32.4
Nutrients 5,585 9.3 41.7
Sediment/siltation 5,045 8.4 50.1
Organic enrichment/low 4,406 7.3 57.4
dissolved oxygen
Fish consumption advisories 3,178 5.3 62.7
pH 2,904 4.8 67.5
Other habitat alterations 2,389 4.0 71.5
Thermal modifications 2,200 3.7 75.2
Biological criteria 2,116 3.5 78.7
1 "General impairment" categories may represent several associated pollutants or impairment listings. For example, the "metals"
category includes 30 specific pollutants or related listings (e.g., iron, lead, contaminated sediments).
TMDL Modeling Requirements
Models are often used to support development of TMDLs—typically to estimate source loading and evaluate
loading capacities that will meet water quality standards. The technical requirements of a TMDL stipulate that
analysis should demonstrate that the allocation of point and nonpoint source loads would result in meeting water
quality standards. The wording of the TMDL requirements also stipulates that WLA and LA must be separately
defined. For modeling purposes, this requirements means that point and nonpoint sources must be evaluated as
separate sources so that they can be simulated under various loading scenarios. Point sources are typically
represented individually to accommodate TMDL regulations, which require "individual waste load allocations" for
point sources. For wet weather or diffuse point sources (e.g., stormwater), the municipal boundaries may need to be
addressed in the modeling analysis as well (USEPA 2002). For nonpoint sources, TMDL guidance identifies that
allocation can be made to individual sources, categories or subcategories of sources. In cases of limited data, LAs
can also be expressed as "gross allotments," allowing for larger scale grouping of nonpoint source.
In the development of a TMDL, load allocations might also be identified that affect sources upstream of a listed
water depending on the transport properties of the pollutant. The need to look at sources upstream of the listed
water necessitates a "watershed-based" approach to TMDLs. The contributions upstream of the listed water are all
potentially part of the solution to the impairment. Although TMDLs are developed for specific listed waters and
their associated watersheds, the TMDL analyses are sometimes developed in "bundles" to address groups of listed
waters that are located within a larger collective watershed. This grouping of TMDL analyses can result in more
efficient TMDL development and review.
EPA's 1991 guidance discusses the option of developing TMDLs using the "phased approach," often referred to as
"adaptive management." Under the phased approach, TMDL development should be based on estimates that use
available data and information, but monitoring for collection of new data would be required (USEPA 199la). The
phased approach provides for further pollution reduction without waiting for new data collection and analysis. The
-------
margin of safety developed for the TMDL under the phased approach should reflect the adequacy of data and the
degree of uncertainty about the relationship between load allocations and receiving water quality (USEPA 1991a).
The TMDL program does not dictate or require the use of any particular model or modeling procedures. NRC
(2001) discusses the use of modeling in TMDL development, supporting the assertion that there is no recommended
model for TMDLs but that any model chosen for TMDL development should meet a set of criteria related to the
specific TMDL issues (e.g., water quality standards, data availability, cost).
TMDL Regulations and Guidance
Water Quality Planning And Management Regulations. Code of Federal Regulations, Title 40, Part 130. Regulations are available
online at http://www.apoaccess.gov/cfr/index.html.
USEPA. 2002. The Twenty Needs Report: How Research Can Improve the TMDL Program. EPA841-B-02-002. U.S.
Environmental Protection Agency, Office of Water, Office of Wetlands, Oceans and Watersheds, Washington, DC.
http://www.epa.gov/owow/tmdl/20needsreport 8-02.pdf
USEPA. 2002. Establishing Total Maximum Daily Load (TMDL) Wasteload Allocations (WLAs) for Storm Water Sources and
NPDES Permit Requirements Based on Those WLAs. Memorandum from Robert H. Wayland, III, Director, Office of Wetlands,
Oceans and Watersheds, and James A. Hanlon, Director, Office of Wastewater Management, U.S. Environmental Protection
Agency, Washington, DC. November 22, 2002. http://www.epa.gov/npdes/pubs/final-wwtmdl.pdf
USEPA. 2001. Protocol for Developing Pathogen TMDLs. EPA 841-R-00-002. U.S. Environmental Protection Agency, Office of
Water, Washington, DC. 132 pp. http://www.epa.gov/owow/tmdl/pathogen all.pdf
USEPA. 1999. Protocol for Developing Nutrient TMDLs. EPA 841-B-99-007. U.S. Environmental Protection Agency, Office of
Water, Washington, DC. 135 pp. http://www.epa.gov/owow/tmdl/nutrient/pdf/nutrient.pdf
USEPA. 1999. Protocol for Developing Sediment TMDLs. EPA 841-B-99-004. U.S. Environmental Protection Agency, Office of
Water, Washington, DC. 132 pp. http://www.epa.gov/owow/tmdl/sediment/pdf/sediment.pdf
USEPA. 1997. New Policies for Establishing and Implementing Total Maximum Daily Loads (TMDLs). Memorandum from Robert
Perciasepe, Assistant Administrator, Office of Water, U.S. Environmental Protection Agency, Washington, DC. Augusts, 1997.
http://www.epa.gov/OWOW/tmdl/ratepace.html
USEPA. 1997. Compendium of Tools for Watershed Assessment and TMDL Development. EPA841-B-97-006. U.S. Environmental
Protection Agency, Office of Water, Washington, DC.
USEPA. 1991. Guidance for Water Quality-Based Decisions: The TMDL Process. EPA 440/4-91 -001. U.S. Environmental
Protection Agency, Office of Water, Washington, DC. http://www.epa.gov/OWOW/tmdl/decisions/
EPA has provided guidance for major pollutant types
that identify modeling and analysis needs and
generalized approaches (USEPA 1999a; USEPA
1999b, USEPA 2001). Other modeling related
guidance from USEPA mostly relates to the
requirements for record keeping and documentation.
All TMDL reports, models, and documentation are
subject to public review and comment. Record
keeping and documentation of all modeling code and
software are recommended as part of the
administrative records. The need for review by
USEPA and open comment periods for stakeholders
has resulted in a strong preference for public domain or
open code modeling systems for application in TMDL
development.
Administrative Records
While not a necessary part of the TMDL submittal, USEPA
recommends preparation of an administrative record containing
documents that support the establishment of and
calculations/allocations in the TMDL. Components of the record
should include all materials relied upon to develop and support
the calculations/allocations in the TMDL, including any data,
analyses, or references that were used, records of
correspondence with stakeholders and USEPA, responses to
public comments, and other supporting materials. This record
is needed to facilitate public and/or USEPA review of the
TMDL.
From: Guidelines for Reviewing TMDLs Under Existing
Regulations Issued in 1992
http://www.epa.gov/owow/tmdl/guidance/final52002.html
-------
Freshwater
Lakes - Metals
Lakes - Nutrients
Rivers - Metals/Pesticides
(active & legacy)
Rivers - Pathogens
Rivers - Nutrients
Rivers - Sediment
Rivers - Temperature
-
Analysis Categories
Understanding the types of impairments that occur throughout
the country can assist in identifying the types of models that are
needed to investigate impairments, diagnose causes of
impairment, and identify management solutions. The waterbody
and general impairment types can be grouped into 10 dominant
analysis categories (Figure 2-1). The grouping of analysis
categories is based on defining the physical processes associated
with the major waterbody types and the characteristics of
impairments and associated pollutants. The physical processes
of water movement characterize the major waterbody types
including lakes, rivers, and tidal estuaries and coastal areas.
Lakes and reservoirs are impounded waters, and the major
processes are associated with relatively static water systems.
Rivers are typically flowing systems characterized by velocity
and volume of flow. Riverine studies normally focus on
concentration of pollutants in flowing water. In tidal waters,
analyses focus on the dominant processes related to tidal flux,
salinity, and mixing. Similar to the physical conditions
associated with waterbodies, pollutants can also be broadly
grouped into nutrients, metals, pathogens, sediment and temperature. Nutrients are typically associated with
eutrophication-related effects; metals and toxics are associated with sediment and water column criteria violations;
elevated concentrations of sediment and deposition of fine-grained sediment are associated with fish habitat
impairment; and in-stream temperature is a stressor associated with fish habitat.
These analysis categories illustrate the diversity of conditions that result in the selection of an appropriate analytical
approach. Recent and current TMDL development efforts have resulted in the development of many TMDLs
representative of the 10 major categories of frequently observed waterbody-impairment combinations. Examination
of recent TMDLs and experience in the development of TMDLs were used to develop a general summary of the
typical sequence of models or analysis techniques that have been used (Table 2-2; 10 major combinations, and 1
combination having two types of pesticides—active and legacy). This table is not intended to provide specific
guidance on the selection of modeling approaches for TMDLs but to illustrate the typical sequences of approaches
and the diversity of techniques employed. In addition, the table demonstrates how the selection of the appropriate
modeling techniques is tightly linked to waterbody type and impairment. The table is organized according to the
defining features of the analyses used in developing a TMDL.
Figure 2-1. Ten major analysis categories.
• Impairment Conditions - TMDLs are a plan to meet water quality standards. An understanding of when
and under what conditions impairment occurs is needed to determine the type of analysis needed.
• Delivery of Pollutants - Source loading can be delivered directly from discharges and from precipitation-
driven processes. The type of impairment is often related to timing of pollutant delivery.
• Modeling Approach - The selected approach will include a combination of watershed loading and
receiving water response models or other estimation techniques.
• Model Output Used to Calculate Loading Capacity - Output from the model(s) is processed to provide a
representation of the loading capacity that meets the water quality standards.
• Typical Implementation - This feature qualitatively or quantitatively discusses the types of practices that
could be used to achieve the loading target.
• Sample Case Studies - Case studies illustrate the types of approaches described in each column of the
table.
-------
Several observations relevant to modeling for TMDL development can be made from Table 2-2:
• TMDLs often require multiple models to address the watershed and receiving water components (see
Category II, which includes Generalized Watershed Loading Functions [GWLF] and BATHTUB).
• Some pollutants are addressed through the use of statistical/analytical techniques that are not formal
"models" or "modeling systems" (see Category VII).
• Although many models are available, the 11 categories presented can be addressed by using a relatively
small set of models.
• Processing of model output is needed to specifically evaluate the TMDL's compliance with the water
quality standard.
• The example models shown are typically public domain and/or EPA-supported models.
• Explicit modeling of best management practices (BMPs) is not typically included or required in a TMDL
analysis.
However, examining the historical use of models for TMDL development is only a reflection of how the currently
available models have been applied and does not explain the limitations in current models to address specific needs
or local considerations. Many additional waterbodies will require TMDL development, and, in some areas, TMDLs
will be revised to improve on previous analysis or reallocate source loading. If more efficient modeling systems or
more robust, process-based models are developed, users are likely to adopt the new technology.
-------
Table 2-2. Summary of Analysis Sequences for Analysis Categories
Category
I
Impairment
Conditions
River - Pathogens
Storm events or warm weather, dry
season periods
Lake - Nutrients
Summer/dry season
River- Nutrients
Summer/dry season/year-round
Delivery of Pollutants Storm event runoff or dry weather
discharge, direct deposition
Stormwater runoff, dry weather
inflows, point sources
Dry weather inflows (point source
discharges, nonpoint sources,
groundwater)
Modeling Approach
General
Approach
Watershed
Loading
Receiving
Water
Response
Wet weather and dry weather
pathogen analysis
Flow, concentration, and load
estimation using HSPF
In-stream response using HSPF (data
collection consideration)
Eutrophication analysis to identify
nutrient loading thresholds to meet in-
lake targets
Load estimation using GWLF,
AGNPS, AnnAGNPS, SWAT,
SWMM, or HSPF
Lake response using BATHTUB.
More detailed option using CE-QUAL-
W2 or EFDC.
Low- or high-flow analysis of nutrient
loading thresholds to meet in-stream
targets
Load estimation based on tributary
and point source low-flow monitoring
Stream response using mass
balance, QUAL2E low-flow model, or
WASP
Model Output Used
to Calculate Loading
Capacity
Number of exceedance days based
on model output or monitoring data
and comparison with reference
watershed
Loading of nitrogen and phosphorus
needed to meet lake target as
simulated by lake model
Loading or concentration for critical
low-flow or average summer, or high-
flow periods
Typical Targeted management of pathogen
Implementation sources: stormwater, rural uses,
septics
Targeted management of nutrient Targeted dry or weather reductions
sources: stormwater, rural uses, open from point sources, dry season
space uses, septics, point sources nonpoint sources
Sample Case Santa Monica, CA
Studies http://www.swrcb.ca.aov/rwgcb4/html/
meetings/tmdl/tmdl ws santa monica
.html
Lake Ontelaunee, PA
http://www.epa.gov/reg3wapd/tmdl/pa
tmdl/Lake%20ontelauneeTMDL/inde
x.htm
Wissahickon Creek, PA
http://www.epa.gov/reg3wapd/tmdl/pd
f/wissahickon tmdl/index.htm
-------
Table 2-2. Summary of Analysis Sequences for Analysis Categories (continued)
IV
V
VI
Category
River- Pesticides/Urban
(Active Pesticide Sources)
River - Pesticides/Legacy
(No Current Pesticide Sources)
River/Estuary- Toxics
Impairment Mixed. Associated with application
Conditions dates and days when transport
occurs
Mixed. Associated with disturbance or Mixed
resuspension of historical deposits
Delivery of Pollutants
Urban runoff, typically storm drains.
Dry weather discharges including
irrigation and dumping
Historic delivery. Resuspension due
to storm events, aquatic life
Municipal and industrial wastewater,
urban runoff, agricultural runoff, other
sources
I
0)
•g
General Identification of reduction needed to
Approach meet water column toxicity-based
targets
Identification of reduction needed to
meet sediment, fish tissue, or water
column water quality toxicity-based
targets
Identification of reduction needed to
meet sediment, fish tissue or water
column toxicity-based targets
Watershed
Loading
Source characterization
Tributary monitoring
Source characterization
Receiving Allowable loading determination
Water based on calculation from identified
Response target at design flow or a range of
flows
Allowable loading determination
based on calculation from identified
target at design flow or a range of
flows
Allowable loading determination
based on calculation from identified
target at design flow or a range of
flows
Model Output Used
to Calculate Loading
Capacity
Allowable load for design flow or
annual period
Allowable load for design flow or
annual period
Allowable load for design flow or
annual period
Typical Reduction or elimination of active
Implementation pesticide sources
Removal or stabilization of deposits,
long-term attenuation
Reduction or elimination of active toxic
sources
Sample Case San Francisco Bay Area Urban
Studies Creeks Pesticide Toxicity/Diazinon
TMDL, CA
http://www.swrcb.ca.aov/rwgcb2/urb
ancrksdiazinontmdl.htm
Newport Bay, CA
http://www.epa.gov/region09/water/tm
dl/final.html
Newport Bay, CA
http://www.epa.gov/region09/water/tm
dl/final.html
-------
Table 2-2. Summary of Analysis Sequences for Analysis Categories (continued)
Category
VII
VIM
River - Sediment
River - Temperature
IX
River - Biological
Impairment
Conditions
Nonseasonal: estuary infilling, pool filling
Spring: spawning/incubation
All seasons: rearing
Winter: migration (turbidity-related)
Summer/dry-warm weather
Multiple/dry-wet season
Delivery of
Pollutants
Storms and throughout the wet season over a wide
range of flows
Summer heat input
Depends on
pollutants/stressors
associated with the
impaired conditions
General . Long-term loading analysis based on sediment
Approach budget and reference approach. Sediment
source analysis if full budget not possible
• Turbidity/total suspended solids (TSS) events
• Sedigraphs (combination of flow and
turbidity/TSS data)
g Watershed
Temperature estimation based on
flow, solar inputs, stream geometry,
meteorologic conditions, vegetative
shading, and other factors
Biological reference
approach, load estimation
for identified pollutants
.
<
0)
"8
Loading
Load estimation using sediment budget or
sediment source analysis
Estimation of inputs based on sediment yields
and delivery from land use/erosion categories
Temperature estimation based on
models of flow, travel time,
solar/meteorologic conditions.
Shade models do not address
watersheds with dams or high levels
of irrigation return flows, or cooling
water discharges.
Load estimation of
identified pollutant(s)
contributing to biological
impairment using GWLF or
similar model
Receiving . Load target determined from comparison with
Water desired reference watershed
Response . Rate of infilling
• Geomorphic/habitat targets derived from
literature
SSTEMP or SNTEMP stream flow Comparison of estimated
and temperature analysis, watershed/source loads
QUAL2E stream flow and with loads in reference
temperature analysis watershed
Model Output Used
to Calculate
Loading Capacity
Average annual sediment load from dominant
sources to meet reference conditions.
Identification of achievable reductions by source
category
Heat loading
Shade dominated streams
Effective shade allocations
(percent of stream shade)
Annual loading
benchmarked to reference
watershed
Typical Targeted management of sediment sources for
Implementation long-term restoration
Targeted management of vegetation
and stream system, dam releases,
irrigation withdrawals, or return flows
Targeted management of
relevant pollutant sources
Sample Case Garcia River, CA
Studies http://www.epa.aov/reaion09/water/tmdl/final.html
North Fork Eel River, CA
http://www.epa. aov/region09/water/t
mdl/final.html
Cooks Creek, VA
http://www.deg.state.va.us/
tmdl/tmdlrpts.html
SNTEMP = Stream Network Temperature Model
SSTEMP = Stream Segment Temperature Model
10
-------
Table 2-2. Summary of Analysis Sequences for Analysis Categories (continued)
XI
Category
Estuary- Nutrients
Coastal - Pathogen
Impairment
Conditions
Die-off of macrophytes, floating maps, algal blooms
Spring runoff or winter and summer dry weather
Delivery of Pollutants Annual/long-term nutrient loading from runoff, nutrients
associated with sediment, groundwater
Runoff/wet weather sources or dry weather sources
Direct deposition
General
Approach
Long-term loading, nutrient cycling, and response of
estuaries
Wet weather loading and response of estuaries
_c Watershed
ro Loading
Load estimation using GWLF, HSPF, analyses of
monitoring data, or similar model
Load estimation using HSPF or direct analysis of
monitoring data
^ Receiving
"g Water
Response
Estuary response using Tidal Prism, WASP, EFDC, or
similar model
Response using WASP, EFDC, or similar model
Alternatively determine correlation of coastal impairment
with tributary loading
Model Output Used Annual loading based on meeting estuary target condition
to Calculate Loading
Capacity
Wet and dry weather exceedance frequencies and
associated loading
Typical Targeted management of nutrient and sediment sources:
Implementation stormwater, rural uses, open space uses, septics, point
sources, irrigation return flows, fertilizer management
Targeted management of pathogen sources: stormwater,
rural uses, septics
Sample Case
Studies
(Several available nationally)
Santa Monica, CA
http://www.swrcb.ca.aov/rwacb4/html/meetinqs/tmdl/tmdl
ws santa monica.html
Model Selection Considerations
When addressing any impairment, selecting the appropriate model is crucial in developing a feasible, defensible and
equitable TMDLs and load allocations. The primary factor in determining the modeling approach for a TMDL, as
demonstrated by the analytical categories, is the pollutant and associated endpoint that represents compliance with
water quality standards. The TMDL endpoint is a numeric threshold that is equated with compliance with water
quality standards. Some TMDL endpoints are derived directly from numeric water quality criteria and have a
defined magnitude, duration, and frequency (e.g., zinc expressed as a concentration in pg/1, 4-day average, l-in-3
year frequency of exceedence). Some endpoints are derived based on interpreting narrative criteria to derive a
numeric endpoint. For example, a waterbody impaired by nuisance algal blooms could result in a TMDL that
defines nutrient loads (total phosphorus [TP] and total nitrogen [TN]) that will result in meeting a summer
chlorophyll a endpoint of 20 pg/1. Endpoints designed to address acute (short-term) impairments are typically based
on instantaneous maximums or daily averages while chronic (long-term) problems (e.g., eutrophication, sediment
loading and deposition) are represented by endpoints with longer durations (e.g., monthly average concentration,
annual loading). The applicability of a model for a specific TMDL application is evaluated based on the ability to
simulate at a time-scale and resolution appropriate for evaluation of the endpoint's magnitude, duration and
frequency. For example, if an endpoint is based on a maximum daily concentration, a model that provides output of
only monthly average concentration is not appropriate.
11
-------
Other factors that guide model selection for TMDLs
include defining the waterbody type, sources,
necessary spatial resolution (e.g., gross watershed vs.
subwatershed vs. site-scale) and special local features
(e.g., surface-groundwater interactions) or land
features (e.g., wetlands). If management solutions
will be evaluated, model selection must consider the
types of management techniques, spatial scale of the
information, and degree of specificity required in the
management alternatives analysis.
Special processes or technical considerations that
affect pollutant loading and impairment conditions
and, therefore, TMDL development are of particular
concern in the appropriate selection and application of
models. Models that include more complex processes
are typically more difficult to apply, require
significantly more data for setup and testing, and
might include untested algorithms. However, not all
of the identified technical considerations may be
crucial to include in a particular application—they
may not have enough effect on the outcome of the
results to merit selecting a more complex modeling
approach to address them. The standard practice in
modeling is to identify the dominant processes and
identify the simplest models sufficient to meet the
needs of the project.
In developing TMDLs, some key technical
complexities or issues are often identified. Some have
been addressed in simplified approaches or through
the use of statistical or site-specific models. Many of
these key technical considerations could be addressed
through additional research and integration of new
physically based modeling techniques and are
discussed further in Chapter 6. Specific technical
issues that have been encountered as considerations or limitations in TMDL development include the following:
Water quality standards. Provisions of state or federal law which
consist of a designated use or uses for the waters of the United
States, water quality criteria for such waters based upon such
uses. Water quality standards are to protect public health or
welfare, enhance the quality of the water and serve the purposes
of the Act (40 CFR 131.3(1))
Designated uses. Those uses specified in water quality
standards for each waterbody or segment whether or not they are
being attained (40 CFR 131.3(f))
Criteria. Elements of state water quality standards, expressed as
constituent concentrations, levels, or narrative statements,
representing a quality of water that supports a particular use.
When criteria are met, water quality will generally protect the
designated use (40 CFR 131.3(b))
Numeric Criteria. Numeric criteria or limits exist for many
common pollutants, such as high concentrations of bacteria,
suspended sediment, algae, dissolved metals, etc. An example of
numeric criteria is "dissolved oxygen must be at least five
milligrams per liter" (i.e., dissolved oxygen > 5.0 mg/L). Numeric
criteria are based on laboratory and other studies that test or
otherwise examine pollutant impacts on live organisms from
different species such as frogs, fish, and insect larvae.
Narrative Criteria. Non-numeric descriptions of desirable or
undesirable water quality conditions. An example of narrative
criteria might be that "all waters shall be free from sludge, floating
debris, oil and scum, color and odor-producing materials,
substances that are harmful to human, animal or aquatic life, and
nutrients in concentrations that may cause algal blooms."
Chronic. Defines a stimulus that lingers or continues for a
relatively long period of time, often one-tenth of the life span or
more. Chronic should be considered a relative term depending on
the life span of an organism. The measurement of a chronic effect
can be reduced growth, reduced reproduction, etc., in addition to
lethality (USEPA1991 b).
Acute. Refers to a stimulus severe enough to rapidly induce an
effect; in aquatic toxicity tests, an effect observed in 96 hours or
less is typically considered acute. When referring to aquatic
toxicology or human health, an acute affect is not always
measured in terms of lethality (USEPA 1991b).
Sediment - Stream bank erosion and channel adjustments can be substantial sources of sediment in urban
and rural areas and are difficult to characterize and quantify.
Irrigation - Irrigation can significantly alter the natural hydrology of an area, and irrigation return flows
are a significant source of pollutants in arid regions.
Drainage - Drainage tile can affect the hydrologic response of the watershed and can provide discharges to
rivers.
High water table areas - In areas of high water tables, water fluctuations, and surface-groundwater
interactions affect runoff and pollutant delivery. Wetland areas can retain water and affect water quality.
Wetlands - Large areas of wetlands influence watershed hydrology, loading, and management options, and
areas with wetting and drying can influence tidal areas (i.e., estuaries).
Contaminated sediment - Contaminated sediment is subject to several processes that influence their
interaction with the water column and aquatic biota, including accumulation, movement, burial, and
12
-------
dredging. In many areas, contaminated sediments are the result of historical sources, limiting management
options.
• Pesticides - Application of pesticides can result in water contamination through overspray or rainfall
during or immediately after application. However, pesticide monitoring data are limited, and few tested
simulations are available.
• Ecological impairments - Ecological or habitat related parameters/stressors are poorly understood or
cannot be modeled directly.
• Best management practices - Data on existing management techniques are often not readily available, and
techniques are difficult to represent accurately in models.
A number of issues that affect model selection and application for TMDL development are not related to the
technical representation of the system. General issues related to the overall TMDL process and the use of models
include:
• Data availability - Many TMDLs have limited local monitoring data available to support analysis, limiting
the potential approach options as well as the confidence in modeling approaches (i.e., no data for
calibration). TMDL guidance does not require data collection and supports the use of "available data and
information" when possible (USEPA 1991a).
• Accuracy of models - TMDLs are developed with a wide range of approaches and models. Regardless of
the approach, there are often concerns regarding oversimplifications, insufficient data or lack of confidence
in complex models. No guidance is available that specifically recommends the level of accuracy expected
for modeling studies.
Water quality standards formulation - TMDLs often involve some degree of interpretation of applicable
water quality standards. Standards may be narrative and may require a determination of a representative
numeric value. Other examples include criteria that are not defined precisely (e.g., no frequency or
duration) or situations where the parameter included in the standards is not suited to modeling (e.g.,
turbidity). Models may not be able to directly predict the endpoint associated with the water quality
standard.
• Pollutant focus of TMDLs - TMDLs are developed to
address specific impairments associated with a pollutant
as identified on the 303(d) list (40 CFR 130.7(c)).
Protection of designated uses, such as aquatic life
support, may require an approach that addresses
multiple stressors and the cumulative benefit on the
impaired water. Some stressors, such as low flow or
poor habitat, that affect aquatic life uses are not defined
as pollutants under the CWA (§ 502(6)) and therefore
do not require TMDLs. Watershed studies, which may
or may not include a TMDL, may need to address
broader ecological modeling or multiple stressor
analyses.
Pollutant vs. Pollution
Pollutant. Dredged spoil, solid waste, incinerator residue,
sewage, garbage, sewage sludge, munitions, chemical
wastes, biological materials, radioactive materials, heat,
wrecked or discarded equipment, rock, sand, cellar dirt
and industrial, municipal, and agricultural waste
discharged into water. (CWA § 502(6)).
Pollution. Generally, the presence of matter or energy
whose nature, location, or quantity produces undesired
environmental effects. In 40 CFR 130.2(c), pollution is
defined as "The man-made or man-induced alteration of
the chemical, physical, biological, and radiological
integrity of water."
13
-------
14
-------
Chapter 3 What is a Model?
The term "model" describes the set of equations or algorithms that are used to simulate a physical system. In this
report, "model" also refers to a variety of available software tools that automate the calculation of equations or
groups of equations representing the system. To address a specific technical problem, an analyst may choose to
apply an existing model, apply multiple models alone or in combination, modify an existing model, or develop a
site-specific model. Each application of a model must be designed to meet the analytical needs of the specific
system.
Model Complexity
Models are developed at various levels of complexity,
depending on the application needs. The simplest models
provide general predictions based on a limited set of
environmental or physical factors. One example might be
a loading rate model that provides an estimate of annual
pollutant load as a simple function of land use type.
These simplified approaches group physical processes
and provide generalized estimates of response. Empirical
equations, that build functional relationships based on
long-term studies and statistical analysis, are typically
used in simplified models (e.g., Universal Soil Loss
Equation [USLE]). Simplified techniques, due to their
inherent generalizations, are limited in applicability to
the various pollutants and waterbodies addressed by
TMDLs. On the other end of the spectrum are physically
based models that seek to describe the fundamental
processes that are associated with water, sediment, and
pollutant movement, transport, transformation, and
delivery. Physically based models describe fundamental
processes, such as infiltration, through the use of
scientifically based equations. The most sophisticated models will solve fundamental equations on a detailed spatial
and temporal scale. Physically based models require additional data to estimate the various parameters used in the
solution techniques. For example, infiltration calculations might require detailed information on precipitation,
evaporation, slope, soil conductivity, soil profiles, and vegetated cover.
The complexity of models is also a function of the spatial representation of the heterogeneity of the watershed or
waterbody. The simplest watershed models group large areas by land use category. Similarly, a simplified lake
model represents the lake as one large unit. More detailed watershed models will represent land areas as a network
with grid cells that have defined land and soil features. Other watershed models compromise by using a
hydrologically based network of subwatersheds and stream segments to represent the system. Within each
subwatershed, the individual land use units are "lumped" based on similar characteristics (e.g., land use, soils,
slope). This lumped approach simplifies the physical representation of the subwatershed and does not distinguish
between small parcels and contiguous parcels of land areas within a subwatershed. Models are also distinguished by
their temporal complexity - with simple models that use long timesteps (i.e., annual, seasonal) and detailed models
that have timesteps of hours or minutes. Even the most detailed models are built using a combination of empirical
and physically based techniques, with varying degrees of flexibility for users to select the spatial and temporal scale
of the application.
Additional Modeling Definitions
Field scale. Taking place at the subbasin or smaller level.
Field scale modeling usually refers to geographic areas
composed of one land use (e.g., a cornfield).
Lumped model. A model in which the physical characteristics
for land units within a subwatershed unit are assumed to be
homogeneous.
Mechanistic model. A model that attempts to quantitatively
describe a phenomenon by its underlying causal mechanisms.
Numerical model. Model that approximates a solution of
governing partial differential equations that describe a natural
process. The approximation uses a numerical discretization of
the space and time components of the system or process.
Steady state model. Mathematical model of fate and transport
that uses constant values of input variables to predict constant
values of receiving water quality concentrations.
Dynamic model. A mathematical formulation describing the
physical behavior of a system or a process and its temporal
variability.
15
-------
Alternatives Analysis
Models can help evaluate and quantify the potential effects of alternative plans and operational schemes and help
understand the relationships between natural systems and human influences on those systems. For TMDLs and
watershed studies, models are particularly useful in evaluating the likely benefits and drawbacks associated with
various loading alternatives and their effect on specific quantifiable endpoints that represent compliance with water
quality standards. Central to the application of models is support for alternatives analysis. Modeling analyses can
be used to test multiple scenarios, with various allocations to nonpoint and point sources (i.e., LA and WLAs).
Among the scenarios, one or more may meet water quality standards; however, some distributions of loading may
not be technically feasible or accepted by stakeholders. Using the model to develop an understanding of the
magnitude of each source, its geographic location, and the sensitivity of the receiving water to changes in source
loading supports the selection of feasible or preferred allocations. The allocation or alternatives analysis is typically
performed based on the following discrete steps, as illustrated in Figure 3-1:
Step 1: Application of the Model to Existing
Conditions - This application represents the
current condition (e.g., observed water quality,
current loadings) and is compared to available
monitoring data for model testing and
calibration. Point sources are set at
representative discharge concentrations,
reflective of permit monitoring data.
Representative concentrations may be lower or
higher than permit limits.
Total Load
Load Reduction
Figure 3-1. Steps for performing allocation analysis.
Step 2: Application of the Model to Existing
Conditions with Point Sources at Permit
Limits - This application establishes the
baseline condition, which will be reduced to
meet the allowable load. The point sources are
set at permit limits for flow and pollutant concentration. If no permitted flow is available, the design flow
or historic observed flow can be used. If the permit does not include a permit limit for the affected
pollutant, then the observed concentration can be used.
Step 3: Application of the Model to Future Conditions - When future growth is considered, it can be
added to the nonpoint or point source loading contributions.
Step 4: Develop and Test Allocation Scenarios - Working from the baseline condition (Step 2, or Step 3 if
future growth is considered), sample allocation scenarios are applied with a variety of source reductions.
These scenarios are shown as A, B, and C in Figure 3-1. The results of each scenario are compared with
the applicable water quality standard, and scenarios are adjusted until water quality standards (or loading
capacity) are achieved.
Step 5: Select Final TMDL Scenario - Once the final TMDL scenario is selected, results are processed to
provide the required TMDL elements.
Model Development
Models include suites of equations that represent key processes. Conventional wisdom indicates that the simplest
model sufficient to answer management questions with confidence should be applied. Analysts will need to
consider the level of complexity needed and appropriate for a given application. Additional complexity, in the form
of more detailed simulation of physically based processes and higher spatial and temporal resolution, requires more
experience, time, data, and resources to implement. However, the selected model will need to include sufficient
description of processes to preserve sensitivity to evaluate management techniques or alternatives and the effects on
the relevant performance measures. In other words, if the chosen TMDL model is not capable of quantifying the
potential response of the selected endpoints to changes in source loads, that model is inadequate to the task.
16
-------
The complexity of a model application can also be adjusted by configuring a specific modeling system. Typically,
detailed modeling systems include the flexibility to select processes and define the appropriate spatial and temporal
detail. For example, the Hydrological Simulation Program—FORTRAN (HSPF) provides the ability to use detailed
descriptions of land runoff and erosion processes. But HSPF can also be applied without including explicit erosion
processes. Simplified models are more limited by their original formulation and defining assumptions. For
example, the GWLF model does not include simulation of in-stream processes, and no alternative formulations are
offered. The user of a model makes numerous choices during model setup that define the complexity of a particular
model application. The user defines spatial and temporal resolution, the specific processes simulated, and the level
of detail of the analysis. For example, based on user setup selections, a model such as Stormwater Management
Model (SWMM) can include pollutant simulation using a description of buildup and washoff of dust and dirt or a
simpler approach, where a fixed concentration (e.g., event mean concentration [EMC]) is assigned to runoff. When
applying the model to a specific watershed or waterbody, the user will also determine the spatial resolution of the
model—how many land use categories, subwatersheds, and waterbody features to simulate. These decisions allow
the user to change the level of complexity from a simplified approach to a more detailed analysis.
Integration of management practices will also affect the development of models for TMDL development and
watershed assessments. Point sources that are discrete discharges, located at well-defined discharge points, typically
will need to be represented individually to determine individual WLAs. Other management practices for wet-
weather point sources, including combined sewer overflows, sanitary system overflows, and stormwater, and
nonpoint sources can be represented in varying levels of detail depending on the pollutant type, watershed and
waterbody conditions, and the level of detail of the management planning analysis. In most states, an
implementation plan is not a required element of a TMDL, and detailed description of BMPs is not required.
However, increasingly, TMDL practitioners are using models to demonstrate management techniques that can be
used to achieve the needed load reductions. The 319 program guidelines also identify the need to evaluate load
reductions and identify management measures needed to achieve the load reductions.
The spatial detail required for simulation of BMPs, especially stormwater and nonpoint source management
techniques, place particular challenges on the development of practical and cost-effective model applications. Most
applications use simplified estimates of BMP adoption and benefit to evaluate the potential for load reduction. Land
use-based management might be represented by a general representation of a reduction in loading. For example, a
change in crop practice could be estimated by percentage reduction in cropland loading expressed as a percentage of
the total load. In detailed simulations, individual BMPs can be applied and their effects on water quality simulated.
For example, in an urban watershed, specific stormwater management ponds can be simulated as a hydrologic unit
and the trapping of runoff and pollutants simulated for each pond. Although simulation of individual BMPs can be
achieved using existing modeling systems, the effort for data collection and modeling for watershed-wide
applications is often too high. The selection and placement of the specific BMPs are typically addressed later, as
part of a watershed implementation planning study. Often, implementation planning includes less rigorous modeling
and focuses more on technical and budget-related specifics. In some studies, a "nested" model development process
is used, where a small-scale, detailed evaluation of BMP performance is used as a basis for extrapolation to a larger
watershed. Some studies use small-scale monitoring studies and literature values of BMP effectiveness to support
watershed-wide estimates.
Integrated Modeling Systems and Linked Models
For watersheds with multiple land and water features, such as a land areas, rivers, canals, reservoirs, and estuaries,
more than one model is often needed. In TMDL studies, representation of sources and receiving waters often
requires two or more models. Modelers often connect or "link" models together to describe an entire system. The
use of multiple models is necessary when multiple features of the system cannot be sufficiently described by one
model. These linkages between models (either available modeling software systems or customized systems) can be
static or dynamic. A static linkage takes output from one model and uses it as input to a second model. A dynamic
linkage can be bi-directional, where information from each timestep transfers back and forth between the models
and affects both simulations. Modelers often implement linkages through a simple file transfer system or a common
database. Some models or modeling software systems provide software-enabled linkages so that all file exchange
requirements are automatically performed as the models are applied.
17
-------
Integrated modeling systems may also provide software that facilitates data exchange; uses common spatial and
point data formats; and prepares input files. Often, shared tools support data management, Web-based data
downloads, model setup, and post-processing. Some modeling systems are based on independent models with an
open set of supporting tools. Other systems provide a unified system with a single interface that launches and
manages several models concurrently.
Trends in Model Development
Models are needed to address the new questions that watershed managers ask that reflect the 21st century trends in
policy and environmental decision-making. Many models are applied as part of larger scale watershed management
studies that address multiple objectives, including TMDLs. Questions that models might be used to evaluate include:
• The implications of long-term changes in land use
• Competition among dischargers for limited assimilative capacity
• Conjunctive use of water resources for water supply, recreation, and aquatic life
• Management of harbors and shipping channels to maintain navigation and aquatic life support
• Planning for effective, targeted implementation of TMDLs
• Addressing multiple concurrent programs such as NPDES, TMDL, Endangered Species, Wetlands, and
Source Water Protection
• Cost effectiveness of management alternatives
• Optimization techniques to help select alternatives that minimize cost and maximize benefit
• Implication of global climate changes on long-term changes in water quantity and quality
New technical challenges will be placed on modeling to support environmental decisions and emerging
programmatic needs. This review will ultimately focus on identifying those specific areas where technical
development is needed to support TMDL development and related programs. However, an initial review of existing
trends in model use can help to develop a preliminary list of key technical needs.
• Less-familiar pollutant types will need to be analyzed. Many studies, especially for TMDLs and watershed
and estuary restoration, have been performed to assess nutrients, dissolved oxygen, sediment, pathogens,
and metals. But many more pollutant types will need to be addressed, including chloride, Dichloro-
Diphenyl-Trichloroethane (DDT), polychlorinated biphenyls (PCBs), and mercury. Existing models and
available supporting data are limited in the range of chemical processes and pollutants they can represent.
• The range of source types will need to be expanded. Typical modeling applications have focused on
dominant, general source categories, such as "agriculture," "urban," "forest." New studies will likely need
to address more specific source types or source loading characteristics, such as agricultural specialty crops
(e.g., strawberries), golf courses, or ski areas.
• Improved techniques are needed to address complex hydrologic conditions, such as high water tables with
surface-to-groundwater interactions, pumped and managed systems, dams, decreasing baseflow due to
groundwater pumping, or complex geology (e.g., karst).
• Competition for resources, management of airborne sources (i.e., mercury, nitrogen), and watershed-based
management needs will increasingly require cross-media analyses (air-land-surface water-groundwater).
18
-------
• Contaminated sediment problems and the need to examine restoration alternatives and dredging
implications will require more detailed simulation of sediment transport and sediment chemistry.
• Evaluation of designated use support in waters of the United States and watershed management
implications will require further analysis of aquatic life and terrestrial habitat (i.e., ecological models).
• An expanded focus on cost-effective implementation will drive technical development in modeling systems
to include simulation of management practices, consider management cost, and include optimization
techniques.
• Global climate change and rapid urbanization will require modeling of future conditions under changing
land use and meteorological conditions, requiring more sophisticated land use and meteorologic projection
techniques.
• As models are increasingly used to support decisions that result in significant financial investment for
restoration and infrastructure, the defensibility and accuracy of models will be challenged. Techniques will
be needed to support model testing, calibration support, verification, and uncertainty analysis.
19
-------
20
-------
Chapter 4 Available Models
This review initially focuses on models that are available today for simulation of watershed and receiving water
conditions. The review emphasizes public domain models, although some selected private or proprietary models are
included if they are published, provide significant technical benefit, or are typically used in TMDL development.
Preparing the review categories considered prior model reviews including Kalin and Hantush (2003) and USEPA
(1997) and online systems such as EPA's Council on Regulatory Environmental Modeling (CREM) Knowledge
Base structure (http://cfpub.epa.gov/crem/knowledge base/knowbase.cfm).
Most models focus on particular land-water features; some are dominantly receiving water models, while others are
primarily oriented to calculating watershed loading. Both receiving water and watershed models can incorporate the
ability to simulate management techniques. For this review, two major categories of models are recognized and
used for evaluation and comparison:
• Receiving water models (Hydrology, Water Quality). This group of models emphasizes description of
hydrology and water quality of water conveyance systems, including rivers, canals, reservoirs, lakes and
estuaries. Some include bi-directional flow, pumps, and operations in freshwater systems. Others include
evaluation of tidal systems and the influences of wind, waves, and tides on mixing. Water quality
simulation involves representation of sediment and pollutant transport and transformation. Some models
include ecological processes, such as vegetative growth, aquatic organisms and aquatic productivity. Not
all receiving water models address water quality. Sometimes, water quality functions are provided by
linking hydrologic and water quality models.
• Watershed models. This group of models emphasizes description of watershed hydrology and water
quality, including runoff, erosion, and washoff of sediment and pollutants. Some models include surface-
groundwater interactions and simplified groundwater transport. Some also include internally linked river
transport and water quality processes and reservoirs.
Table 4-1 provides a summary of currently available models included in this review with contact information and
support for watershed, receiving water, and other key features. Some models simulate BMP performance and
treatment capabilities. Models continue to be expanded to address multiple categories of analysis, such as watershed
models that include BMP analysis (e.g., SWMM), or receiving water models that include hydrology and water
quality (e.g., Environmental Fluid Dynamics Code [EFDC]). Some models are identified as "system," to recognize
that these systems support multiple models (e.g., the EPA TMDL Modeling Toolbox includes linkages between
watershed models and receiving water models). Integrated systems are included in the list of models and include the
multiple capabilities of their component models. This review uses the model categories for descriptive purposes but
recognizes that available models may support both watershed and receiving water simulations.
21
-------
Table 4-1. Summary of Available Models
Model Acronym
AGNPS
AGWA
Ann AGNPS
AQUATOX
BAblNb
CAEDYM
CCHE1D
CE-QUAL-ICM/
TOXI
CE-QUAL-R1
CE-QUAL-RIV1
CE-QUAL-W2
CH3D-IMS
CH3D-SED
DELFT3D
DIAS/IDLMAS
DRAINMOD
DWSM
ECOMSED
Full Model Name
Agricultural Nonpoint Source
Pollution Model
Automated Geospatial Watershed
Assessment
Annualized Agricultural Nonpoint
Source Pollution Model
—
Better Assessment Science
Integrating Point and Nonpoint
Sources
Computational Aquatic Ecosystem
Dynamics Model
—
—
—
—
—
Curvilinear-grid Hydrodynamics 3D —
Integrated Modeling System
Curvilinear Hydrodynamics 3D —
Sediment Transport
—
Dynamic Information Architecture
System/Integrated Dynamic
Landscape Analysis and Modeling
System
—
Dynamic Watershed Simulation
Model
Estuary and Coastal Ocean Model
with Sediment Transport
0) W 0)
|1 |£
»S »! S 73
= >< = a •= ,-8
5 ^ 5 tfl E u
1 S | S % Q. | S
Source 8 "§, 8 * 13 | g, T5
a: x a: S 3 oa w w
USDA-ARS — — • • — —
USDA-ARS — • • • • —
USDA-ARS — — • • — —
EPA — • — — — —
University of Western Australia • • — — — —
University of Mississippi • • — — — —
USACE — • — — — —
USACE • • — — — —
USACE • • — — — —
USACE • • — — — —
University of Florida, Department • • — — — —
of Civil and Coastal Engineering
USACE • • — — — —
WL | Delft Hydraulics • • — — — —
Argonne National Laboratory — — • • • —
North Carolina State University — — • • —
Illinois State Water Survey — • • • — —
HydroQual, Inc. • • — — — —
Process-based
*
*
*
*
*
•
*
•
•
•
*
*
•
*
*
*
EFDC
Environmental Fluid Dynamics Code EPA and Tetra Tech, Inc.
22
-------
Model Acronym
EPIC
GISPLM
GLEAMS
GLLVHT
GSSHA
GWLF
HEC-6
HEC-6T
HEC-HMS
HEC-RAS
HSCTM-2D
HSPF
KINEROS2
LSPC
MCM
Mercury Loading
Model
MIKE 11
MIKE 21
MIKE SHE1
Full Model Name
Erosion Productivity Impact
Calculator
CIS-Based Phosphorus Loading
Model
Groundwater Loading Effects of
Agricultural Management Systems
Generalized, Longitudinal-Lateral-
Vertical Hydrodynamic and Transport
Gridded Surface Subsurface
Hydrologic Analysis
Generalized Watershed Loading
Functions
Scour and Deposition in Rivers and
Reservoirs
Sedimentation in Stream Networks
Hydraulic Engineering Center
Hydrologic Modeling System
Hydrologic Engineering Center River
Analysis System
Hydrodynamic, Sediment, and
Contaminant Transport Model
Hydrologic Simulation Program —
FORTRAN
Kinematic Runoff and Erosion Model,
v2
Loading Simulation Program in C++
Mercury Cycling Model
Watershed Characterization
System — Mercury Loading Model
—
—
—
Source
Texas A&M University — Texas
Agricultural Experiment Station
College of Charleston, Stone
Environmental, and Dr. William
Walker (for Vermont DEC)
USDA-ARS
J.E. Edinger Associates, Inc.
USACE
Cornell University
USACE
USACE
USACE
USACE
EPA
EPA
USDA-ARS
EPA and Tetra Tech, Inc.
Tetra Tech, Inc
EPA
Danish Hydraulic Institute
Danish Hydraulic Institute
Danish Hydraulic Institute
i | i >, "8
5^5 = <8
'> "5 '> 2 S2 E s 55 ct
_ _ • _ _ •
_ __•__•
• •____•
_ _••__•
_ _•___•
• •____•
• • •
_ _•___•
• _____•
• •____•
_ •••__•
_ _••__•
- • • • •
• • •
_ _•___•
• • •
• • •
• _•••_•
23
-------
Model Acronym
MINTEQA2
MUSIC
P8-UCM
PCSWMM
PGC-BMP
QUAL2E
QUAL2K
REMM
RMA-1 1
SED2D
SED3D
SHETRAN
SLAMM
SPARROW
STORM
SWAT
SWMM
Toolbox
Full Model Name
Metal Speciation Equilibrium Model
for Surface and Ground Water
Model for Urban Stormwater
Improvement Conceptualization
Program for Predicting Polluting
Particle Passage through Pits,
Puddles, and Ponds — Urban
Catchment Model
Stormwater Management Model
Prince George's County Best
Management Practice Module
Enhanced Stream Water Quality
Model
—
Riparian Ecosystem Management
Model
—
—
Three-Dimensional Numerical Model
of Hydrodynamics and Sediment
Transport in Lakes and Estuaries
—
Source Loading and Management
Model
SPAtially Referenced Regression On
Watershed Attributes
Storage, Treatment, Overflow, Runoff
Model
Soil and Water Assessment Tool
Storm Water Management Model
TMDL Modeling Toolbox
Source
EPA
Monash University, Cooperative
Research Center for Catchment
Hydrology
Dr. William Walker
Computational Hydraulics Int.
Prince George's County, MD
EPA
Dr. Steven Chapra, EPA TMDL
Toolbox
USDA-ARS
Resource Modelling Associates
USACE
EPA
University of Newcastle (UK)
University of Alabama
uses
USACE (Mainframe version),
Dodson & Associates, Inc. (PC
version)
USDA-ARS
EPA
EPA
0) tf) 0) ._
13 •:= 13 >> a>
5 E 5 =
>33>7i 55 ct
_ •____•
- • • • •
_ __•__•
_ •____•
_ •____•
_ __•__•
• • •
• • •
• • • •
_ _••_•_
_ _•__•_
- • • • •
- • • • •
• ••••_•
24
-------
Model Acronym
Full Model Name
Source
I
ai in 01
IE is
?! ?§ '
~ J? ~ a •=
>E .£° S2
01 c 01 01 01
op os is
01 >. 01 5 5
o£ i a: 5 5
^ 5
V^l wf w
I S 2
00 OT OT
TOPMODEL —
Lancaster University (UK),
Institute of Environmental and
Natural Sciences
WAMView
Watershed Assessment Model with
an ArcView Interface
Soil and Water Engineering
Technology, Inc. (SWET) and
EPA
WARMF Watershed Analysis Risk
Management Framework
Systech Engineering, Inc.
WASP
WEPP
WinHSPF
WMb
Water Quality Analysis Simulation
Program
Water Erosion Prediction Project
Interactive Windows Interface to
HSPF
Watershed Modeling System
(Version 7.0)
EPA 92 •____•
USDA-ARS — _••__•
EPA — •••__•
Systems, Inc.
XP-SWMM Stormwater and Wastewater
Management Model
XP Software, Inc.
When MIKE SHE is fully linked to MIKE 11, it can be characterized as a system and is able to simulate receiving water
hydrodynamics
2 Only when WASP is used together with DYNHYD-a hydrodynamic program for WASP
USDA-ARS = U.S. Department of Agriculture, Agricultural Research Service
USAGE = U.S. Army Corps of Engineers
USGS = U.S. Geological Survey
Receiving water model simulation capabilities are examined in greater detail in Table 4-2. The type, complexity and
water quality simulation capabilities are identified for each model. Model type is categorized as follows:
• Steady State. These models operate under a single nonvariable flow condition. Steady state models are
typically used to evaluate a design flow.
• Quasi-dynamic. Quasi-dynamic models allow for limited variation, typically a variation in meteorologic
conditions over the course of day, to examine variability.
• Dynamic. These models allow for variations in both flow and meteorologic conditions on a small
timestep, typically shorter than daily.
Level of complexity in receiving water models is also evaluated based on spatial detail described as one, two or
three dimensions. Most three-dimensional models also have the ability to be applied in one- or two-dimensional
modes.
Descriptions of water quality capabilities are based on support for specific pollutants or parameters.
Watershed model capabilities are reviewed in Table 4-3. For watershed models, the evaluation is based on five
separate factors: type, complexity, timestep, hydrology, and water quality. Types of watershed models are generally
25
-------
classified as landscape only, simulating only land-based processes, and comprehensive models, including land and
conveyance systems (e.g., rivers, pipes). Complexity in watershed models is classified on three levels:
• Export functions are simplified rates that estimate loading based on a very limited set of factors (e.g., land
use).
• Loading functions are empirically based estimates of load based on generalized meteorologic factors (e.g.,
precipitation, temperature).
• Physically based include more physically based representations of runoff, pollutant accumulation and
washoff, and sediment detachment and transport. Most detailed models use a mixture of empirical and
physically based algorithms.
Timestep, a defining characteristic of models, is often a factor in the comparison of model results to evaluate
management alternatives. If a model is limited to simulation of individual events, it is noted next to the model
name. If not specified, the model is capable of continuous simulation that provides an ongoing account of flow and
load. The table identifies the smallest timestep supported by a model (e.g., hourly, daily). If larger output timesteps
are needed, model output can be summarized from smaller timesteps. Hydrology evaluation criteria consider
whether a model includes surface runoff only, or if surface and groundwater inputs are considered. Most
comprehensive watershed models include a groundwater factor to account for baseflow contributions to streams and
rivers. Finally, water quality capabilities are evaluated based on the pollutants or parameters simulated by the
model. Depending on the level of complexity, the model may use various techniques to simulate the behavior of the
individual pollutants.
Management practice simulation capabilities of watershed and receiving water models are evaluated in more detail
in Table 4-4. Some models specialize in the representation of agricultural areas and include capabilities to evaluate
various crops, crop rotations, tillage practices, and impoundments. Other models are primarily oriented to urban
areas and typically include the ability to evaluate various structural solutions, such as detention ponds. The
simulation of management practices is evaluated based on the scale, complexity, hydrology, and water quality. The
final category identified the specific BMP types considered by the model. For BMP modeling assessment, model
scale is defined as field, BMP, or generalized, as follows:
• Field practices refers to models that assess land use management for one or more single uniform land
drainage area. These models are typically used in agricultural applications to examine crop rotation,
tillage, or nutrient management practices on a small scale or as part of a larger watershed modeling
simulation.
• BMP refers to models that can assess one or more individual BMPs and their influence on hydrology or
water quality loading.
• Generalized identifies models that include a technique to estimate the effect of management as a gross or
larger-scale effects, typically through the use of percentage reductions at the land use or subwatershed
scale.
Hydrology evaluation describes the hydrologic processes of storage, overflow, infiltration and routing that typically
describe BMPs. These hydrologic processes are fundamental to a more physically based description of the BMPs
and are a basis for evaluation of related pollutant removal techniques. Water quality evaluation criteria consider the
support for various pollutants. "Types of BMPs" provides a listing of the support for the major or most commonly
encountered practices. "Vegetative practices" refers to BMPs that use vegetation as part of a system to slow runoff
and help stormwater infiltrate the soil and settle particulates. Examples of vegetative practices include stream buffer
zones, disturbed area stabilization (with mulch, sod, permanent vegetation, temporary vegetation), filter strips, and
grass swales.
26
-------
Each model included in the review is also described in a longer fact sheet (Appendix). The fact sheet includes a
narrative discussion of essential features of each model and provides a comprehensive evaluation of the individual
model software, tools, and supporting features. Each of the identified models was evaluated on key technical,
practical, and software related capabilities. The evaluation format for the fact sheet is structured to support future
use in a database format and facilitate comparison of models. The structure of the model fact sheets and definitions
for each category are shown in Table 4-5.
Table 4-2. Summary of Receiving Water Simulation Capabilities
Type
Level of Complexity
Water Quality
Model
ACJUAI UX
O
1 1
>. •"?
"5 »
8 §
55 O
o
ro
ro
c
O
in
ti
O -c
(vert)
ro
c
O
in
°l
— -a
ro 01
c c
0 s=
, '1 •§
£ I at
£ .= m
\--a 3
Sedime
in
Nutrient
Toxics
in
ro
13
Q
O
00
"8
Dissolv
oxygen
Bacteric
BASINS
CAEDYM
CCHE1D
CE-QUAL-ICM/TOXI — —
CE-QUAL-R1
CE-QUAL-RIV1
CE-QUAL-W2
CH3D-IMS
CH3D-SED
DELFT3D
DWSM
ECOMSED
EFDC
GISPLM
GLLVHT
GSSHA
HEC-6
HEC-6T
HEC-RAS
27
-------
Type Level of Complexity Water Quality
o
Si I ra ra ra
i & o § § §
>, "? £ « c ' C
Model | 1 | g| o| ||
53 o Q o 5 H 5 H 5
HSCTM-2D — — • — • —
MCM — — • • — —
MIKE 11 • — • — • —
MINTEQA2 • — — — — —
QUAL2E — • — • — —
QUAL2K — • — • — —
RMA-11 — — • • • •
SED2D — — • — • —
SED3D — — • • • •
SHETRAN — — • • — —
_ _ _ _ _
Toolbox — • • • • •
W AM View — — • • — —
WASP — — • • • •
WinHSPF — — • • — —
WMS — — • • • —
XP-SWMM — — • • — —
BOD = Biochemical oxygen demand
il!s, I* i
m "~ "^ "~ " O t/) ^) ^t
(/) Q) 3 Q fl) O .;*• s> (5
Z) CO Z I— ^ CO Q O DQ
________
(Hg)
— — — — — — — —
--------
• _•__•••
• _•__•••
• ••-_••_
________
________
________
_••__•••
-
__•
28
-------
Table 4-3. Summary of Watershed Simulation Capabilities
Type
Model
AGNPS (event
based)
tt
V)
E
2
5
•
g included
_c
Stream rou
•
Level of
Complexity
in
'c
01
O
E
g
O
O
Q.
X
HI
—
in
O
5
Loading fu
—
E
2
Physically
•
Timestep
>,
! * f I
3
-------
Type
Model
PCSWMM
in
g
Tl
5
ncluded
O)
_c
Stream rou
•
Level of
Complexity
in
o
'o
E
%
o
o
Q.
X
HI
in
O
1
a
Ol
c
^
3
•
in
s
Physically
•
Timestep
_>,
'i5
1 t
OT Q
•
|
Annual
Hydrology
Surface
jndwater
8
O)
•c
Surface an
•
•c
User-defin
Sediment
Water Quality
Nutrients
in
01
•c
|o
Toxics/pes
U)
ro
'3
Q
O
00
Bacteria
•
PGC-BMP
SHETRAN
SLAMM
SPARROW
STORM
SWAT
SWMM
Toolbox
TOPMODEL
W AM View
WARMF
WEPP
WinHSPF
WMS
XP-SWMM
30
-------
Table 4-4. Summary of Management Practice Simulation Capabilities
HSPF
Type Level of Hydrology Water Quality Types of BMPs
Complexity
in w
s s
= ti * 8
a! '« S 2 =
£ -a "8" .3 o- a. -g
+t Q) ^ C L (/) ^ fl) -5
I i!&3i?i^ls2H4iii£«
Model 2Q. | ra 2'gs:§£j§l'sjj]5.s'§|'S£S
.3*^ a) ^32>i^O(/)SIfi
i^ 00 CD Q fftO Cfy*^^n ^n ^nno ^^gQ
AGNPS •• • ___••_••••__•_•__
AnnAGNPS •• • —— — •• — •••• — — • — • — —
AQUATOX — • — _____•_••___•____
BASINS ••• •_____••••__•_•_•
DIAS/IDLMAS — • • ______•__________
DRAINMOD •__ • •••____•______•_
DWSM _•_ •__••_••••__••___
EPIC •__ • •••__••••__••_•_
GISPLM — • • _______•_________
GLEAMS ••_ •_••__••••_______
GSSHA ••_ •__••_•_____•• ••
GWLF • — • ______•••_____•__
KINEROS2
LSPC
MUSIC — • — •••••• — — — — — — •••••
P8-UCM _•_ ••••_••••_•_•••_•
PCSWMM _•_ •••••••_•___••__•
PGC-BMP _•_ ••••••______•••••
REMM — • — •— — •• — •••• — — — — • — —
SLAMM — • • • •••__•••_•_•••••
STORM — • — — ••_____________•
31
-------
Model
SWAT
SWMM
WARMF
WEPP
WinHSPF
XP-SWMM
Type Level of Hydrology
Complexity
d)
g -a 13
I 1 ! & 1 1 ? 1 1
11 1 i 1 s 1 1 s l
U.OQO Q CO O .£ £ 3 CO
• •— •__••_•
— •—•••••••
_•_ •_____•
• • • • • •
• • • _____••
— •—•••••••
Water Quality
c
'in
Ig! 5 1
S- o o J2 § 'c
S 2 i is I I
£ s; 01 Ji ra 01
Q. Z Q. S 00 Q
• ••--•
— • — — — •
• • — •• —
• •••• —
— • — — — •
Type
Infiltration practices
*
•
—
9
—
•
s of BMPs
Vegetative practices
Wetlands
* -
— —
— •
— —
Other structures
•
—
•
—
•
32
-------
Table 4-5. Overview of Models - Review Categories
Categories
Description
Contact Information
Includes:
Contact name
Affiliation
Address
Phone number
E-mail
Web site
Download Information
Includes location, contact, availability, and cost, if applicable, for downloading the model and
related files on the Internet
Model Overview/Abstract
Provides a general summary of model purpose and capabilities
Model Features
Identifies the key model features characterizing the model type and simulation capabilities
(e.g., lumped, nonpoint source)
Model Areas Supported
Grades the model's support for the following key features:
• Watershed
• Receiving water
• Ecological
• Air
• Groundwater
Each feature is graded as follows:
• High—fully supported/physically based
• Medium—some simplifying assumptions
• Low—empirical representation
• None—no support
Model Capabilities
Includes the following subcategories:
Conceptual basis—summarizes of the general formulation of the model
Scientific detail—discusses specific technical components and solution techniques used in
the model
Model framework—discusses the structure of the model
Scale
Assumptions
Model Strengths
Model Limitations
Application History
Includes the following subcategories:
Spatial scale — identifies the smallest operational unit of the model (e.g., cell, watershed,
size range)
Temporal scale— identifies the model's timestep
Lists key operational or technical assumptions included in the model
Lists key strengths (technical, operational, or systems/software-related) of the model
Lists key limitations (technical, operational, or systems/software-related) of the model
Identifies past applications that demonstrate model utility and applicability and identifies
documentation of model application
Model Evaluation
Summarizes any available formal testing of model, peer review, or other supporting
documentation
33
-------
Categories
Description
Model Inputs
Listing of key inputs and parameters (or categories of parameters) needed for model setup
and application
Users' Guide
Identifies the availability of a user's manual and where to obtain it
Technical Hardware/Software
Requirements
Listing of key computer and operational related information related to:
• Computer hardware
• Operating system
• Programming language (code and interface)
Linkages Supported
Lists related or linked models or modeling systems
Related Systems
Lists other or alternate interfaces for the model
Sensitivity/Uncertainty/Calibration Identifies supporting tools or measures of sensitivity/uncertainty and calibration components
Model Interface Capabilities
Lists model interface tools (e.g., pre- and post-processors, data display tools, data
preparation tools)
References
Lists key references (selected recent or relevant articles)
34
-------
Chapter 5 Applicability of Models
Although a variety of models is available, the assessment of modeling needs must also consider the applicability of
models for TMDLs and watershed programs. In this chapter, the applicability of models is evaluated by matching
model capabilities to a set of application criteria. Applicability considers the defining characteristics of project
applications including pollutants, land and water characteristics, management alternatives, data, and user interfaces.
The coverage provided by available models is discussed and key gaps identified. The ability of existing integrated
modeling systems to address the application needs, through linkage of models and supporting utilities, is also
discussed.
Application Criteria
The application criteria were designed to evaluate the capabilities of currently available models to support TMDL
development and evaluation of implementation options. The application criteria match specific TMDL modeling
needs with the modeling capabilities and processes described, the land and water features simulated, and the utilities
that support model application. The application criteria were used to evaluate the capability of models to perform
TMDL analyses, including training, level of effort, and user interface capabilities. In the previous chapter, the
models were evaluated on their basic capabilities in watershed, receiving water, and BMP simulation. For this
chapter, all available models were evaluated in each table, recognizing that each model could support multiple
criteria. Integrated modeling systems (e.g., BASINS, Toolbox, and WMS) were evaluated on the capabilities of
their component models.
A structured sequence of application criteria was used to
evaluate capabilities for each of five major categories. The
first application category, TMDL endpoints, considers the
ability of models to predict the magnitude, frequency and
duration of the typical endpoints (Table 5-1). Prediction of
endpoints is essential for evaluating loading capacity in
TMDLs and watershed simulation modeling. For example,
a wide range of models—simple models that provide only
annual loads or complex models that perform subhourly
simulation—can evaluate annual phosphorus loading.
Evaluation of a dissolved oxygen endpoint might require a
model to evaluate hourly dissolved oxygen fluctuations.
The second application category was designed to identify
the capabilities of models to address specific land and
waterbody characteristics (Table 5-2). Some models are
designed to address one or more waterbody types (e.g.,
lakes and rivers) while others are limited to a single
waterbody type. The purpose of this category is to develop
an inventory that characterizes the capabilities and scope of
the available models. In practical application, as noted in
Chapter 2, one or more models may be used in
combination to address multiple land and water features
present in an individual watershed.
Model Application Tables
TMDL Endpoints (Table 5-1). Considers the model's ability
to simulate typical TMDL target pollutants and expressions
(e.g., load vs. concentration). Characterizes the models
depending on the timestep of the simulation for the target—
steady state, storm event, annual, daily or hourly.
General Land and Water Features (Table 5-2). Rates
models according to their ability to simulate general land
uses and waterbody types.
Special Land Processes (Table 5-3). Rates models on
their ability to simulate special land processes such as
wetlands, hydrologic modification, urban BMPs and rural
BMPs.
Special Water Processes (Table 5-4). Rates models on
their ability to simulate special processes occurring in
receiving waterbodies such as air deposition, stream bank
erosion, algae and fish.
Application Considerations (Table 5-5). Rates models on
the following practical considerations affecting their
application—experience required, time needed for
application, data needs, support available, software tools and
cost.
A uniform scoring system is defined in the "key" below each
table.
Special application categories were also identified to highlight the land and water processes that are sometimes, but
not always, needed in models (Table 5-3 and 5-4). The purpose of separating these application categories from the
general categories is to highlight the differences between models and identify those that have incorporated
35
-------
specialized physically based algorithms that might be needed for specific applications. With these categories
separated, the emerging capabilities are more clearly discernable.
The last application category examines the model interface and application considerations, including data needs,
user interfaces, and availability of code (Table 5-5). These functional descriptions help evaluate model "usability"
based on the data requirements and software systems. The criteria are generalized for each model, although
complexity of a specific model application will vary depending on the number of endpoints, land uses, and processes
simulated. This table evaluates models on their typical application complexity. In some cases, highly detailed
models can also be applied very simply and cost-effectively by experienced users. For example, HSPF could be
applied very simply and quickly to a small, homogeneous watershed. Simple models, however, have very little
variation in the level of complexity. The design of these criteria recognizes that TMDLs are often highly
constrained in data availability, application resources, and schedule. The considerations in model selection and
application show a benefit in many cases for models that have low data requirements and are easy to apply. The
criteria scores recognize these considerations by showing "solid dot" or high value for low data needs and "dashes"
or low value for high data needs. Scores for user interfaces show solid dots for good software support.
Following Tables 5-1 through 5-5 is a discussion of the capabilities and limitations of the models based on the
model review.
36
-------
Table 5-1. TMDL Endpoints Supported
0000— — — O— — — 000— — 00— — — ——— —
AGNPS
AnnAGNPS
AQUATOX
BASINS
CAEDYM
CCHE1D
CE-QUAL-
ICM/TOXI
CE-QUAL-R1
CE-QUAL-RIV1
CE-QUAL-W2 •••••••••••••— — — — — — • •
CH3D-IMS
CH3D-SED
DELFT3D
DIAS/IDLMAS
DRAINMOD — — — — • •
DWSM
ECOMSED
EFDC
EPIC
GISPLM
GLLVHT
GSSHA
37
-------
HEC-6
HEC-6T
HEC-HMS
HEC-RAS
HSCTM-2D
HSPF
KINEROS2 ___________ « « «
LSPC •••••••— — —•••• — •——• • •— •— •
MCM ___________________ __e — e —
Mercury Loading a
Model
MIKE 11
MIKE 21 •••••••••••••• — •— — — • •— •• •
MIKE SHE ________________________ _
MINTEQA2 + + + + + + +— — — — — — —++——+ — — — ++ —
MUSIC ,,,,__,____,,___________ _
P8-UCM
PCSWMM
PGC-BMP •••••••— — —•••• — •——• • •— •— •
QUAL2E + + + + + + + + +— — — — — — — — — — + + — — — —
QUAL2K + + + + + + + + + +— — — — — — — — — + + — — — —
REMM eeeeeee— — — eeee— — — — — — — ——— —
38
-------
DMA 1 1
SFD2D
c Fmn
QUFTPAM
CI AIV/IIV/I
o
c c "§
•^ ^ "c
TO m a)
i ^ o
c go
I I I | I
Amm onia concentration
TN : TP mass ratio
DJ ssolved oxygen
Chlorophyll a
Algal density (mg/m2)
M et TSS load
TSS concentration
Sed iment concentration
Se diment load
Sulfate concentration
IVi etals concentrations
p esticides concentrations
Herbi cides concentrations
Toxics concentrations1
Pathogen count
(e.g., fecal colifonn)
Temperature
Methylmercury tissue
concentration
IVI etals sediment concentration
Mercury sediment concentration
Synt netic organic chemicals
sediment concentration
SPARROW 0000— — O— — — — — — O— — 00— — — — — — —
STORM — — 00— — — — — — — 000— — — — — O — — — — —
SWAT eeeeeeeee— — eee — eee— e
SWMM
Toolbox
TOPMODEL
WAMView
WASP
WARMF eeeeeeeee — eeeeeeee— e e— ee —
WEPP
WinHSPF
WMS
XP-SWMM
1 Entries under this category indicate that models have the capacity to simulate user-defined toxic chemicals
2 GWLF calculations are performed on a daily basis, but results are presented on a monthly basis.
Key:
— Not supported
+ Steady State
o Storm
<3 Annual
e Daily
• Hourly (or less)
39
-------
Table 5-2. General Land and Water Features Supported
Reservoir/ Estuary Coastal
Model Urban Rural Agriculture Forest River Lake impoundment (tidal) (tidal/shoreline)
AGNPS —••— — — — — —
AnnAGNPS —••— — — — — —
AQUATOX —— — —00 O — —
BASINS 9 • • • • 9 9 — —
CAEDYM —— — —•• • • •
CCHE1D o O O O • — — — —
CE-QUAL-ICM/TOXI —— — —•• • • •
CE-QUAL-R1 —— — — — o O — —
CE-QUAL-RIV1 ______ _ _ _
CE-QUAL-W2 —— — —•• • • •
CH3D-IMS —— — —•• • • •
CH3D-SED —— — —•• • • •
DELFT3D —— — —•• • • •
DIAS/IDLMAS — (t (t • — — — — —
DRAINMOD —••• — — • — —
DWSM
ECOMSED
EFDC
EPIC
GISPLM
GLLVHT
GSSHA
GWLF
HEC-6
HEC-6T
HEC-HMS
40
-------
Reservoir/ Estuary Coastal
Model Urban Rural Agriculture Forest River Lake impoundment (tidal) (tidal/shoreline)
HEC-RAS ______ _ _ _
HSCTM-2D ______ _ « _
HSPF o • • • • O o — —
KINEROS2 O • • O O — O — —
LSPC o • O • • • • — —
MCM ______ « _ _
Mercury Loading o O O O O O o - -
Model
MIKE 11 ______ « _ _
MIKE 21 —— — —•• O • •
MIKE SHE •••••— • — —
MINTEQA2 ______ _ _ _
MUSIC ______ _ _ _
P8-UCM 00000— 9 — —
PCSWMM
PGC-BMP
• 00000 0 — —
.
QUAL2E ______ _ _ _
QUAL2K
REMM
RMA-11
SED2D
SED3D
SHETRAN
SLAMM
SPARROW
STORM
SWAT
SWMM
41
-------
Model
Reservoir/ Estuary Coastal
Urban Rural Agriculture Forest River Lake impoundment (tidal) (tidal/shoreline)
Toolbox •••••• • • •
TOPMODEL
WAMView
WARMF
— •••o— — — —
>•••>> 0 — —
• • • • o • • — —
WASP —— — —•• • • •
WEPP —••• — — — — —
WinHSPF
WMS
XP-SWMM
>••••> 0 — —
>••••> 0 — —
.
Key:
- Not supported
o Low-Simplified representation of features, significant limitations
<3 Medium-Moderate level of analysis, some limitations
• High-Detailed simulation of processes associated with land or water feature
42
-------
Table 5-3. Special Land Features Supported
Model
AGNPS
*§ o
_
II I
in o ra
E a.
Urban Land Management
Rural Land Management
3 8
•
SI I
§ a-
. = ,
t3°)50)
=
0 .a
-
e practices
fl I f I I jjlil I II 11
< g J I 00 W 2 t. W S" S O !> £ Z ;§!. — ~ ° i H O. !>
— O O O
AnnAGNPS
o o o o
AQUATOX _____ _______ _ _ _ _ a _
BASINS ooooo oo — ooo — • • — — oo
CAEDYM _____ _______ _ ___,_
CCHE1D _____ _______ _ _____
CE-QUAL-ICM/TOXI o O — — — _______ _ ___,_
CE-QUAL-R1 _____ _______ _ _____
CE-QUAL-RIV1 _____ _______ _ _____
CE-QUAL-W2 _____ _______ _ ___,_
CH3D-IMS o — — — — _______ _ ___,_
CH3D-SED o — — — — _______ _ ___,_
DELFT3D o — — — — _______ _ ___,_
DIAS/IDLMAS ____a _______ _ _____
DRAINMOD _a_a_ _______ « _,,,_
DWSM ____, _______ _ __,aa
ECOMSED _____ _______ _ ___,_
EFDC
• O
43
-------
Model
EPIC
GISPLM
GLLVHT
GSSHA
GWLF
HEC-6
HEC-6T
HEC-HMS
HEC-RAS
HSCTM-2D
HSPF
KINEROS2
LSPC
MCM
Mercury Loading
Model
MIKE 11
MIKE 21
Urban Land Management Rural Land Management
•58 ?
-11 »
"c c <8|).i.'°g <8 ZT § ,2
ill 11 i c f i s jjil! H s
1 1 | . £| « | | | 1 £§ « g 8 |
1 i 1 1 I H5 1 i 1 . 1
li i li i jiiij j H 1 1
< § J3 i OQ SSzsHSSw'Q o > 1 z£_ 2* o t j= £ .g
_ _ _ _ a ___o_o_ o o___o
- , - . 0 ----,-- - 9 - - . -
— — — — — 0 — — — — 0 — 0 0 — — — 0
--,,,----,-- - - ,
09090 —0—0 — 00 • • — — 90
— — — 90 — — — — 0 — — — 9 — — 9 —
— 9 — 90 —0—0 — 00 9 9 — — 9 0
MIKE SHE
— 9 • • O ——0—000 — 00 — — —
44
-------
Model
MINTEQA2
MUSIC
P8-UCM
PCSWMM
PGC-BMP
QUAL2E
QUAL2K
REMM
RMA-1 1
SED2D
SED3D
SHETRAN
SLAMM
SPARROW
STORM
SWAT
SWMM
Toolbox
Urban Land Management
«f «
1 I
all i I
oo_ .".? o i
1 1 I 1 I fi« | 1 I I
« Eoni fliO^oj i =
- » ---.-..
— O — — 9 OOOOOOO
— O — 99 9 O 9 9 O O —
— 9 — • • ____a__
— 9 • • 0 — 0 0—000
— 0 — 00 0000000
— 9 — — — — — —— — — —
— O — — — O O — O O O —
— O — 99 9 O 9 9 O O —
<:>••<:»<:» o c» o c» • c» c»
Rural Land Management
Nutrient control practices
(fertilizer, manure mgmt.)
Vgricultural conservation
iractices (contouring, terracing,
ow cropping)
Irrigation practices
Tile drains
Ponds
- - - - "
- - - - 3
0 0 — — 9
9 - - - .
0 00 — —
— — — — 9
0 0 — — —
• • • • 9
O O — — 9
9 9 9 • •
Vegetative practices
—
—
—
—
—
—
—
»
—
9
45
-------
Urban Land Management Rural Land Management
Model
TOPMODEL
WAMView
WARMF
WASP
WEPP
WinHSPF
WMS
XP-SWMM
Air deposition
Wetland
Land-to-land simulation1
o • •
- » -
000
O 9 O
— o —
Hydrologic modification
»
—
0
»
»
BMP siting/placement
»
—
0
o
»
Street sweeping
o
—
—
—
»
Nutrient control practices
(fertilizer, pet waste mgmt.)
»
—
0
o
o
Stormwater structures (manhole,
splitter)
o
—
—
—
»
Detention/retention ponds
»
—
0
o
»
Constructed wetland processes
•
—
—
—
o
Vegetative practices
»
—
0
o
o
Infiltration practices
»
—
0
o
—
Nutrient control practices
(fertilizer, manure mgmt.)
»
•
•
•
o
Vgricultural conservation
iractices (contouring, terracing,
ow cropping)
Irrigation practices
Tile drains
Ponds
a a . .
- - - •
• — — o
. - - o
0 - - 0
Vegetative practices
"
—
0
o
—
Land-to-land simulation: Model capacity to transfer runoff, sediment, and nutrients from land to land instead of from land to streams
or other receiving water.
Key:
- Not supported
o Low-Simplified representation of features, significant limitations
<3 Medium-Moderate level of analysis, some limitations
• High-Detailed simulation of processes associated with land feature
46
-------
Table 5-4. Special Water Features Supported
o
IE
O 3
•a °-
— VI
Model 2 o
0=2
VI
I
Near-field ana
(mixing zone)
Airdepositior
Surface/
groundwater
interactions
'in
2
MI
Stream bank i
o
Q.
t/)
Sediment trar
Ol
§
r*t
Sediment diac
^
c
Phytoplanktol
floating algae
Periphyton/
macrophytes
"O
Planktonic an
benthic algae
^
iZ
AGNPS
AnnAGNPS
AQUATOX
BASINS
(t (t
CAEDYM
CCHE1D
CE-QUAL-ICM/TOXI _
CE-QUAL-R1
CE-QUAL-RIV1
CE-QUAL-W2
CH3D-IMS
CH3D-SED
DELFT3D
DIAS/IDLMAS
DRAINMOD
DWSM
ECOMSED _ _ _ _
EFDC • • o O
EPIC _ _ _ o-
GISPLM _ _ _ _
GLLVHT _ _ _ _
GSSHA _ _ _ «
GWLF _ _ _ _
HEC-6 _ _ _ _
47
-------
in
O
1
01
Q.
O
E
Model 2
"in
a.
2
Q.
in
I
-field analysi.
ng zone)1
II
eposition
<
ace/
ndwater
actions
W oi.E
c
O
S
Ol
££
1
E
8
•c
ment transpo
V)
in
in
ment diagene
V)
opiankton/
ing algae
•5,13
££
U)
= ^
i!
0.2
ktonic and
hie algae
||
in
[L
HEC-6T
HEC-HMS
HEC-RAS
HSCTM-2D
HSPF
KINEROS2
LSPC
MCM
Mercury Loading
Model
MIKE 11
MIKE 21
MIKE SHE
MINTEQA2
MUSIC
P8-UCM
PCSWMM
PGC-BMP
QUAL2E
QUAL2K
REMM
RMA-11
SED2D
SED3D
SHETRAN
48
-------
Model
SLAMM
!_ I 1
2 in ~m^~ c &
SL| sg a sz =
°= TJ 2 « 5 o S
E°: *£> a ||| ^
1 2 "1 3 I 1 2 |
Bo ra.o . t 2 « S*
tig aip .- =2* i:
Q =. z i. < w o).E w
— — — O —
Sediment transport
—
in
'in
01
1 =
I sl -8 SS
•» _o> c -e o 51
•£ ra « £.= 'E "
1 H tl 11 I
(0 Q- 5= Q- E Q. J2 u.
— — — — —
SPARROW ___________
STORM ___________
SWAT
— — — 9 —
9
— — — — —
SWMM a____a_____
Toolbox
TOPMODEL
WAMView
WARMF
WASP
9 9 O O 9
— — — 9 —
9 — — 9 0
O — 9 9 —
— — O — —
9
—
0
—
—
.
— — — — —
— — — — —
— — — — —
9999 —
WEPP ___________
WinHSPF
WMS
XP-SWMM
0 — 99 —
9
• — O O —
0
9
9
— — — — —
— — — — —
— — — — —
Near-field is the region of a receiving water where the initial jet characteristic of momentum flux, buoyancy flux, and outfall
geometry influence the jet trajectory and mixing of an effluent discharge. This is a specialized feature included only in the most
detailed receiving water models.
Key:
— Not supported
o Low-Simplified representation of features, significant limitations
9 Medium-Moderate level of analysis, some limitations
• High-Detailed simulation of processes associated with water feature
49
-------
Table 5-5. Application Considerations
Model
Experience Time Needed for Support Software
Required Application Data Needs Available Tools
Cost
AGNPS
AnnAGNPS
AQUATOX
BASINS
CAEDYM
CCHE1D
CE-QUAL-ICM/TOXI
CE-QUAL-R1
CE-QUAL-RIV1
CE-QUAL-W2
CH3D-IMS
CH3D-SED
DELFT3D
DIAS/IDLMAS
DRAINMOD
DWSM
ECOMSED
EFDC
EPIC
GISPLM
GLLVHT
GSSHA
GWLF
HEC-6
HEC-6T
HEC-HMS
-
• • • O • •
9 9 0 0 • •
0 0 0 0 • 0
0 0 0 0 • •
50
-------
Model
HEC-RAS
HSCTM-2D
HSPF
KINEROS2
LSPC
MCM
Mercury Loading Model
MIKE 11
MIKE 21
MIKE SHE
MINTEQA2
MUSIC
P8-UCM
PCSWMM
PGC-BMP
QUAL2E
Experience Time Needed for Support Software
Required Application Data Needs Available Tools
9
— — o o o
— — 0 • •
0 0 0 0 •
— o o 9 •
0
• •909
— — 0 • •
— — O • •
— — O • •
9 • 9 0 0
9 9 9 • •
• • • O O
— O O 9 •
9 • O O •
9 • 9 9 0
Cost
•
•
•
•
•
a
•
—
—
—
•
0
•
—
a
•
QUAL2K a « a a a «
REMM
RMA-1 1
SED2D
SED3D
SHETRAN
SLAMM
SPARROW
STORM
SWAT
O 9 O O —
-
— — 0 9 9
-
— — O • •
09999
9 9 • • 9
O • 9 O —
0
•
—
—
•
a
0
•
•
•
SWMM _ o ooo*
51
-------
Experience Time Needed for Support Software
Model Required Application Data Needs Available Tools Cost
Toolbox o o o • • •
TOPMODEL a a • 9 • •
WAMView o — O (t • (t
WARMF o O 000 —
WASP _ _ 000*
WEPP a « « o O •
WinHSPF _ o O • • •
WMS — o O O • —
XP-SWMM _ o O O • —
Key:
Experience: Time Needed for Application: Data Needs:
- Substantial training or modeling expertise required (generally - > 6 months o |_|jqn
requires professional experience with advanced watershed o > 3 mOnths
and/or hydrodynamic and water quality models.) a Medium
o > 1 month _ . _.,,
o Moderate training required (assuming some experience with • LOW
basic watershed and/or water quality models) * < 1 montn
o Limited training required (assuming some familiarity with
basic environmental models)
• Little or no training required
Support Available: Software Tools: Cost:
- None - None - Significant Cost (>$500)
o Low o Low o Nominal Cost (<$500)
o Medium o Medium o Limited Distribution
• High • High • Public Domain
Capabilities and Limitations of Currently Available Models
The currently available models address many of the TMDL needs identified in Tables 5-1 through 5-5, at various
levels of complexity and difficulty. Observations can be drawn from examination of each table to evaluate model
capabilities and limitations for the development of TMDLs and waterbody restoration plans.
• Endpoints supported (Table 5-1). Of the more than 65 models reviewed, most of the models support
some type of analysis of sediment and nutrients. Many models address related measures of eutrophication
such as chlorophyll a and nutrient concentrations. Some models, in particular some of the more detailed
models, support simulation of pathogens. However, significantly fewer models support evaluation of
metals (20 models) and toxics (12 models). Contaminated sediment, herbicides and pesticides are the least
frequently supported endpoints. Only a few highly specialized models support mercury simulation.
• General Land and Water Features. All waterbodies and general land use types are supported by models;
however, most models specialize in one of the following: land/watershed system, freshwater rivers, lakes,
and tidal areas. Many comprehensive watershed models (e.g., HSPF, LSPC, SWMM) include watersheds,
rivers and simplified lakes. Simulation of tidal waters is addressed by specialized models (e.g., EFDC,
CH3D, ECOM) that are limited to receiving water simulation. Development of watershed inputs for
receiving water models is provided by linkage with watershed models or evaluation of monitoring data.
Integrated modeling systems may include linked models that provide support for watersheds, rivers, lakes,
52
-------
and estuaries (e.g., EPA TMDL Toolbox). Support for linkage between watershed and receiving water
models, especially tidal waters, is extremely limited.
• Specialized Land Features. Support for specialized land features is limited to a few models. The most
commonly supported feature is simulation of impoundments such as ponds or reservoirs, often needed in
the evaluation of stormwater management. However, although most receiving water models can simulate
ponds, typically the models are too complex for practical application. More typically applied are simplified
tools included in watershed models to assess stormwater management ponds (e.g., P8-UCM). Atmospheric
deposition is an important consideration in large watershed or estuaries. The application criteria show that
some watershed models assess dry and wet weather deposition but do not always consider a separate
atmospheric deposition term. HPSF is one of the few available models with an explicit function to
assessing atmospheric deposition of nutrients. Practices such as fertilizer and manure application are
included in many agriculturally oriented watershed models (e.g., AnnAGNPS, SWAT). However,
irrigation and tile drainage are less frequently included. Wetlands and BMPs that include constructed
wetlands are also infrequently included in models. Similar to modeling of impoundments, wetland
simulation, although possible using complex receiving water models, is not always practical for general
application. For example, EFDC has the capability to evaluate wetland systems within the context of a
larger simulation of a tidal waterbody. For urban areas, street sweeping is included only as a distinct
practice in traditional urban models such as SWMM, SLAMM, and P8-UCM but not included in many
other watershed models such as HSPF and GWLF.
• Specialized Water Features. Only the most detailed physically based models support multiple specialized
water features. Few models address surface-groundwater interactions in detail. Few hydrodynamic models
can address near-field mixing zone studies. Only rarely do models support calculation of stream bank
erosion. More models address stream sediment transport at varying levels of complexity. Few models
include sediment diagenesis, which can be an essential factor in the internal recycling of nutrients in lakes
and estuaries. This feature is increasingly significant in performing long-term projections of restoration
potential. Some support evaluation of multiple processes related to algal, periphyton, and macrophyte
species. This is an important feature for many eutrophication-related TMDLs because simulation of
aquatic vegetation is needed to determine allowable loading of nutrients. Of the reviewed models,
ecological processes such as fish and food chain simulation are supported only by AQUATOX. Many of
the three-dimensional models used for large-scale estuary applications integrate atmospheric deposition as
a source (e.g., ECOM, EFDC, WASP). However, very few models incorporate irrigation and drainage
processes, except agriculturally oriented systems (e.g., SWAT) or specialized drainage models (e.g.,
DRAINMOD).
• Application Considerations. Detailed models typically require a high level of experience, significant
amounts of data, and time for setup and testing. Most of the models reviewed require experience and
training to apply; application and interpretation of even the simplest models still require some experience in
environmental analysis. The data requirements and time needed for application are typically associated
with complexity. Some models have technical support available (e.g., list servers), while others have no
formal network for support. Some of the propriety models provide technical support as part of the services
included with purchase of the systems (e.g., XPSWMM, MIKE SHE). Levels of support vary and are
typically more limited for research or public domain models. Many models include interfaces and software
tools, such as post-processors, which can help to make application and interpretation of model results more
efficient. Software tools often focus on the typical use of the model. Models that are actively used for
watershed and TMDL development typically include specific software tools for calculating TMDL
allocations (e.g., LSPC, WARMF). Integrated systems provide support for data, software, and analysis,
although the complexity of the systems still requires training and experience for application (e.g., BASINS,
TMDL Toolbox). Few models include tools for evaluation of model accuracy, support for
calibration/validation, or sensitivity analysis. Cost of the models and systems varies from free distribution
of public domain or open source code systems to significant costs (i.e., more than $1,000) for privately
maintained and distributed models (e.g., MIKE SHE, XP-SWMM).
Of the dominant pollutants identified in TMDL listing across the country, many still have significant limitations in
the availability and sensitivity of simulation techniques. Pathogens are simulated by few watershed models, and
53
-------
accuracy of source characterization is limited. Metals are also simulated by few watershed models with the major
limitation being speciation of metals and pH-related processes. Nutrients are addressed relatively well by both
watershed and receiving water models, building on a long history of eutrophication studies in lakes and estuaries.
However, some of the more specific endpoints related to nutrients are not well described and require more
development of ecological models. In addition, river models are less likely to include dynamic simulation of
attached algae and dynamic calculation of the input of dissolved nutrients. Simulation of stream sediment, stream
bank erosion, and channel formation is not supported by most watershed models. However, sediment-related
aquatic life impairment comprises more than 8 percent of TMDL listings and is likely associated with additional
listings for Biological Criteria.
The review of TMDL requirements and comparison with available models demonstrates that, although many of the
technical needs are addressed, the capabilities are distributed among multiple models, techniques are not uniformly
available, and selection of any single model is likely to result in limitations of simulation capabilities. The available
models also reflect the genesis of their development—as agricultural or urban models, tidal modeling systems, or
ecosystem models—that have been only partially adapted to TMDL and restoration plan development. The
specialized land and water features, supported by few of the available models, include many of the technical
considerations that are present in impaired waterbodies throughout the United States. The most promising aspect of
model development for TMDL and restoration planning support is the integrated modeling systems discussed further
in the following section.
Integrated Modeling Systems
In this section, the unique characteristics and system of supporting tools that comprise integrated systems are
discussed further. In the previous sections, integrated modeling systems were reviewed for and categorized by the
comprehensive capabilities of their component models. In this section, the applicability and capabilities of the
integrated modeling systems are described and compared. These systems are evaluated for their capabilities in
providing more comprehensive solutions of the need for simulation and development of management plans for
TMDLs.
Integrated systems are compilations of data support tools and multiple models that provide a workspace or
environment for executing multiple analytical steps. Integrated systems demonstrate a high level of support for
TMDL needs, based on the application criteria examined in Tables 5-1 through 5-5. Integrated systems can satisfy
multiple application criteria because they include the capabilities of multiple models, provide various software tools,
and include data and analysis support. Much of the recent development of integrated systems for TMDL
applications has focused on improving efficiency and consistency of modeling applications. Integrated systems
typically provide linkages between data and models and include a set of tools to quickly and efficiently build, test,
and apply models to support environmental decision-making. The development of integrated systems has generally
focused on functionality and not on the fundamental research and development of new models and physically based
processes.
In 1996, the first major release of an integrated modeling system in the public domain was the BASINS modeling
system. BASINS provided a linkage between spatial and point data and a watershed model (HSPF) through the use
of emerging geographic information system (GIS) technology. Bundled within the BASINS system was a series of
tools that facilitated data analysis and development of model input files. Semi-automating various data analyses and
facilitating spatial data processing significantly reduced the time and effort required for performing a basic
watershed characterization and developing input files for HSPF. The linkage of GIS and modeling technology
significantly advanced the ability of users to evaluate watershed systems including point and nonpoint sources under
a variety of conditions. Since the original release, the BASINS system has continued to add models (e.g., SWAT,
PLOAD, AQUATOX, KINEROS) and additional systems (e.g., AGWA) to improve the functionality of data
download and management tools.
The TMDL Modeling Toolbox, sponsored by EPA Region 4 and EPA Office of Research and Development,
Watershed and Water Quality Modeling Technical Support Center, provides a loosely linked system that is designed
to facilitate TMDL development. In this framework, the models are developed and supported individually, but
linkage is facilitated by data exchange tools and common data management and GIS interfaces. The Toolbox has
emphasized support for linking watershed and receiving water models. Adding EFDC and WASP to the Toolbox
54
-------
supports the capability to evaluate watersheds, rivers, lakes, and estuaries. Additional specialized tools were added
to address sediment and mercury impairments. WAMView provides an assessment tool for areas with high water
tables. The recent addition of the Conservational Channel Evolution and Pollutant Transport System (CONCEPTS)
provides a tool that can assess channel sediment and stream channel adjustments. The system also includes
watershed report generation tools that facilitate characterization and a database system for managing and analyzing
water monitoring data.
The WMS provides another format for an integrated modeling system, including various data management tools and
watershed models. The WMS includes GIS, support tools, and watershed models. Most recently, WMS has added
support for a grid-based model for hydrologic simulation (GSSHA) and HSPF for water quality modeling. This
system, once tested, could provide more practical applicability for grid-based models for watershed simulation and
TMDLs.
The three most commonly used and available integrated modeling systems are BASINS, the EPA TMDL Toolbox,
and WMS (fact sheets are provided in the Appendix). The three systems are compared in more detail in Table 5-6,
with particular focus on the included analytical tools and model linkages.
Examination of the recent history in development of the three major integrated systems discussed here shows a
continuing expansion in three general areas—data analysis tools, available types and complexity of models, and
linkages between models. Data analysis tools within integrated model systems include pre-processing tools to help
understand watershed and waterbody conditions, perform diagnostic analysis of waterbody conditions or sources,
and process data for use in model input files. Data analysis tools are also used in the evaluation and interpretation of
model output datasets. Models are being added to integrated modeling systems to provide a variety of simple (e.g.,
PLOAD) and more detailed models addressing receiving waters (e.g., EFDC, WASP) or specific source types (e.g.,
SWAT for agricultural applications).
There is also increasing interest in developing systems that link models to each other and facilitate the linkage with
GIS tools (to spatially locate linkage points) and provide file conversion and data management capabilities. For
larger scale TMDL applications, multiple models are often needed to evaluate source loading and receiving water
response (e.g., watershed and estuary models). For some TMDL applications, receiving water models are needed to
compare predicted waterbody conditions to water quality standards. For example, to address dissolved oxygen
impairment in an estuary, an estuary model might be used to simulate eutrophication processes and the sensitivity of
dissolved oxygen concentrations to changes in nutrient loading, and a watershed model might be used to estimate
the magnitude and sources of nutrient loading. Watershed model outputs in the form of discharge time series (i.e.,
daily or hourly flow and concentrations) provide input data at critical boundary points of the estuary model.
Linkages between watershed and receiving water model are typically one-directional, because the data pass directly
from one model to other, and the two models can be run separately. The watershed model output points are selected
above the head of tide so that the flow is one-directional and is not controlled by the tidal fluctuations. Linkages
between ground and surface water models are more complex, with typically bi-directional dynamic linkages.
Surface and groundwater models both need to consider water table elevation and soil moisture content, and the two
models may need to run concurrently with frequent data exchange to maintain continuity and assess water
conservation. Examples of model-to-model linkages that are supported by the three integrated modeling systems
discussed here include:
• Watershed to receiving water—LSPC to EFDC
• Receiving water hydrodynamic model to water quality model—EFDC to WASP
• Watershed model to water quality/ecological model—HSPF or SWAT to AQUATOX
55
-------
Table 5-6. Capabilities of Integrated Modeling Systems
01
I,
V)
BASINS
http://www.epa.gov/ost/basins/
EPATMDLToolbox
http://www.epa.gov/athens/wwqtsc/
http://wcs.tetratech-ffx.com
WMS1
http://www.ems-i.com/index.html
in
•3
GIRAS land use
Soils
Digital Elevation Model (DEM)
Reach File, version 3 (RF3) and
National Hydrography Dataset (NHD)
STORE! water quality data summary
Permit Compliance System data
WDM weather data
303(d) listed waters
Multi-resolution Land Characteristics
(MRLC) land use
Soils
DEM
RF3 and NHD
Agricultural Census
Population data
National Resources Inventory
erosion data
STORE! water quality data summary
Pesticide and fertilizer data
Monthly weather data summary
DEM
Triangulated Irregular Network (TIN)
Land use
Soils
Watershed images
Hydrography
Precipitation
Stream stage
tn
I
I
ro
c
Theme Manager
Import Tool
Data Download Tool
Grid Projector
GenScn
WDMUtil
AGWA
Manual Delineation Tool
Automatic Delineation Tool
PEST (parameter
estimation/calibration) Predefined
Delineation Tool
Land Use, Soil Classification, and
Overlay
Land Use Reclassification
DEM Reclassification
Water Quality Observation Data
Management
Lookup Tables
WCS - characterization reports
(including 12 physical reports, 4
water quality reports, and 5 loading
reports)
WCS - manual delineation
WCS - automatic delineation
WCS - NHD download tool
LSPC data preprocessor
SWMM data preprocessor
NPSM data preprocessor
SNPP - WASP stream network
preprocessor
WRDB - water resources data
access and analysis
Watershed Sediment Loading Tool
Watershed Mercury Loading Tool
EFDC Grid Generator
EFDC Interface and Post-processor
Map Module - defining watershed
data and maps
GIS Module - manipulating spatial
data
Terrain Data Module - processing
terrain data
Drainage Module -watershed
delineation
Hydrologic Modeling Interface
River Modeling Interface
Stochastic Simulations
Scatter Point Module - interpolate
data from scattered points to grids
2D Grid Module -surface
visualization
01
S
. PLOAD
. WinHSPF
. KINEROS
. SWAT
. QUAL2E
. AQUATOX
. LSPC
. PC-SWMM
. EFDC
. QUAL2K
. WASP
. WAMView
. CONCEPTS
. HEC1 (HMS)
. TR-20
. TR-55
. MODRAT
• Storm Drain
. CE-QUAL-W2
• National Flood
Frequency Program
(NFF)
Rational Method
HSPF
HEC-RAS (steady flow analysis)
UNET (unsteady flow analysis)
BRI-STARS (flow and sediment
transport analysis)
GSSHA
in
01
HSPF-AQUATOX
1 ra . SWAT-AQUATOX
•S
c
LSPC - EFDC
EFDC-WASP
. HEC1 - HEC-RAS
Distributed by Environmental Modeling Systems, Inc. (EMS-I, http://www.ems-i.com/index.html). Similar to WMS, EMS-I also
distributes a groundwater modeling system (QMS) and a surface water modeling system (SMS).
56
-------
These types of linkages provide the capability to assess the complex waterbodies that often require consideration in
TMDLs. Numerous other linkages are performed in practice but are not yet supported by a specific modeling
system. For example, HSPF has been linked with MODFLOW to address groundwater-surface water interactions.
In TMDL applications, linkages are often essential for identifying the allowable loading capacity and determining
the source loading components.
The newest trends in model integration are toward Web-based or partially Web-based systems that use the Internet
to extract, manage, and manipulate data. Integral to Web-based data systems is the adoption of uniform data storage
and management standards and formats. Other research has emphasized building universal and flexible systems that
can be used to build model linkages. However, practical application of many of the emerging systems is not yet
demonstrated. Selected emerging systems include:
• Modular Modeling System (MMS). MMS provides a framework for linking modules to provide a
comprehensive modeling system that can be used to develop and test physical process algorithms
(http://www.brr.cr.usgs.gov/projects/SWjrecip rirnoff/mms/). The individual modules can be new code
or created as executable objects from existing models. However, MMS places strict requirements on the
structure of source code that comprises the modules. Application of the system requires experience in
modeling and software development.
• Flexible Modeling System (FMS). FMS (http://www.gfdl.noaa.gov/~fms/) is designed for the
construction of climate models, and is oriented to facilitating parallel and vector solution techniques.
Modules or kernels are developed for high-performance solutions that are linked together based on the
FMS structure specifications. Independent groups of researchers can collaborate by evaluating different
subsystems concurrently. The system includes specific standards and a shared software environment.
• Spatial Modeling Environment (SME). SME (Maxwell and Costanza 1994; Maxwell and Costanza
1997; Maxwell 1999; http://giee.uvm.edu/SME3') provides an environment for linking models and solving
analysis with parallel supercomputers. The modeling environment is graphically based and draws from a
generic object database. The environment facilitates sharing modules and reusing components in new
configurations. Early applications of the system for ecological and nutrient modeling include Patuxent
River, Buzzards Bay, and the Everglades Landscape Model (ELM). One application of the SME is the
Land-Use Evolution and Impact Assessment Model (LEAM) that provides an approach for simulating the
evolution of urban systems by using the Cellular Automata approach combined with the open architecture
tools (http://www.rehearsal.uiuc.edu/projects/leamA).
57
-------
58
-------
Chapter 6 Case Studies
This chapter includes case studies of two TMDL applications for mercury and nutrient impairments, respectively.
Case studies are included to provide an illustration of how models are applied for the purposes of developing
TMDLs. These case studies were selected as demonstrations of the typical techniques used in recent years in
developing TMDLs for two of the most critical pollutant types, mercury and nutrients. Each study had particular
complexities that required the use of models to support the analysis and involved the use multiple models.
Selection of the case studies considered the type of pollutants, the use of multiple models, and the availability of the
final report and supporting documents. Each TMDL considered in the case studies is also complete and approved by
the state and appropriate EPA region.
The first case study discusses the development of mercury TMDLs in Arivaca Lake and Pena Blanca Lake, Arizona.
A combination of watershed loading model, spreadsheet analyses, and mercury lake cycling model is used to
describe various sources of mercury loading, in lake processes, and bioaccumulation of mercury in fish.
The second case study summarizes the TMDL development for nutrients in the Cahaba River, Alabama. This case
study demonstrates the use of multiple models to address elevated nutrient concentrations during low flow that cause
excessive periphyton (attached algae) growth in the Cahaba River. A combination of watershed hydrology
modeling, spreadsheet analyses, and river modeling are used to examine the relationship between various low-flow
sources and resulting algal growth.
Each case study provides a description of the steps in the TMDL development process, with a clear emphasis on the
development of the modeling aspects.
• Background and Problem Identification, including watershed characteristics, listing information and water
quality standards and TMDL targets
• Sources
• Model Selection
• Model Setup
• Model Evaluation
• Model Application
Development of Mercury TMDLs in Arivaca Lake and Pena Blanca Lake, Arizona
Background and Problem Identification
Watershed Characteristics
Arivaca Lake and Pena Blanca Lake are impoundments in rural southern Arizona in the Santa Cruz watershed
(hydrologic unit code [HUC] 15050304) near the Mexican border. Arivaca, impounded in 1970, has a full-pool
surface area of 89 acres, a volume of 1,037 acre-feet, and a maximum depth of 25 feet. Pena Blanca, impounded in
59
-------
1958, has a full-pool surface area of 49 acres, a volume of 1071 acre-feet, and a maximum depth of 60 feet. Both
lakes establish strong summer stratification, which typically breaks down in October.
The region has a semi-arid climate, with abundant rainfall only in July and August. Most of the remainder of the
annual precipitation occurs in the winter months. The watersheds are predominantly evergreen forest and shrub and
brush rangeland, and tributary inflow to both lakes is intermittent. Both lakes are managed as recreational fisheries.
Listing Information
Both Arivaca and Pena Blanca lakes were placed on Arizona's 303(d) list following detection of elevated levels of
mercury in fish tissue and issuance of Fish Consumption Advisories. The criterion used by Arizona to establish Fish
Consumption Advisories is an average concentration in target species of greater than 1 mg/kg wet weight (ppm). In
Pena Blanca Lake, average concentrations in yearly samples of largemouth bass collected from 1994 to 1997 ranged
from 1.31 to 1.53 mg/kg mercury, with individual fish samples ranging up to 2.02 mg/kg. Similar concentrations
have been observed in Arivaca Lake, with largemouth bass averages ranging from 1.03 to 1.5 mg/kg. The two lakes
were determined not to support their designated uses offish consumption.
TMDL Targets
The applicable numeric targets for the Arivaca and Pena Blanca TMDLs are the Arizona water quality standard of
0.2 ug/1 total mercury in the water column and the Fish Consumption Guideline criterion of 1 mg/kg total mercury
concentration in fish tissue. Water column mercury concentrations have not been found in excess of the ambient
water quality standard; however, fish tissue concentrations have consistently exceeded the guideline value. Fish in
the lakes accumulate unacceptable tissue concentrations of mercury even though the ambient water quality standard
appears to be met. The most binding regulatory criterion is the fish tissue concentration criterion of 1 mg/kg total
mercury, which is selected as the primary numeric target for calculating the TMDL.
Mercury bioaccumulates in the food chain. Within a lake fish community, top predators usually have higher mercury
concentrations than forage fish, and tissue concentrations generally increase with age class. Top predators are often
target species for sport fishermen, and Arizona's Fish Consumption Guideline is based on average concentrations in
a sample of sport fish. Therefore, the criterion should not be applied to the extreme case of the most-contaminated
age class of fish within a target species; instead, the criterion is most applicable to an average-age top predator.
Within Arivaca Lake and Pena Blanca Lake, the top predator sport fish is the largemouth bass. A site-specific
spreadsheet model was developed to evaluate lake water quality model output and predict mercury concentrations in
fish tissue for each age class at each trophic level. Average mercury concentrations in fish tissue of target species are
assumed to be approximated by average concentration in 5-year-old largemouth bass. In the May 1995 sampling of
Pena Blanca Lake, the average mercury tissue concentration in largemouth bass (1.31 mg/kg) was slightly lower
than the average concentration in 5-year-old largemouth bass (1.35 mg/kg), and the average concentrations in all
other sampled species were lower than that in largemouth bass. Therefore, the selected target for the TMDL analysis
is an average tissue concentration in 5-year-old largemouth bass of 1.0 mg/kg or less.
Source Assessment
There are no permitted point source discharges and no known sources of mercury-containing effluent in the Arivaca
or Pena Blanca watersheds. External sources of mercury load to the lake include natural background load from the
watershed, nonpoint loading from past mining activities, and atmospheric deposition.
Watershed Background Load
The watershed background load of mercury derives from mercury in the parent rock and from the net effects of
atmospheric deposition of mercury on the watershed. Because no near-field significant sources of mercury
deposition were identified, mercury from atmospheric deposition onto the watershed is treated as part of a general
watershed background load in this analysis. Atmospheric deposition of mercury occurs throughout the world, and
mercury enters these watersheds through both wet deposition (precipitation) and dry deposition. As described below,
atmospheric deposition is estimated to contribute more than 12 micrograms of mercury per square meter per year
(ug/m2/yr). The direct atmospheric loading to the lake is greater than the total estimated load of mercury from the
60
-------
watershed to the lake; however, portions of the atmospheric mercury deposition do not enter the lake because they
are recycled to the atmosphere or sequestered within the watershed.
Mercury is also present within the parent rock formations of the Pena Blanca and Arivaca watersheds. Cinnabar
(HgS), the primary naturally occurring ore of mercury, typically consists of 86.2 percent mercury and 13.8 percent
sulfide. Cinnabar occurs as impregnations and vein fillings in near-surface environments from solutions associated
with volcanic activity and hot springs. Cinnabar also may occur in placer-type concentrations produced from the
erosion of mercury-bearing rocks. In the Pena Blanca watershed, cinnabar has been reported to occur as traces in
irregular and lensing fissure veins in association with argentiferous galena, pyrite, marcasite, and chalcopyrite.
The net contributions of both atmospheric deposition and weathering of native rock were assessed by measuring
concentrations in sediment of tributaries to Arivaca Lake and Pena Blanca Lake. Based on these data, as well as
three background sediment samples from just outside the Arivaca Lake watershed (in areas expected to be relatively
uncontaminated by anthropogenic sources of mercury), most of the sediment samples from the Arivaca and Pena
Blanca watersheds may be considered at or near background mercury levels. In the Pena Blanca watershed,
sediment mercury concentrations were below 100 parts per billion (ppb), except for samples at and just downstream
of an old gold mine tailings pile found during reconnaissance. The sample just below the tailings pile showed an
extremely elevated concentration of 555,000 ppb. In the Arivaca watershed, sediment mercury concentrations were
below 150 ppb, except for samples at and just downstream of the Ruby Dump site, in the southern end of the
watershed. Samples within Ruby Dump had mercury levels as high as 1467 ppb.
Nonpoint Loading from Past Mining Activity
The mining of precious metals such as gold and silver was common in the Parajitto mining area surrounding Pena
Blanca Lake. It was also common in the area surrounding Arivaca Lake, but apparently not within the Arivaca
watershed itself.
Before the introduction of cyanidation technology at the beginning of the 20th century, mercury-amalgamation of
precious metal ores it was a common practice throughout the western United States, using mercury to amalgamate
gold ore in ball mills. Ball mill process mercury is likely to be of greater concern for environmental impact because
the residue is more likely to contain soluble species of mercury than low-solubility cinnabar outcrops. Studies of the
highly contaminated Carson River area in Nevada demonstrate that the dominant form of mercury present in
amalgamation-process tailings is still elemental mercury, approximately a century after peak mining activity, while
stream sediments in the tailings area were dominated by elemental and exchangeable forms of mercury. Significant
conversion to relatively insoluble cinnabar occurs only when these materials are transported to more anoxic,
reducing environments with concentrations of labile sulfur in excess of 0.1 percent by weight. Thus, the mercury
contained in ball mill tailings is likely to be more mobile and more bioavailable than the mercury contained in
cinnabar in the watershed soils and tailings residue from hard rock mines (which were not processed by mercury
amalgamation).
One ball mill site has been identified in the Pena Blanca watershed, associated with the St. Patrick Mine and with a
tailings pile adjoining an intermittent stream bed. In June 1999, 30 samples were collected from the tailings pile,
mill site, and adjacent streambed site. Samples from outside of the tailings pile generally revealed low levels of
contamination (from nondetectable up to 15 mg/kg). Seven samples collected from the tailings pile gave higher
results, ranging from 63 to 460 mg/kg total mercury. These results confirm that the tailings pile is a mercury hot
spot.
Reconnaissance efforts in the Arivaca watershed have not located any obvious ball mill sites within the watershed.
Although the possibility cannot be ruled out, the likelihood of finding previously unknown additional mill sites or
tailings piles in the Arivaca drainage is low. No detectable mercury was found at two known mine shaft sites in the
watershed. However, somewhat elevated levels of mercury were found within the old Ruby Dump, in the southern
end of the watershed. Ruby Dump is located in the southern portion of Arivaca watershed, at the very upstream end
of Cedar Canyon Wash. The dump apparently served the town of Ruby and the Montana Mine. This former mining
town is located about 1 mile southwest of the dump site, outside the Arivaca watershed.
61
-------
Atmospheric Deposition
The third component of mercury loading is direct atmospheric deposition to the lake surface. Atmospheric
deposition is often separated into near-field, or local sources, and far-field sources. Elevated near-field deposition is
often found downwind of coal-fired power plants, smelters, and lime kilns. A variety of potential near-field sources
was evaluated in the United States and Mexico. Based on the lack of major nearby sources, particularly sources
along the axis of the prevailing wind, near-field atmospheric deposition of mercury attributable to individual
emitters is not believed to be a major component of mercury loading to the Arivaca and Pena Blanca watersheds.
Long-range atmospheric deposition is a major source of mercury in many parts of the country. A study of trace
metal contamination of reservoirs in New Mexico indicated that perhaps 80 percent of mercury found in surface
waters was coming from atmospheric deposition. In other remote areas (e.g., in Wisconsin, Sweden, and Canada),
atmospheric deposition has been identified as the primary (or possibly only) contributor of mercury to waterbodies.
Wet deposition of mercury has been measured by the Mercury Deposition Network in its first year of operation
(February 1995-February 1996), the Mercury Deposition Network found a volume-weighted average concentration
of 10.25 ng/L total mercury in precipitation at 17 stations located mainly in the upper Midwest, Northeast, and
Atlantic seaboard (http://nadp.nrel.colostate.edu/ NADP/mdn/mdn.html). Volume-weighted average concentration
of mercury did vary by station, ranging from 3.62 ng/L at Acadia National Park, Maine, to 13.56 ng/L at Bondville,
Illinois. Average weekly wet deposition at the 17 stations ranged from 63 ng/m2 to 280 ng/m2.
Only limited monitoring of atmospheric deposition of mercury is available in the Southwest and none in Arizona.
Dry deposition were measured in the Pena Blanca watersheds (Caballo data), but the monitoring location is about
150 miles closer to the subject lakes than the Pena Blanca lakes. Lack of geographically closer monitoring
introduces considerable uncertainty; however, as shown below, direct atmospheric deposition appears to account for
only a small portion of the total mercury load to the lake. Even if the direct atmospheric loading rate is
underestimated by a significant amount, it would have only a minor effect on the predicted lake response. The
Caballo data therefore were selected to characterize mercury wet deposition to the lake surfaces. The short period of
record available was extrapolated to provide estimates across the period of simulation. Two approaches were
considered to make this extrapolation: development of a relationship between mercury concentration and rainfall
volume, and calculation of average deposition rates. The first approach is based on the observation that mercury wet
deposition concentrations are typically inversely related to rainfall volume. There is considerable scatter in this
relationship in the Caballo data, particularly at low precipitation volumes. Given this scatter and the short period of
record available, the concentration approach was rejected. Instead, it was assumed that cumulative deposition mass
was a more robust estimator than concentration. To make maximum use of the available data, the series of all
possible running 12-month sums were calculated and then averaged, yielding an average annual deposition rate of
4.125 ug/m2-yr (79 ng/m2-wk). This annual sum was then apportioned to months based on the observed deposition
pattern from May 1997 through April 1998.
The Caballo station does not measure dry deposition. Although there are few direct measurements to support well-
characterized estimates, dry deposition of mercury often is assumed to be approximately equal to wet deposition
(e.g., Lindberg et al., 1991), as is reported in the Pena Blanca Lake. Because the climate at Arivaca and Pena Blanca
is wetter than at Caballo, the distribution of wet and dry deposition is likely to be different. Total mercury deposition
at Arivaca/Pefia Blanca is assumed to equal that estimated for Caballo, New Mexico, but Arivaca/Pena Blanca are
estimated to receive greater wet deposition and less dry deposition than Caballo because more of the paniculate
mercury and reactive gaseous mercury that contribute to dry deposition will be scavenged at a site with higher
rainfall.
Model Selection
The TMDLs needed to be completed on a short time line established in a consent decree and with limited resources.
This condition required focusing efforts on those aspects of the problem that were most important to the decision
process. Of the many physical and chemical processes linking mercury sources to bioaccumulation in fish, some
were, of necessity, addressed through simple scoping models, while for other components, highly sophisticated
methods were used. Focusing the modeling effort was a key factor in successful completion of the TMDLs and
depended on two factors: use of a cross-sectional comparison (using a reference lake) and collection of a high-
quality dataset.
62
-------
The Mercury Cycle
Development of the risk hypothesis, and therefore selection of appropriate modeling approaches, requires an
understanding of how mercury cycles in the environment. Mercury chemistry in the environment is quite complex.
Mercury has the properties of a metal (including great persistence due to its inability to be broken down) but also
has some properties of a hydrophobic organic chemical due to its ability to be methylated via a bacterial process.
Methylmercury, easily taken up by organisms, tends to bioaccumulate; it is very effectively transferred through the
food web, magnifying at each trophic level. This transfer can result in high levels of mercury in organisms high on
the food chain, despite nearly immeasurable quantities of mercury in the water column. In fish, mercury is not
usually found in levels high enough to cause the fish to exhibit signs of toxicity, but wildlife that habitually eat
contaminated fish are at risk of accumulating mercury at toxic levels, and the mercury in sport fish can present a
potential health risk to humans.
Mercury and methylmercury form strong complexes with organic substances (including humic acids) and strongly
sorb onto soils and sediments. Once sorbed to organic matter, mercury can be ingested by invertebrates, thus
entering the food chain. Some of the sorbed mercury will settle to the lake bottom; if buried deeply enough, mercury
in bottom sediments will become unavailable to the lake mercury cycle. Burial in bottom sediments can be an
important route of removing mercury from the aquatic environment.
Methylation and demethylation play an important role in determining how mercury will accumulate through the
food web. Hg(II) is methylated by a biological process that appears to involve sulfate-reducing bacteria. Rates of
biological methylation of mercury can be affected by a number of factors. Methylation can occur in water, sediment,
and soil solution under anaerobic conditions and, to a lesser extent, under aerobic conditions. In lakes, methylation
occurs mainly at the sediment-water interface and at the oxic-anoxic boundary within the water column. The rate of
methylation is affected by the concentration of available Hg(II), microbial concentration, pH, temperature, redox
potential, and the presence of other chemical processes. Demethylation of mercury is also mediated by bacteria.
Note that both Hg(II) and methylmercury (MeHg) sorb to algae and detritus, but only methylmercury is assumed to
be passed up to the next trophic level (inorganic mercury is relatively easily egested). Invertebrates eat both algae
and detritus, thereby accumulating sorbed MeHg. Fish eat the invertebrates and either grow into larger fish (which
have been shown to have higher body burdens of mercury) or are eaten by larger fish. At each trophic level, a
bioaccumulation factor must be assumed to represent the magnification of mercury concentration that occurs up the
food chain.
Typically, almost all of the mercury found in fish (more than 95 percent) is in methylmercury form. Studies have
shown that fish body burdens of mercury increase with increasing size or age of the fish, with no signs of leveling
off.
Although it is important to identify sources of mercury to the lake, there may be fluxes of mercury within the lake
that would continue nearly unabated for some time, even if all sources of mercury to the lake were eliminated. In
other words, compartments within the lake probably currently store a significant amount of mercury, and this
mercury can continue to cycle through the system even without an ongoing outside source of mercury. The most
important store of mercury within the lake is likely to be the bed sediment. Mercury in the bed sediment may cause
exposure to biota by being:
• Resuspended into the water column, where it is ingested or it adsorbs to organisms that are later ingested.
• Methylated by bacteria. The methylmercury tends to attach to organic matter, which may be ingested by
invertebrates and thereby introduced to the lake food web. Methylmercury poses the real threat to biota due
to its strong tendency to accumulate in biota and magnify up the food chain.
63
-------
Data Collection
Unfortunately, much of the historic database on environmental mercury concentrations is suspect, due to high
detection limits and potential contamination due to lack of ultra-clean sampling and analysis techniques. Further,
while methylmercury concentrations are key to predicting bioaccumulation, the methylated component is rarely
measured.
To develop a defensible TMDL linkage analysis, high-quality data were needed to describe mercury distribution and
movement in the waterbodies and their watersheds, but time and resources for data collection were limited. During
two sampling events representing stratified and unstratified lake conditions, data were collected using ultra-clean
techniques on mercury species in biota, mercury species and general chemistry in the water column and sediments
of both lakes. Sampling of inflow was not possible, given the highly intermittent nature of runoff. Watershed
sampling therefore focused on evaluating mercury concentrations and sediment characteristics in the beds of the
intermittent stream networks feeding each lake. Another important aspect of data collection was detailed field
reconnaissance. While there are no permitted point sources or active mines in either watershed, the field
reconnaissance, together with evidence from stream sediments, identified an important mercury source area in old
ball-mill tailings in the Pefia Blanca watershed.
Summer stratified surface water concentrations in both lakes were low (about 4 ng/1 total mercury in Pefia Blanca
and 8 ng/1 in Arivaca), but increased with depth, reaching 20 to 40 ng/1 near the sediment. Methylmercury
concentrations below the hypolimnion were 4 ng/1 in Pefia Blanca and 14 ng/1 in Arivaca. Higher concentrations
were found in lake sediment, ranging up to 470 ppb in Pefia Blanca and 192 ppb in Arivaca. Surveys of watershed
sediments revealed higher concentrations. Within Pefia Blanca watershed, concentrations ranged up 554,937 ppb dry
weight and in Arivaca watershed up to 1,222 ppb dry weight; however, most samples were less than 100 ppb.
Background concentrations collected in an area just outside the Arivaca watershed believed to be relatively
uncontaminated by anthropogenic sources ranged from 12 to 197 ppb mercury, consistent with other studies of
background mercury levels in soils. Thus, both watersheds appear to be largely in the range of background
measurements, with isolated mercury hot spots.
Cross-Sectional/Reference Site Approach
The complex nature of mercury cycling in the environment can introduce considerable uncertainty into linkage
analysis modeling. From examination of a single waterbody, it is difficult to determine the relative contributions of
gross mercury loading, internal mercury cycling, and rates of mercury methylation and food chain accumulation to
observed body burdens in fish.
Additional constraints on the analysis can be developed by examining several lakes within the same region
simultaneously (cross-sectional approach). Explaining the differences in mercury load, cycling, and bioaccumulation
among several lakes provides a robust basis on which to develop the conceptual model. Therefore, the linkage
analysis for Arivaca Lake was developed simultaneously with analyses for Pefia Blanca Lake and Patagonia Lake.
Patagonia Lake is within the same region, yet it has acceptable fish tissue mercury concentrations. Patagonia thus
serves as an unimpaired reference site for the cross-sectional analysis. The basic physical characteristics of the three
lakes and their watersheds are compared in Table 6-1.
All three lakes lack known point source discharges of mercury and have a fairly similar distribution of rural
rangeland and forest land uses. The Patagonia watershed has far more historical gold mining operations (but also a
much larger watershed area), but it is not known how many (if any) of the Patagonia mines are associated with
mercury-contaminated ball mill sites. EPA has not detected elevated sediment mercury in the Patagonia watershed.
Physically, Patagonia differs from Pefia Blanca and Arivaca because it has a much larger volume, a larger
contributing watershed, and a shorter hydraulic residence time. Patagonia is also the deepest of the three lakes.
EPA collected data from all three lakes and their watersheds in July 1998, providing a valuable basis for cross-
sectional comparison. All three lakes were strongly stratified with anoxic hypolimnia at the time of sampling.
64
-------
Table 6-1. Cross-Sectional Comparison of Studied Lakes
Characteristic
Surface area (acres)
Volume at full pool (acre-feet)
Average depth (ft)
Maximum depth (ft)
Estimated hydraulic residence time
(yrs), 1985-98 average
Watershed area (acres)
Rangeland (acres)
Evergreen Forest (acres)
Cropland and Pasture (acres)
Urban and Residential (acres)
Water (acres)
Pena Blanca Lake
49
1071
21.8
60
0.36
8820.6
845.9
7906.7
0
33.2
34.8
Arivaca Lake
90
1050
11.7
25
0.33
12696.4
5761.3
6421.1
420.3
26.5
67.3
Patagonia Lake
200
11000
29.1
86
0.16
145904
55509.7
88503.8
1204.2
408.2
278.1
Producing mines identified in MILS 4 inactive
88 inactive
6 active
Mines producing gold
2 inactive
51 inactive
1 active
Note: "Active" mines include those on temporary shutdown as of the 1995 MIL. Prospects are omitted from the tabulation.
At the time of the July sampling, all three lakes had similar total mercury concentrations in the sediment but very
different concentrations in the water column. Lake sediment concentrations in Pena Blanca were somewhat elevated
relative to Arivaca and Patagonia. All three lakes showed significant amounts of methylmercury in sediment, but
Patagonia, unlike Arivaca and Pena Blanca, did not have much methylmercury in the water column. This condition
seems to explain why fish have unacceptable levels of mercury contamination in Arivaca and Pena Blanca but not in
Patagonia.
The July data emphasize that there may be little correlation between the total mercury mass stored in lake sediments
and mercury concentration in fish. Sediment concentrations in Patagonia Lake of both total mercury and
methylmercury were higher than those observed in Arivaca, yet Patagonia Lake has acceptable fish tissue
concentrations while Arivaca does not. Sediment concentrations of total mercury in Pena Blanca were three times
those in Arivaca, but total mercury concentrations in the water column were about twice as high in Arivaca as in
Pena Blanca. These observations—indicating that total mercury concentrations in sediment are not linearly related to
fish body burden—suggest that the linkage analysis requires a model that can describe the relationship between
external mercury load and methylmercury generation.
Why are mercury levels in the water column higher in Arivaca and Pena Blanca than in Patagonia, despite rather
similar sediment concentrations? Strong clues emerge from the water column chemistry results from the July
sampling. As shown in Table 6-2, sulfate is strongly elevated in the hypolimnion of Patagonia relative to the other
lakes, while alkalinity and pH are also elevated, and dissolved organic carbon (DOC) is somewhat depressed.
65
-------
These observations suggest that relatively high sulfate concentrations (under alkaline conditions) promote
precipitation of cinnabar in Patagonia, thus reducing water column concentrations. Differences in sediment
chemistry might also play an important role. The sediment of Patagonia Lake has a stronger reducing environment
and lower organic carbon content than the other two lakes. Finally, Patagonia is the deepest lake, which might
reduce growth of algae and photosynthetic bacteria at the sediment interface.
Table 6-2. Comparison of Summer Hypolimnetic Water Chemistry between Studied Lakes
Parameter
Sulfate (mg/L)
Alkalinity (mg/L)
pH
DOC (mg/L)
Total Hg (ng/L)
MeHg (ng/L)
Total Hg in sediment (|Jg/kg)
MeHg in sediment (|Jg/kg)
Patagonia
185
156
7.5
7
2
0.8
148
0.45
Arivaca
0.2
91
6.6
24
38
14.3
129
0.30
Pena Blanca
7
86
7
10
20
3.9
360
0.95
Risk Hypotheses
In sum, the key differences among the lakes appear to be in water chemistry and in consequent effects on mercury
speciation and cycling, rather than in gross total mercury load (as indicated by sediment concentration). Prior to
model development, this understanding was summarized in the following risk hypothesis:
1. Mercury concentrations in fish are driven by summer methylmercury concentrations in the epilimnion.
2. Summer methylmercury concentrations in the epilimnion are driven by mixing from methylmercury
concentrations in the hypoxic zone just below the thermocline.
3. Methylmercury concentrations below the thermocline are determined primarily by water chemistry and its
effect on mercury methylation in the anoxic portion of the water column and cycling between the water and
sediment, and secondarily by mercury concentration in the sediment or gross mercury loads.
4. Total mercury concentration in the sediments is driven by watershed loads but reflects accumulation over
relatively long periods of time and changes slowly.
For each lake, the linkage analysis components described in the following sections are designed to provide a
quantitative investigation of this risk hypothesis. The linkage tools are separated into several general components.
The first two components address the watershed, and the third and fourth address the lake itself. First is a watershed
hydrologic and sediment loading model, which represents the movement of water and sediment from the watershed
to the lake. This model supports the second component, an analysis of watershed loading of mercury to the reservoir.
A lake hydrologic model is the third component. Finally, a model of lake mercury cycling and bioaccumulation is
used to address the cycling of mercury in the lake among and between abiotic and biotic components. When
combined, these components are the TMDL linkage analysis.
66
-------
Model Setup
Watershed Hydrologic and Sediment Loading Model
An analysis of watershed loading could be conducted at many different levels of complexity, ranging from simple
export coefficients to a dynamic model of watershed loads. Data are not available, however, to parameterize or
calibrate a detailed representation of flow and sediment delivery within the watersheds. Therefore, a relatively
simple, scoping-level analysis of watershed mercury load, based on an annual mass balance of water and sediment
loading from the watershed, is used for the TMDL. Uncertainty introduced in the analysis by use of a simplified and
uncalibrated watershed loading model must be addressed in the Margin of Safety.
Watershed-scale loading of water and sediment was simulated using the GWLF model (Haith et al. 1992). The
complexity of this model falls between that of detailed simulation models, which attempt a mechanistic, time-
dependent representation of pollutant load generation and transport, and simple export coefficient models, which do
not represent temporal variability. GWLF provides a mechanistic, simplified simulation of precipitation-driven
runoff and sediment delivery, yet is intended to be applicable without calibration. Solids load, runoff, and
groundwater seepage can then be used to estimate paniculate and dissolved-phase pollutant delivery to a stream,
based on pollutant concentrations in soil, runoff, and groundwater.
GWLF simulates runoff and streamflow by a water-balance method, based on measurements of daily precipitation
and average temperature. Precipitation is partitioned into direct runoff and infiltration using a form of the SCS
Curve Number method. The Curve Number determines the amount of precipitation that runs off directly, adjusted
for antecedent soil moisture based on total precipitation in the preceding 5 days. A separate Curve Number is
specified for each land use by hydrologic soil grouping. Infiltrated water is first assigned to unsaturated zone
storage, where it may be lost through evapotranspiration. When storage in the unsaturated zone exceeds soil water
capacity, the excess percolates to the shallow saturated zone. This zone is treated as a linear reservoir that discharges
to the stream or loses moisture to deep seepage, at a rate described by the product of the zone's moisture storage and
a constant rate coefficient.
Stream flow may derive from surface runoff during precipitation events or from groundwater pathways. The amount
of water available to the shallow groundwater zone is strongly affected by evapotranspiration, which GWLF
estimates from available moisture in the unsaturated zone, potential evapotranspiration, and a cover coefficient.
Potential evapotranspiration is estimated from a relationship to mean daily temperature and the number of daylight
hours. In the arid Southwest, evapotranspiration often exceeds moisture supply, so stream runoff occurs sporadically
in response to precipitation exceeding infiltration capacity. All the streams feeding Arivaca Lake are classified by
USGS as intermittent and lack a consistent base flow component.
Monthly sediment delivery from each land use is computed from erosion and the transport capacity of runoff, and
total erosion is based on the Universal Soil Loss Equation (Wischmeier and Smith 1978), with a modified rainfall
erosivity coefficient that accounts for the precipitation energy available to detach soil particles (Haith and Merrill
1987). Thus, erosion can occur with precipitation but no surface runoff to the stream; delivery of sediment, however,
depends on surface runoff volume. Sediment available for delivery is accumulated over a year, although excess
sediment supply is not assumed to carry over from one year to the next.
GWLF application requires information on land use, land cover, soil, and parameters that govern runoff, erosion,
and nutrient load generation.
Watershed Mercury Loading Model
Estimates of watershed mercury loading are based on the sediment loading estimates generated by GWLF through
application of a sediment potency factor. A background loading estimate was first calculated, then combined with
estimates of loads from individual hot spots.
The majority of the EPA sediment samples showed no clear spatial patterns in sediment mercury concentrations,
with the exception of the "hot spot" areas identified at Ruby Dump and the St. Patrick Mine tailings pile. Therefore,
background loading was calculated using the central tendency of sediment concentrations from all samples
excluding the hot spots. The background sediment mercury concentrations were assumed to be distributed
67
-------
lognormally, as is typical for environmental concentration samples, and an estimate of the arithmetic mean was
calculated from the observed geometric mean and coefficient of variation. Applying this assumption to the GWLF
estimates of sediment transport yields an estimated rate of mercury loading from watershed background of 178.9
g/yr. This load is ultimately derived from a combination of atmospheric deposition on the land, naturally occurring
mercury in rocks underlying the watershed, and dispersed human activities.
Loading from the Ruby Dump and St. Patrick mine area are calculated separately but are also based on the GWLF
estimate of sediment load generated per hectare of rangeland (the land use surrounding the hot spots), as reduced
by the sediment delivery ratio for the watershed.
Based on assumptions regarding hot spot size and sediment load multipliers, less than 1 percent of the watershed
mercury load to Arivaca Lake appears to originate from Ruby Dump, which is the only identified hot spot in the
watershed. Given the uncertainties in estimation of erosion rates and the incomplete status of the USFS
characterization of the St. Patrick Mine ball mill tailings, the assessment of mercury loading from this area should be
judged to be only a rough, order of magnitude estimate. The estimate suggests, however, that this source plays a
significant role in mercury loading to Pena Blanca Lake. Given the assumptions used to estimate loads,
approximately 69 percent of the watershed mercury load to Pena Blanca appears to originate from the tailings pile
and contaminated downstream sediments. This large percentage is a result of the high average concentration
reported for the tailings (287,000 ppb) relative to the average mercury concentration in sediments in the remainder
of the watershed (48 ppb).
Direct Atmospheric Deposition to Lake
The direct deposition of mercury from the atmosphere onto the lake surfaces was calculated by multiplying the
estimated atmospheric deposition rates times the lake surface area. Although Patagonia Lake has a higher total
annual mercury load, the load per volume of inflow is much lower than those in the two impaired lakes.
Atmospheric deposition directly to the lake surface does not appear to be a major source of total mercury load, as it
is estimated to account for only about 1 percent of the total annual load to the lakes. Atmospheric deposition to the
watershed could, however, constitute a significant portion of the net loading from the watershed.
Lake Hydrologic Model
No monitoring data for inflow, water stage, or outflow are available for Arivaca Lake or Pena Blanca Lake. The lake
levels are not actively managed, and releases occur only when storage capacity is exceeded. Therefore, lake
hydrology was represented by a simple monthly water balance, using the following assumptions:
• Inflow from the watershed is given by monthly predictions from the GWLF model application.
• Direct precipitation on the lake surface is estimated from local observed monthly precipitation depth times
the lake surface area at the beginning of the month.
• Evaporation from the lake surface is estimated from pan evaporation data and a pan coefficient of 0.7. This
estimate represents the ratio between mean annual lake surface evaporation and average annual evaporation
from Class A evaporation pans for this area of southern Arizona, and is within the range recommended by
Dunne and Leopold (1978).
• Net gain from or loss to groundwater seepage through the lake bed is assumed to be zero, lacking any
evidence to the contrary.
• Potential storage at the end of the month is calculated as the sum of initial storage plus inflow plus direct
precipitation minus evaporation.
• The stage-area-discharge curve is used to estimate the surface area and elevation of the lake surface
corresponding to the potential storage at the end of the month. If the lake surface elevation is computed to
be higher than the spillway elevation, the excess volume is assumed to spill downstream.
• Actual storage at the end of the month is the smaller of potential storage and full-pool storage.
68
-------
Surface area and elevation of the lake surface at the end of the month are updated to reflect actual storage.
Lake Mercury Cycling and Bioaccumulation Model
Cycling and bioaccumulation of mercury within the lake was simulated using the Dynamic Mercury Cycling Model
(D-MCM; Tetra Tech 1999). D-MCM is a Windows 95/NT™-based simulation model that predicts the cycling and
fate of the major forms of mercury in lakes, including methylmercury, Hg(II), and elemental mercury. D-MCM is a
time-dependent mechanistic model, designed to consider the most important physical, chemical and biological
factors affecting fish mercury concentrations in lakes. It can be used to develop and test hypotheses, scope field
studies, improve understanding of cause and effect relationships, predict responses to changes in loading, and
support design and evaluation of mitigation options.
The major processes in D-MCM include inflows and outflows (surface and groundwater), adsorption/desorption,
paniculate settling, resuspension and burial, atmospheric deposition, air/water gaseous exchange, industrial mercury
sources, in-situ transformations (e.g., methylation, demethylation, MeHg photodegradation, Hg(II) reduction),
mercury kinetics in plankton, and bioenergetics related to methylmercury fluxes in fish.
Model compartments include the water column, sediments, and a food web that includes three fish populations.
Mercury concentrations in the atmosphere are input as boundary conditions to calculate fluxes across the air/water
interface (gaseous exchange, wet deposition, dry deposition). Similarly, watershed loadings of Hg(II) and
methylmercury are input directly as time-series data. The user provides for hydrologic inputs (surface and
groundwater flowrates) and associated mercury concentrations, which are combined to determine the watershed
mercury loads.
The food web consists of six trophic levels—phytoplankton, zooplankton, benthos, nonpiscivorous fish, omnivorous
fish, and piscivorous fish. Fish mercury concentrations tend to increase with age, and thus are followed in each year
class. Bioenergetics equations for individual fish (Hewitt and Johnson 1992) have been adapted to simulate year
classes and entire populations.
Input parameters for conducting the simulations for all three lakes in this study can be broadly separated into five
categories:
• Hydrologic dynamics and lake physical characteristics (morphometry, stratification)
• External loading rates of Hg (atmospheric, nonpoint source, and point source)
• Thermodynamic and kinetic rate constants
• Water, and sediment chemistry
• Biotic data
Model Evaluation
Sensitivity Analysis on Hotspot Sediment Load Multipliers
Watershed mercury loading from the known mercury hot spots (Ruby Dump and the St. Patrick mine ball mill site)
was calculated separately from the general sediment-associated watershed mercury loading, as discussed previously.
For both hot spots, sediment load per hectare from the hot spot is assumed to be four times greater than that for
normal rangeland. For the Ruby Dump site, this multiplier is intended to reflect the lack of vegetation at the site; for
the ball mill site, it reflects the fine consistency of tailings. This sediment load multiplier factor may be thought of in
terms of the USLE equation (RE*K*LS*C*P), which predicts sediment loss. The K factor for rangeland in the
watershed was set to 0.08, based on data in the STATSGO soil coverage. This K factor represents a sandy soil. The
consistency of the mine tailings, however, has been compared to that of talcum powder. A typical K factor for very
69
-------
fine sand with low organic content is 0.42, which is 5.2 times the K factor for rangeland soil (and would therefore
increase sediment loading estimates by 5.2 times). This factor is compensated somewhat by the fact that sediment at
the ball mill site is likely to be a mixture of native sandy soil and the finer tailings, and by lower slopes at the ball
mill site than the average for rangeland soils in the watershed (because the site is located at the bottom of the
canyon).
For both sites, a simple sensitivity analysis was carried out on the sediment load multiplier to estimate its effect on
overall load estimates. In the case of the Ruby Dump site, doubling the load multiplier has very little effect on the
overall estimated load (Table 6-3). The sediment load multiplier is a more important parameter in the Pena Blanca
watershed (Table 6-4), however, because the ball mill site appears to be the dominant contributor of mercury loading
to the lake.
Table 6-3. Sensitivity Analysis on the Ruby Dump Sediment Load Multiplier
Sediment Load Multiplier Percent of Load Attributed to Ruby Dump
1 0.1%
2 0.2%
4 0.4%
8 0.7%
Table 6-4. Sensitivity Analysis on the Ball Mill Site Sediment Load Multiplier
Sediment Load Multiplier Percent of Load Attributed to St. Patrick Ball Mill Tailings
2 54.7%
4 70.7%
8 82.8%
Mercury Cycling Model
As stated previously, it was necessary to focus the modeling efforts, applying the most sophisticated modeling
approaches to the processes that seem most important in driving fish mercury concentrations (i.e., cycling and
accumulation of mercury within the lake). Therefore, though other modeling components (e.g., the GWLF
watershed model) remained uncalibrated, more effort was expended on the D-MCM component, including
examining known areas of scientific uncertainty in the model, modifying the model to better represent the lakes in
this study, and calibrating the model to the three lakes used in the cross-sectional approach.
Scientific Knowledge Gaps in D-MCM
The current version of D-MCM has updated mercury kinetics and an enhanced bioenergetics treatment of the food
web. The predictive capability of D-MCM is evolving but is currently limited by some scientific knowledge gaps,
including:
• The true rates and governing factors for methylation and Hg(II) reduction
• Factors governing methylmercury uptake at the base of the food web
• The effects of anoxia and sulfur cycling
70
-------
For example, there is evidence that anoxia and sulfides can affect mercury cycling and influence water column
mercury concentrations in lakes (e.g., Benoit et al. 1999, Gilmour et al. 1998, Watras et al 1995, Driscoll et al.
1994), but the underlying mechanisms and controlling factors have not been quantified.
Another important assumption in the current version of D-MCM is that all of the Hg(II) on particles is readily
exchangeable. This results in longer predicted response times for lakes to adjust to changing conditions or mercury
loads than likely would occur. It is quite plausible that a significant fraction of Hg(II) on particles is strongly bound,
reducing the amount of Hg(II) available for mercury cycling and the time required for fish mercury concentrations to
adjust to changes in mercury loadings. The magnitude of this error potentially can be quite large for oligotrophic
lakes with very low sedimentation rates and very long paniculate mercury residence times in the surficial sediments.
For systems that have very high sedimentation rates such as many reservoirs, the practical consequence of this
assumption could be small.
Modifying D-MCM for this Modeling Application
Because strong anoxia in the hypolimnion is a prominent feature during summer stratification for the Arizona lakes
simulated in this study, D-MCM was modified to explicitly allow significant methylation to occur in the
hypolimnion. In previous applications of D-MCM, the occurrence of methylation has been restricted to primarily
within surficial sediments. That the locus of methylation likely includes or is even largely within the hypolimnion
(at least for Arivaca and Pena Blanca lakes) is supported by (1) the detection of significant, very high
methylmercury concentrations in the hypolimnia of Arivaca and Pena Blanca lakes and (2) almost complete losses
of sulfate in Arivaca Lake in the hypolimnion resulting from sulfate reduction. An input was added to the model to
specify the rate constant for hypolimnetic methylation, distinct from sediment methylation.
Calibration of D-MCM
D-MCM was calibrated to the three study lakes by compiling and inputting data specific to each lake on:
• Hydrology and lake physical characteristics (morphometry, stratification)
• External loading rates of mercury (from the atmosphere, watershed, and Ruby Dump)
• Thermodynamic and kinetic rate constants
• Water and sediment chemistry
• Biotic data
Data specific to each of the three lakes were input into the model first, followed by data derived from calibrations
for other lakes where site-specific data were lacking for Arivaca and Pena Blanca lakes. For instance,
thermodynamic and kinetic rate constants specific to the lakes are not available and were obtained from previous
calibrations of D-MCM to lakes in other regions.
Calibration proceeded by running the model with a daily timestep for 10 years and adjusting the model so that
concentrations of mercury in largemouth bass matched observed averages for each lake. Because the hydrology of
these lakes is so dynamic and "flashy," more weight was placed on matching largemouth bass Hg concentrations
than on trying to match predicted and observed water chemistry data precisely. This decision was based on the
following:
• Limited water chemistry data that indicate that chemistry in these systems varies rapidly.
• Hydrologic budgets that show that the hydraulic residence time of all three lakes is relatively short (less
than 0.4 year).
71
-------
• The lack of truly local atmospheric loading data adequate for resolving and validating short-term dynamics
in any of the lakes.
• The fact that mercury concentrations in older cohorts of largemouth bass reflect dietary intake throughout
their life history and are rather insensitive to short-term variations in water column chemistry and Hg
loading dynamics.
The calibrations used the same kinetic (rate constant) assumptions for all three lakes letting only differences in
loading, hydrology, and chemistry dictate differences in response. The following paragraphs give a brief overview
of how the input data were assembled and input to the model.
Calibration of the model assumed that there were no good a priori reasons to use differing rate or thermodynamic
constants for each lake to account for differing mercury behavior. Initial application to Arivaca and Pena Blanca
resulted in large overestimates of the amount of mercury predicted in fish. The particle-Hg(II) partition coefficients
were adjusted for particles in the sediment and water column to yield stronger paniculate binding, thus reducing the
dissolved pool available for methylation. Higher partition coefficients are appropriate for the epilimnion because the
hypolimnion becomes seasonally anoxic, which can reduce the ability of inorganic particles to sorb trace elements.
To further improve the model calibration, focus was placed on one feature of the model known to be potentially
inadequate—the ability of the model to predict the amount of labile Hg truly available for desorption. Previous
simulations with D-MCM have illustrated that, although initial sorption of Hg to particles may be well characterized
by conventional sorption models such as the Freundlich and Langmuir isotherms, desorption of "aged" Hg bound to
particles may not follow the same models. In others, some Hg may become irreversibly bound after adsorption has
initially occurred, and the amount of mercury ultimately available for desorption is less than the initial sorption
models would predict. The final model calibration assumed that watershed background mercury loads were 62
percent available for desorption, while loads derived from ball mill tailings at Pena Blanca Lake and from Ruby
Dump at Arivaca Lake were wholly available. This approach yielded a good match to observations of mercury
concentrations in water and in fish (Table 6-5).
A comparison of model-predicted internal fluxes in the three lakes shows that the key difference between Patagonia
versus Pena Blanca and Arivaca lakes is the rate of hypolimnetic methylation of mercury.
Table 6-5. D-MCM Calibration for Pena Blanca, Arivaca, and Patagonia Lakes
5-year Bass Hg (mg/kg
Lake Parameter Type Methyl Hg (ng/L) Hg(ll)totai (ng/L) wet)
Pena Blanca
Arivaca
Patagonia
Observed
Predicted
Observed
Predicted
Observed
Predicted
3.92
0.00-4.26
14.3
0.00-12.07
0.78
0.00-0.12
11.38
0.00-17.69
1.46-8.3
0.00-6.28
1.14
0.00-11.38
1.42
1.40
1.18
1.18
0.14
0.05
Model Application
The application of the linkage models provides an estimate of the loading capacity of the two lakes, or the rate of
external mercury loading consistent with achieving the numeric targets. The TMDL represents the sum of all
72
-------
individual allocations of portions of the waterbody's loading capacity. Allocations are made to point sources
(wasteload allocations) and nonpoint sources or natural background (load allocations). In many cases, it is
appropriate to hold in reserve a portion of the loading capacity to provide a MOS, as provided for in the TMDL
regulation.
After calibration, the model was used to identify the loading capacity and load reductions necessary to meet the
numeric target in 5-year-old largemouth bass. The response of mercury in 5-year-old largemouth bass to changes in
external loads is nearly linear for these lakes (after a period of several years' adjustment). This is because sediment
burial rates are high and sediment recycling is low, with the majority of the methyl mercury that enters the food
chain being created in the anoxic portion of the water column. The numeric target of 1 mg/kg in 5-year-old
largemouth bass is predicted to be met with a 37 percent reduction in total watershed mercury loads to Pena Blanca
Lake and a 16 percent reduction in total watershed mercury loads to Arivaca Lake.
Because there are no permitted point sources in the watersheds, there are no wasteload allocations in the TMDL.
Load allocations represent assignment of a portion of the TMDL to nonpoint sources.
The current knowledge of mercury sources in the watershed and transport to the lake requires use of a "gross
allotment" approach to the watershed as a whole, rather than assigning individual load allocations to specific tracts
or land areas within the watershed. Loading from geologic sources also has not been separated from the net effects
of atmospheric deposition onto the watershed. Information is currently available to separate sources for load
allocations into three components: (1) direct atmospheric deposition onto the lake surface, (2) loading from
watershed hotspots, and (3) generalized background watershed loading, including mercury derived from parent rock
and soil material, small amounts of residual mercury from past mining operations, and the net contribution of
atmospheric deposition onto the watershed land surface.
The allocations are summarized in Table 6-6. It is assumed that the direct atmospheric deposition load was
essentially uncontrollable at these lakes and is therefore not reduced. For Pena Blanca, loading attributed to the St.
Patrick Mine hotspot is considerable and is already proposed to be addressed by a removal action; therefore, load
reductions are focused on that site. For Arivaca, watershed background is most important, and load reductions must
be achieved there. The allocations are developed as annual average loads and address fish tissue concentrations
associated with bioaccumulation of mercury. Because methyl mercury accumulates in tissue, concentrations in tissue
of fish integrate exposure over a number of years; therefore, annual mercury loading is more important for the
attainment of uses than daily loads.
Table 6-6. TMDL Allocations for Pena Blanca and Arivaca Lakes
Pena Blanca Lake
Arivaca Lake
Allocations
Atmospheric Deposition
Hotspot Loads
Watershed Background
Total
Unallocated
Loading Capacity
(g-Hg/yr)
2.3
18.6
58.6
79.5
65.2
144.7
(g-Hg/yr)
2.3
133.0
58.6
193.9
—
—
Reduction
0.0
114.4
0.0
114.4
—
—
(g-Hg/yr)
4.2
0.7
111.2
116.1
38.7
154.8
(g-Hg/yr)
4.2
0.7
178.9
183.8
—
—
Reduction
0.0
0.0
67.7
67.7
—
—
73
-------
Development of a Nutrient TMDL in the Cahaba River, Alabama
Background and Problem Identification
Watershed Characteristics
The Cahaba River watershed (HUC 03150202) in central Alabama encompasses much of the Birmingham
metropolitan area (Figure 6-1). The Cahaba River, a tributary of the Alabama River, has three segments listed for
impairment by nutrients on the Alabama Department of Environmental Management (ADEM) 2002 §303(d) list.
Historically, nutrient impacts have been documented as nuisance blooms and persistent growth of periphyton. The
purpose of the modeling effort by Tetra Tech was to evaluate in-stream nutrient dynamics and predict the
effectiveness of source reductions based on allocations to point and nonpoint sources of TP. This modeling case
study describes: (1) nutrient TMDL numeric criteria development and interpretation (35 ug/L growing-season
median at specific points), (2) watershed and receiving model design and application, and (3) interpretation of
modeling results to determine TMDL load allocations. The Cahaba River Nutrient TMDLs were proposed as a
cooperative effort between ADEM and EPA Region 4 (ADEM 2004b).
Listing Information
ADEM listed one of the segments for nutrient impairment in 1996, but the remaining two segments listed for
nutrients were added to ADEM's 1998 303(d) list by EPA Region 4, based on consultation with the U.S. Fish and
Wildlife Service (USFWS) regarding combined nutrient and siltation impacts to overall habitat degradation.
USFWS, in addition to other agencies such as EPA Region 4, ADEM, and the Geological Survey of Alabama,
attributed excessive periphyton growth in the Cahaba River as the cause of associated impacts to the aquatic life use,
including impacts to threatened and endangered species of mussels, fishes, and snails.
74
-------
Cahaba River Basin
303(d) Listed Cahaba River
/\/ Streams
Incorporated Areas
Figure 6-1. Location of Upper Cahaba River watershed, Alabama.
Water Quality Standards and TMDL Targets
The state of Alabama has narrative rather than numeric criteria for nutrients, which this requires determination of a
numeric target prior to determining TMDLs. In other words, allowable pollutant loads must be determined to meet
numeric TMDL target(s) established by interpretation of the narrative criteria. Early in TMDL development for the
Cahaba River, EPA Region 4 and ADEM undertook a cooperative effort to determine appropriate nutrient targets to
protect the aquatic life designated use as well as threatened and endangered species viability. These efforts included
literature reviews, consultation with national experts, and an ecoregion-based reference-stream evaluation of
ambient total phosphorus.
After evaluating periphyton growth-limiting nutrient thresholds in the literature, ADEM decided to focus on
determining ambient water-column total phosphorus concentrations during the growing season that would limit
periphyton growth and preserve habitat viability. ADEM decided to use an ecoregion reference-stream approach to
determine the appropriate TP target. The ecoregion reference approach is recommended by EPA (USEPA 2000) in
the absence of sufficient data to support a strictly effects-based target. In a reference-stream approach, ambient
nutrient levels in comparable least-affected streams are evaluated to determine ambient nutrient levels that protect
aquatic habitat and that prevent excessive periphyton growth.
The 303(d)-listed portion of the Cahaba River system is in the southeast portion of the Ridge and Valley province
(Ecoregion 67), as shown in Figure 6-2. Six sites located on "least-impacted" reference streams were assessed for a
few years of monthly ambient water-column nutrient data. ADEM previously established the reference reaches and
their associated watersheds using various methods to characterize their condition and determine if they were good
75
-------
candidates. Such methods include watershed surveys, land use coverage, inventorying point and nonpoint sources,
conducting field reconnaissance, and ultimately collecting chemical, physical and biological data to ensure their
condition and verify the streams are of high quality and fully meet designated uses. A summary of nutrient data at
these sites is listed in Table 6-7.
S Ecoregion 67 Reference Sites
A/303(d) Listed Segments
Level III Ecoregions
Piedmont (45)
Southeastern Plains (65)
Ridge and Valley (67)
Southwestern Appalachians (68)
Interior Plateau (71)
Southeastern Coastal Plain (75)
Figure 6-2. Locations of sites utilized for ecoregion reference-stream analysis.
Table 6-7. Summary of Median Nutrient Concentrations for April-October at the ADEM Reference Stations in Ecoregion 67
Station
DRYC-2
DRYT-9
FRMB-8
HNMB-4
MAYB-1
TCT-5
Samples
7
6
13
11
20
13
11-digitHUC
03150106240
03150106330
03150202090
03160111070
03150202080
03150106330
Median TP
(ug/L)
24
32
26
28
21
19
Median TN
(ug/L)
295
203
228
273
173
167
Stream Name
Dry Cr (Calhoun Co.)
Dry Cr (Talladega Co.)
Fourmile Cr (Bibb Co.)
Hendrick Mill Branch
Mayberry Cr
Talladega Cr
Basin
Coosa
Coosa
Cahaba
Black Warrior
Cahaba
Coosa
76
-------
The nutrient target was determined to be the 75th percentile TP of all data collected at these sites within the growing
season (April-October). This target was calculated to be 35-ug/L median TP during the growing season (ADEM
2004a).
Source Assessment
Nutrient loading causing excessive periphyton growth was found to be a result of both point sources and urban
nonpoint source runoff from municipal separate storm sewer systems (MS4s). The drainage area contributing to the
105 river miles of 303(d)-listed segments in the upper Cahaba River is approximately 1,027 square miles, including
a dozen NPDES-permitted municipal wastewater treatment plants (WWTPs) with discharges greater than 1.0
million gallons per day (MOD), as shown in Figure 6-3. In the study period 1999-2001, these WWTPs operated
advanced secondary treatment processes with little or no implementation of nutrient removal. In drought periods,
wastewater can comprise up to 60 percent of the total streamflow at certain locations, exacerbating eutrophic
conditions in critical periods. Natural streamflow in the river is further modified in low-flow periods by a reservoir
on a tributary, the Little Cahaba River, and a major (~80 MOD) drinking water withdrawal in the mainstem river.
Available data for the TMDL included biweekly nutrient samples collected by local municipalities for three years at
various locations, in addition to a few USGS streamflow gages and ADEM long-term trend data.
Major WWTPs
Minor WWTPs
Water Supply Intake
County Boundaries
' US Hwy 280
' 303(d)-Listed Cahaba River
Lake Purdy
' Streams
Cahaba River Basin
Incorporated Areas
[
BIRMINGHAM RIVERVIEW WWTP CWRS
HOOVER (INVERNESS) WWTP|
HOOVER RIVERCHASE WWTP
20 Miles
Figure 6-3. Locations of major (>1.0 MGD) NPDES-permitted point source discharges in the Upper Cahaba River
watershed
Data analysis to support the water quality modeling effort, specifically prediction of in-stream TP concentrations,
required compilation of extensive NPDES discharge monitoring reports (DMRs) for the major (>1.0 MGD) WWTPs
and estimated discharge and nutrient loading for the minor WWTPs.
77
-------
Nonpoint Source Data Analysis
Nonpoint source TP loads to the Cahaba River were determined by evaluation of in-stream data collected at
locations not influenced by point sources. These nonpoint source evaluation sites and contributing watershed areas
are shown in the land use map in Figure 6-4.
Characteristic nonpoint source TP concentrations for each land use category were derived by the simple correlation
of median TP with MRLC land use classification (urban, forest, and other). Descriptions of each site are listed in
Table 6-8.
O Nonpoint Source Calibration Stations
303(d) Listed Cahaba River
.•'-.'•'NHD Streams
| ~\ NFS Calibration Subbasins
j ' Cahaba Watershed
MRLC Landuse (C03160202!
^^^ Urban
Barren or Mining
Transitional
Cropland or Grasses
Pasture or Grasses
Forest
Upland Shrub Land
Grass Land
Water
Wetlands
16 Miles
Figure 6-4. Water quality sampling sites used to assess nonpoint source concentrations of TP and other nutrients.
Table 6-8. Sites Within the Cahaba River Watershed Not Impacted by Point Sources
Station ID
CR1IS
SC1IS
SC2IS
SC3IS
SC4IS
ST2
ST3
ST4
Location
Cahaba River at Hwy 11 Civitan Park — Trussville
Shades Cr at Elder St near Eastwood Mall in Birmingham
Shades Cr at Columbiana Rd — Lakeshore Drive Junction
Shades Cr at Hwy 150 Galleria area — Hoover
Shades Cr at Dickey Springs Rd (02423630) nr Greenwood
Little Shades Creek above Cahaba River
Patton Creek above Cahaba River
Patton Creek at Patton Church Rd.
Percent
Forest
82
67
57
67
71
56
55
47
Percent
Other
14
11
10
9
12
9
9
8
Percent
Urban
3.7
21.5
32.9
24.5
16.6
34.5
35.5
44.3
Median TP
(ug/L)
50
70
70
66
75
160
130
145
78
-------
The correlations of TP/percent urban and TP/(urban, forest, other) are shown in Figure 6-5. Estimated
concentrations for 100 percent urban land use (-285 ug/L TP) determined by this correlation corresponded very
closely with evaluated MS4/stormwater data (Pitt et al, 2004). Further validation of general application of this
correlation is that estimated forest concentrations (zero percent urban area) are virtually identical to evaluated
medians of reference stream (least-impacted watershed) data at ~25 ug/L TP (ADEM 2004a). Based on the
correlation, nonpoint source concentrations of TP could be empirically applied to tributary streamflows based on
each tributary watershed's land use composition of urban, forest, and other land use percent areas.
_J
"3)
a
Q.
1-
180 -i
140. -
190. -
mn
pn
fin -
do -
on
o
• Median TP
Linear (Median TP) ^
•
+^S^
y = 258.25X + 26.847 ^^^^
R2 = 0.615^^^^
^^ \
^+ +
* ^^ \
0% 10% 20% 30% 40% 50%
Percent Urban Area (MRLC)
Figure 6-5. Nonpoint source concentrations of TP as a function of percent urban area.
Based on the correlation, it was possible to make the estimates of "existing" TP concentrations by land use as
follows: Urban, 285 ug/L; Forest, 25 ug/L; Other, 60 ug/L. (The "Other" category was assumed to be between
Urban and Forest, and the best fit was chosen; the value chosen for "Other" had minimal effect on overall results).
Although it was important to evaluate nonpoint source tributary concentrations of TP, it was obvious from
assessment of in-stream data that sources of TP were dominated by municipal point sources, because high TP
measurements, often in excess of 1 mg/L, corresponded to drought periods of low streamflow and virtually no urban
stormwater runoff. At the time the point and nonpoint source water quality data compilation was complete, it
became apparent that simple mixing of high-TP point source effluent with low-TP streamflow was the dominant
process controlling concentrations of TP in critical locations of the river immediately upstream of reaches where
periphyton growth was of greatest concern.
Model Selection
The modeling system applied to evaluate nutrient TMDLs for the Cahaba River system includes: (1) the LSPC,
which is an enhanced watershed model based on HSPF algorithms (Bicknell et al. 1996) and including a Windows-
based GIS interface; (2) EPDRIV1, a one-dimensional unsteady-flow hydrodynamic model based on CE-QUAL-
RIV1; and (3) a custom-developed Spreadsheet Model which combines results from both simulation models to
interpret scenarios for management options. Load allocations and wasteload allocations necessary to meet the
nutrient target were determined using the Spreadsheet Model, so that the target would be met as a growing season
median at three critical locations.
At the beginning of the effort to model pollutant dynamics in the Cahaba River, because no single model could be
expected to adequately predict hydrology and nutrient fate and transport in such a complex system, it was
determined that at least two models would be required—a watershed model to simulate hydrologic response (i.e.
79
-------
runoff as a function of precipitation, geomorphology, and land use) and a receiving water model that would account
for both hydrodynamic transport and water quality kinetics (i.e., eutrophication).
Development of a comprehensive modeling system to assess the Cahaba system required consideration of urban and
rural hydrology, in addition to nutrient fate and transport in the mainstem river. LSPC was selected as the
hydrologic model for tributaries and the overall watershed, and EPDRIV1 was chosen to assess the mainstem river
receiving water (hydrodynamic transport and water quality). Design of the system was such that tributary and
subbasin hydrology was determined by using the LSPC model based on precipitation records, geomorphology, and
land use classification. Although LSPC features the capability to predict water quality constituent concentrations in
runoff, nutrient concentrations from runoff and tributaries were determined instead by empirical estimates described
in the evaluation of nonpoint sources above.
Overall, the watershed and receiving water models accurately represented nutrient dynamics in the Cahaba River,
but it was necessary to simplify the issue in order to evaluate existing conditions and propose TMDL allocations.
To combine the dynamic elements of watershed hydrology, urban nonpoint source and background phosphorus
loading, and predict in-stream mixing and dilution of major point source inputs, a mass-balance spreadsheet model
("Cahaba Spreadsheet Model") was created with Microsoft Excel, using and combining information from USGS
streamflow gages, daily LSPC model-predicted hydrology, land use classification and associated TP concentrations,
river geometry, EPDRIV1 dynamically predicted stream velocity, and historical WWTP data from NPDES DMRs.
In this way, the Cahaba Spreadsheet Model is essentially a postprocessor for the dynamic model results for tributary
steamflow and transport, but also incorporates empirical and raw data for upstream flow, tributary TP, point source
flow and point source TP.
Model Setup
Watershed Hydrologic Model
Configuration of LSPC was supported with the Arc View application known as the Watershed Characterization
System (which uses datasets specific to states within EPA Region 4) and the WCS-to-LSPC autoconfiguration tool,
which automatically delineates watershed subbasins guided by streamlined user input, overlays and tabulates land
use areas, and calculates reach lengths and slopes based on DEM and NHD data.
WCS was used to automatically delineate 300 subbasins within the watershed. Using the WCS-to-LSPC tool, land
use areas for each MRLC classification were tabulated for each subbasin and converted to the LSPC database
format, which uses Microsoft Access. A pictorial example of land use areas within subbasin boundaries for a few
headwaters subbasins is shown in Figure 6-6.
More than 300 subbasins were delineated, and each was assigned to one of a dozen hourly and daily precipitation
stations featuring complete records for three years (1999-2001). Average subbasin size for the LSPC watershed
model is approximately 3 mi2. The model was run for a three-year period corresponding to available precipitation
data at all of the sites, although some had to be patched with adjacent stations for short periods, and, in some cases,
daily rainfall was disaggregated to hourly totals. Locations of the 12 precipitation stations are illustrated in Figure
6-7.
LSPC model output for daily streamflow at 88 subbasins corresponding to tributary subbasins were passed to the
EPDRIV1 mainstem hydrodynamic model. Both daily streamflow and monthly median tributary flows were
incorporated into the Cahaba Spreadsheet Model. An example of the LSPC model interface is shown in Figure 6-8.
80
-------
I I Headwaters above Trussville
/\/ Streams
Land Use Classification
| Urban
^B Barren or Mining
Transitional
Crops or Grasses
Agriculture - Pasture
^| Forest
Upland Shrub Land
Grass Land
^| Water
Wetlands
Figure 6-6. MRLC land use aggregation calculated by LSPC subbasin delineation.
| | Theissen Polygons
s Weather Stations
/\/ 303(d) Listed Segments
/\/ Streams
/\/ Lake Purdy
| I Cahaba River Basin
CentrevilleWSMO
Figure 6-7. Precipitation sites with 3 years of hourly or daily data used in the LSPC watershed model.
81
-------
15PC - [tkatftHutMenwit -
»r pa,
g*t*i
tj
..C '.x.
.=_*..*.
/ w
if r~ ^r
J
J IX
r v-
Figure 6-8. Example of LSPC GIS interface for selecting headwater subbasins.
Mainstem Hydrodynamic Model
In-stream hydraulics and transport in the mainstem Cahaba River were calculated using the EPDRIV1 model, the
Cahaba application of which was originally configured by Jefferson County Environmental Services Division in
conjunction with a SWMM watershed model (now superseded by the LSPC watershed model). For TMDL
development, the EPDRIV1 model's extent was expanded upstream and downstream to encompass 303(d) listed
segments of the Cahaba River, using additional cross-section surveys from Federal Emergency Management Agency
(FEMA) flood studies for a total of 160 cross sections and 105 river miles. A minimum flow was instituted at times
of zero flow in the headwaters to allow the model to run. Examples of model cross sections are shown in Figure
6-9.
Output tributary streamflow predictions from LSPC were linked to the EPDRIV1 input fileset at 88 cross-sections
via a custom post-processing routine (LSPCRIV1) written in FORTRAN. Due to hydraulic modification at the
drinking water withdrawal and operation of the Lake Purdy reservoir, streamflow was corrected mid-river to
correspond with the USGS streamflow record at a nearby location, and calibrated at downstream locations based on
calculated hydrodynamic transport and additional LSPC-derived inflows.
82
-------
Figure 6-9. Examples of EPDRIV1 cross-sections derived from FEMA survey data.
Mass-Balance Spreadsheet Model
The Cahaba Spreadsheet Model estimates monthly median TP concentrations at 160 points along the Cahaba River
(corresponding to EPDRIV1 cross sections) from Trussville to Centreville, for each month in the study period 1999-
2001 and based on historical and projected point source loads, historical flows, and estimated nonpoint source loads.
The transport scheme was simplistic in nature, accounting for TP loss from the system by simple first-order decay
based on time of travel (cf.. SPARROW work of Smith et al. 1997). The decay parameter was chosen to be 0.25
day"1 by best fit to measured in-stream data.
Inputs for the Cahaba Spreadsheet Model included the following combination of data sources:
• River geometry (segment length) from EPDRIV1 cross-section input
• USGS monthly median streamflow at Trussville (upstream boundary) and Caldwell Mill (below US 280
dam)
• Predicted monthly median streamflow at 88 tributary points from LSPC watershed model
• Estimated empirical nonpoint source nutrient concentrations based on percent urban land use for all 88
tributary subwatersheds
• Reported and estimated monthly WWTP effluent discharge and nutrient concentrations from DMR reports,
as available
• Predicted monthly median in-stream velocities from EPDRIV1 simulation output
83
-------
A schematic of the basic process of combining these data in the Cahaba Spreadsheet Model is shown in Figure 6-10.
Geometry
(160
reach
lengths)
ll
li
u. a
Upstream Boundary - TP calculated by land use (%urban)
—Inflow volume based on USGS Streamflow at Trussville
_ „ (02423130)
Tributary Inflows - TP calculated by land use (%urban)
—Inflow volumes calculated by LSPC watershed model
- •* for 88 subbasins
Point source inflows - TP from average or DMRs
~ "* —Inflow volumes from monthly DMRs
Time of Travel - Computed from median monthly
RIV1 model velocity
—Corrected to approximate volumetric residence time
~--TP "loss" approximated as first order exponential decay
Figure 6-10. Schematic of the functional relationship of data inputs in the Cahaba Spreadsheet Model.
Calculations of in-stream TP are performed on a monthly timestep, the smallest timestep on which point source
nutrient data are available. Because the nutrient target was established as a growing-season median, the monthly
results from the Spreadsheet Model have been aggregated to examine in-stream TP concentrations for three years of
growing seasons.
Model Evaluation
Watershed and Mainstem Streamflow
Streamflow predictions from LSPC were calibrated to USGS Streamflow in multiple locations, including the upper
Cahaba River headwaters and tributaries such as Shades Creek. An example of hydrologic predictions in a tributary
of the Cahaba River is shown in Figure 6-11.
Hydraulic transport results were calibrated in EPDRIV1 using friction factors and with particular attention to the
hydrologic discontinuity at the US 280 dam, water withdrawal, and controlled discharge from Lake Purdy.
Ultimately, the EPDRIV1 model average velocities were evaluated to derive daily time of travel between cross-
sections, which was transferred to the Cahaba Spreadsheet Model.
84
-------
Daily Total Precipitation (Year 2001)
-Year 2001 Observed Flow
10000 -,
Modeled Flow
r 0
1000 — ^t=
to
LL.
o,
1
D)
O
100
^N ^N
Figure 6-11. Example of hydrologic model calibration: model and observed streamflow.
Spreadsheet Model: Mass Balance for Flow and Total Phosphorus
Mass balance for flow in the Cahaba Spreadsheet Model was established by combining flows from USGS daily
streamflow data at the upstream boundary and below the US280 dam, 88 watershed tributary daily inflows to the
mainstem, and 9 major and 22 minor point sources on a monthly basis. Monthly median combined flows, used
because long-term water quality analysis ultimately was performed monthly, compare favorably to USGS data, as
shown in Figure 6-12.
10000.00,
Median Flow
Minimum
Maximum
USGS Min
USGS Mean
200.0
100.0
80.0
Figure 6-12. Example of monthly streamflow predictions compared to USGS data at seven sites.
Mass balance for total phosphorus was calculated by mixing upstream cross-section TP concentrations with tributary
and point source TP concentrations (where applicable) and their associated monthly median streamflow values.
Losses during downstream transport were estimated by a first-order exponential loss factor based on EPDRIV1 time
of travel. An example of predicted longitudinal TP concentrations (in this case, monthly median concentrations) is
shown in Figure 6-13. Data from grab samples from random times during a given month were not expected to
85
-------
precisely match the monthly median of daily model results, but the results are very similar and trends are identical
(primarily dominated by longitudinal locations of major point sources and tributaries).
Estimated Historical TP Concentrations from Trussville Downstream
3500
3000
2500
2000
1500
•= 1000
500 -
Estimated TP
TP Data
Max TP
MinTP
200.0
180.0
160.0
140.0
River Mile
120.0
100.0
80.0
Figure 6-13. Estimated monthly median TP concentrations in the Cahaba River in September 1999, from Trussville
(upstream at left) to Centreville (downstream at right). Gray lines indicate maximum and minimum monthly predicted TP
concentrations based on streamflow variation.
Model Application
Spatial Interpretation of Nutrient Target
Once the integrated results of the Spreadsheet Model made it possible to evaluate long-term (growing-season) TP
conditions to compare to the ecoregion reference stream target of 35 ug/L TP, it was still necessary to determine the
spatial applicability of the target. ADEM determined that evaluating TP concentrations at three sites (rather than
every reach of the river) would be sufficient to prevent periphyton growth at levels that would potentially affect
aquatic life uses. These sites were determined to be upstream of critical sites where periphyton had been confirmed
to be a major problem. At the critical sites, it was apparent that the greatest periphyton growth effects observed had
been due to not only historically high TP levels but also geomorphological conditions: wider and shallower reaches,
more sun exposure, and higher availability of substrate. Based on these observations, the spreadsheet model was
used to evaluate growing-season TP conditions at three sites upstream of critical periphyton reaches. The selection
of three sites also simplifies ADEM's future workload of follow-up monitoring for comparison of in-stream ambient
conditions to the TMDL target. The three sites selected for TMDL evaluation are illustrated in Figure 6-14.
86
-------
A/
Instream Compliance Points
Water Supply Intake
NPDES Municipal Point Sources
NPDES Small Point Sources
US Hwy 280
Lake Purdy
303(d)-Listed Cahaba River
Streams
County Boundaries
Cahaba River Basin
JEFFERSON CO
Old Montgomery HwyJ/
~\| Shelby Co. Hwy 52
Figure 6-14. Cahaba River nutrient TMDL evaluation points upstream of critical reaches.
TMDL Results: Determining Wasteload Allocations and Load Allocations
Based on the three critical years (1999-2001), the TMDLs were determined as the necessary load allocations and
wasteload allocations to achieve the growing-season median water quality target of 35 ug/L (April-October) at the
three designated critical locations along the river. Point source TP concentrations at maximum permitted flow were
reduced such that the target was not exceeded at the three evaluation sites for the average of the three (1999-2001)
growing-season median values. In addition, nonpoint source concentrations were reduced from urban land areas by
65 percent in the final scenario. A longitudinal perspective of growing-season median TP concentrations, the final
TMDL scenario is shown in Figure B-15.
ADEM and EPA determined the Cahaba River Nutrient TMDLs to be achieved in three phases of implementation
over 15 years. Wasteload Allocations and Load Allocations necessary to meet the three phases of the TMDL are
shown in Table 6-9.
87
-------
200
Growing Season Median TP (1999-2001) fromTrussville Downstream
• • • TP Target
^^^^Growing Season Median
1999 Growing Season Median
2000 Growing Season Median
2001 Growing Season Median
180
160
140
River Mile
120
100
80
Figure 6-15. TMDL scenario, growing-season median TP for 1999-2001 and three=year average
Table 6-9. TMDL Summary for Cahaba River Phased TP Reductions
Reduction
Phase
Phase 1
(Initial
Reductions
2005-2010)
Phase 2
(Intermediate
Reductions
2010-2015)
Phase 3
Final Reductions
(2015-2020)
Growing Season
(April-October)
Wasteload
Concentration (ug/L
TP)
400 major WWTPs
(21.0MGD design)/
2000 minor WWTPs1
(<1.0MGD design)2
200 major WWTPs
(21.0MGD design)/
500 minor WWTPs
(<1.0MGD design)1
43 major WWTPs
(51.0 MGD design)/
300 minor WWTPs
(<1.0 MGD design)1
Nonpoint Source Load
Allocation and MS4
Wasteload Allocation by
MRLC Land Use
Classification (ug/L TP)
285 urban/MS4
25 forest
60 other
214 urban/MS4
25 forest
60 other
100 urban/ MS4
25 forest
60 other
Continuous Point LA and MS4 WLA
Source Percent Percent Reduction by
Reduction from MRLC Land Use
1999-2001 loads Classification
36% 0% urban/MS4
0% forest
0% other
62% 25% urban/MS4
0% forest
0% other
81% 65% urban / MS4
0% forest
0% other
Minor WWTPs with a current permit limit less than that proposed will keep their current limit.
2Margaret WWTP (0.5 MGD), due to its headwaters location, is required to meet the following: Phase 1-1000 ug/L; Phase 2-250
ug/L; Phase 3-150 ug/L
MS4 and urban nonpoint source loads (considered identically as land use-based sources in this analysis) were
determined by the empirical method using percent urban, forest, and other land uses described above. Urban loads
were derived from empirical data and fraction of USGS MRLC land use classification designated as urban types
(high-intensity residential, low-intensity residential, and high-intensity commercial/industrial/transportation). MS4
loads included in the WLA are defined as urban area loads within designated NPDES MS4 boundaries, but urban
area loads outside of MS4 areas are defined as part of the LA, in order to be consistent with EPA guidelines. No
-------
reductions are required from forested areas or "other" land use classifications. Reductions in loading from MS4 and
urban areas are required beginning in Phase 2 because marginal benefits of such reductions would be negligible in
Phase 1. Continuous point sources (WWTPs) dominate in the low-flow summer period; therefore, only WWTPs are
required to make a reduction in Phase 1.
Table 6-10 shows existing and predicted in-stream growing season (April-October) median total phosphorus
concentrations for the three phases of implementation at the three critical evaluation points on the Cahaba River.
Table 6-10. Existing and Predicted Instream Growing Season Median TP in the Cahaba River
Evaluation Site
Cahaba River at Roper Rd.
Cahaba River at Old
Montgomery Hwy
Cahaba River at Shelby Co.
Hwy 52
Existing Condition
(1999-2001)
1999-2000 Instream
Growing Season
Median Conditions1
(ug/L TP)
1,140'
895
560
Phase 1 (Initial
Reductions 2005-2009)
Predicted Instream
Growing Season
Median Conditions1
(ug/L TP)
124
247
156
Phase 2
(Intermediate
Reductions 201 0-201 4)
Predicted Instream
Growing Season
Median Conditions1
(ug/L TP)
67
126
82
Phase 3 (Final
Reductions 2005-2010)
Predicted Instream
Growing Season
Median Conditions1
(ug/L TP)
31
35
25
Instream conditions are evaluated as median value of growing season April-October.
2Downstream of Trussville site existing conditions shown due to lack of data at Roper Rd.
The Cahaba Spreadsheet Model, using and simplifying results from the dynamic watershed model LSPC and
mainstem river model EPDRIV1, has proven extremely useful as a management tool to judge at a screening level the
results of proposed management actions (i.e., NPDES point source wasteload allocations for nutrient TMDLs).
Although the watershed system exhibits extremely dynamic hydrology and water quality conditions, at the high
concentrations of TP caused by primarily WWTP effluent in low-flow critical conditions, consideration of the
dominant process of mixing and dilution has proven sufficient to evaluate the system for management alternatives,
namely a 3-tier phased implementation schedule of nutrient TMDLs (ADEM 2004b). Furthermore, the system of
linking models together to take advantage of the strengths of each can be extended in a modular fashion. For
example, output of LSPC and EPDRIV1 has been used to drive an experimental application of WASP6 simulating
periphyton conditions in the Cahaba River, and the water quality module of EPDRIV1 can be used if necessary in
the future, although its data requirements are more extensive than those of the Cahaba Spreadsheet Model.
89
-------
90
-------
Chapter 7 Research Needs
The research needs for supporting modeling for TMDL development are as varied as the watersheds and water
impairments throughout the United States. Examining research needs for modeling must address many issues,
including the interface between users and models; the fundamental science and physically based processes employed
by models; and the supporting framework of reference data, training and application guidance. This section builds
on the previous model evaluations and case studies to identify key recommendations for model capabilities,
supporting tools and systems.
Methodology for Identifying Research Needs
In Chapters 1 and 2, the needs for TMDL modeling were
discussed and some of the critical limitations in existing
models identified. Chapter 3 focused on the general
types and formulation of models. In Chapter 4, the
available models were identified and described. In
Chapter 5 the models' capabilities to meet specific
TMDL analysis needs were discussed. Applicability of
models was evaluated across multiple performance
criteria, including hydrology, sediment, water quality
simulation, BMP simulation, and user interface and
supporting tools. Examination of the model review
tables also highlighted areas where models can provide
detailed and comprehensive representation (e.g.,
eutrophication processes). In addition, integrated
modeling systems were discussed and the current
availability of modeling linkages and supporting tools
reviewed. Chapter 6 demonstrated the use of multiple
linked models in the evaluation of TMDLs for mercury
and nutrients. The review and case studies demonstrated
some of the current strengths and limitations of modeling
systems. A variety of supporting tools were needed in
the analysis and the case studies showed a diversity of
approaches used. Depending on the pollutant type and
critical conditions the emphasis of the modeling varied.
The case studies indicated that continued work is needed
in supporting the various linkages and expanding the
science of nutrient/algae and mercury simulation.
Dominant Trends in Application Needs
Integration of water systems. Many traditional models
specialize in individual waterbody types such as lakes,
rivers, or estuaries. Watershed-based studies may require a
comprehensive assessment on a larger scale and the ability
to evaluate multiple management alternatives. Increasingly,
modeling studies need to address the linked behavior of
multiple surface waterbodies or multimedia, including rivers
and lakes, estuaries, and coastal regions. In other
applications, considerations of interactions between air,
surface, and groundwater need to be evaluated.
Model Complexity. The typical perception of scientific
progress is a trend to develop increasingly detailed physical
representations of systems. This trend is demonstrated by
the use of physically based models, models that include
solutions of fundamental equations, and 3-dimensional
representations of waterbodies. However, in TMDL and
watershed-based applications, there is an equally strong
need to employ simplified solution techniques. Users are
searching for reasonable simplifications, practical solutions,
and easily applied tools. The need for simple techniques is
driven by data, time, and resource constraints.
Management Planning. TMDLs and related watershed-
based programs (e.g., NPDES stormwater, 319 NPS plans)
are moving toward implementation planning and
assessment. Implementation planning requires analysis of
BMP performance, management alternatives, and
consideration of cost. Implementation planning might
include examination of multiple solutions and selection of
preferred approaches based on a broader assessment of
efficiency, cost and social-political considerations.
To identify specific recommendations for future development, the modeling needs were evaluated in the context of
TMDL requirements and watershed applications, current trends, and evolving model application needs. In addition,
the needs analysis recognizes that future model development efforts can exploit emerging technology in computer
hardware and software, Internet access, and current research. Historically, several leaps in modeling interfaces and
user support tools occurred with the emergence of GIS technology. The convergence of disparate technologies and
science continue to provide an opportunity for innovation and significant expansion of the capabilities of models and
applicability of more detailed, physically based simulation techniques.
Modeling systems or interface development has always been closely linked to the availability of new computing
technology. Early advances in modeling were closely linked to the availability of personal computers (1980s) and
91
-------
later the emergence of GIS (1990s). Technological advances continue to provide opportunities to facilitate model
application. Creative adoption of new technologies can profoundly change the ways that users work with data and
models in the future. New technologies also can influence collaboration during model development and review and
interpretation of model results. Key recent and potentially relevant technological advances include:
• Significant increases in computing speed and parallel processing capabilities can support more complex
and computationally intensive applications.
• Reduction in cost and increase in size of data storage devices allow users to store and process much larger
datasets efficiently.
• The proliferation of broadband access can facilitate data access, remote data storage, and on-line analysis
capabilities (e.g., databases, mapping, searches). Broadband can affect all stages of model development
and application.
• Research and development of "grid-based" computing technologies and "middleware" can provide
techniques for Internet-based data exchange, data management, and distributed computing techniques.
• Visualization software is now more widely available and affordable. Broader availability of visualization
software encourages use of dynamic displays and interpretative tools for modeling results.
• Remote sensing systems and software are available and can more reliably provide spatially heterogeneous
land cover, stream quality, soil moisture, and precipitation information.
The following sections provide specific recommendations developed by combining an understanding of the needs
with the availability of modeling tools and the emerging technological trends. Some recommendations for
supporting tools and guidance are also identified. Application of models also relies on the availability of appropriate
guidance, reference documents, and supporting information. Specific needs for model support are identified as well.
The recommendations provided below are organized in the following major areas:
• Model capabilities
• Data
• Model defensibility
• Systems development/supporting tools
• Integrated modeling systems
Model Capabilities
Currently available models support assessment of a range of sources and
waterbody types. Described below are recommendations for enhancing and
broadening the capabilities of models to encompass specific source behavior,
provide more sensitivity to management changes, and improve resolution of
results. These changes could be implemented by enhancing existing modeling
systems or through development of new models or supplemental components.
Where appropriate, enhancements or updates for specific models are
recommended. The recommendations are grouped according to sources (e.g.,
irrigation, septic systems) and functional groups, including hydrology,
sediment, nutrients, pathogens, ecological simulation, and evaluation of
management practices.
This section discusses model
capabilities for the following topics:
Sources
Hydrology
Sediment Loading
Mercury
Pathogens
Management Practice Simulation
Hydrodynamics
Sediment Transport
Nutrients/Eutrophication
Ecological/Habitat
92
-------
Sources
Development of a TMDL allocation includes the distribution of the load among sources. Sources are often grouped
in categories and subcategories such as agricultural and urban areas. Source groupings may vary in detail and type,
depending on the regional dominance of source types and the desire of the TMDL developers to explore source
behavior and management options. In some applications, dominance of a particular source type or the dynamic
nature of source behavior (e.g., time variable management or change) will require new modeling approaches or more
refined, detailed model development. Listed below are the major research needs that have been identified for
improving representation of source behavior in models.
• Variable Land Use and Cover. Existing watershed models should be enhanced through improved spatial
data interfaces or new models developed to better simulate time-variable land uses or changes in land use
management activities over time. Currently, watershed models are applied based on a fixed or static land
use coverage. Future or alternate land use configuration is represented by using a separate, static land use
coverage. Remote sensing technology could provide a more dynamic representation of historic land use
conditions. Land use planners also simulate projections of land use conversion over time. A system that
can integrate land use simulation and watershed modeling in one framework could be used to dynamically
incorporate land use conditions and the effects on watershed loading. Variable land use and cover
representation can improve the accuracy of predicted conditions compared to historic flow and water
quality measurements. This type of dynamic land use change could be especially important for western
watersheds where timber harvesting and fires can dramatically change conditions in a matter of years.
Similarly, areas experiencing rapid urban growth could be modeled more accurately through
implementation of a variable land use schema. The LEAM provides an approach for simulating the
evolution of urban systems by using the Cellular Automata approach combined with the open architecture
tools of STELLA and SME (http://www.rehearsal.uiuc.edu/projects/leam/). LEAM simulates land use
transformation and effects on multiple metrics, including water quality. Potential linkage of this
technology with more detailed representations of water quality processes will benefit TMDL evaluations of
dynamic land use conditions.
• Tile Drainage. Existing watershed models (e.g., GWLF, HSPF/LSPC) should be enhanced to better
simulate the hydrologic and water quality effects of tile drainage. Many areas of the country include
extensively drained areas that affect the hydrologic response of the system. Models that include
capabilities for representation of drainage and examples of their application are needed.
• Irrigation. Irrigation practices have a significant effect on the water balance, runoff, and water quality of
watersheds. Irrigation transfers water from surface and groundwater sources and increases water available
for evapotranspiration. Better tools for assessing effects of irrigation on water quantity and quality could
be built into existing watershed models. Significant modeling work has been done to simulate irrigation for
agronomic purposes. For example, watershed models such as HSPF and GWLF do not explicitly
incorporate irrigation practices.
• Roads. Paved roads provide significant impervious coverage, and unpaved roads are often major sources
of sediment. Existing models can be used to represent a variety of surfaces through use of pervious and
impervious land use coverages, but a more systematic approach (similar to WEPP) could be incorporated in
frequently used watershed models (e.g., SWMM, HSPF, SWAT) to enhance their ability to accurately
account for loadings from different types of roads.
• Septic Systems. Septic systems are a potential source of nutrients and pathogens in many rural and
suburban areas. Watershed models do not typically include a standard approach for representation of septic
systems. Potential septic system loading is often estimated as a function of failure rate and provided as a
direct input (i.e., discharge) to watershed models. Septic system loading is typically a function of location
(proximity to waterbodies), failure rate, and water table elevation. Currently available watershed models
should be enhanced to simulate nutrient and pathogen loads from septic systems.
• Coalbed Methane. In western states (e.g., Wyoming), coalbed methane extraction has resulted in the
discharge of waters with elevated concentrations of sodium and total dissolved solids. The water has
limited suitability for domestic and animal consumption—its high saline and sodium content make it
93
-------
unsuitable for agriculture—and it has the potential to damage wildlife habitat and surface water supplies.
These issues have been the central focus of TMDL development efforts in the Tongue River, Powder River,
and Rosebud Creek watersheds in southeastern Montana and northeastern Wyoming
(http://www.deq. state.mt.us/wqinfo/TMDL/TonguePowderRosebudTMDL.asp). Currently, the effect of
coalbed methane is simulated using simple mass balance approaches or with continuous watershed
modeling that addresses only direct discharges. Both of these approaches simplify the in-stream kinetics,
however, and do not address the potentially significant affects on groundwater from leaking containment
ponds. Modeling could be improved by better integrating detailed groundwater models with existing
watershed models, as well as obtaining additional data on the fate and transport of discharge water that is
pumped to containment ponds.
Hydrology
Hydrology in watershed models is well understood, and various levels of simulation are available in numerous
models (e.g., GWLF, SWAT, HSPF, SWMM, GSSHA). Depending on the specific application, annual, seasonal,
daily, hourly or smaller timesteps can be used, and simulation can include sensitivity to rainfall intensity. Many
models include various formulations for representation of snow accumulation and melting processes as well.
However, improvements could still be made in the application and understanding of existing models, the structural
formulation of hydrologic systems (e.g., representation of land surface), and surface-groundwater interactions. In
addition to the traditional hydrology modeling systems, there has been a growing trend toward developing
physically based distributed watershed models over the past decade. These models are necessary because decision
makers are asking questions that require a more refined examination of the spatial heterogeneity of watershed
systems. Distributed models have the potential to examine changes in management techniques, provide sensitivity
to fine variations in location and changes in slope or regrading, and represent a diversity of soils and land use
characteristics. Distributed models can predict at spatial scales smaller than what can be modeled with lumped
parameter models. A distributed model's primary advantage is its physically realistic representation of hydrological
and pollutant transport processes, instead of the generalizations used by lumped models. Key limitations are the
availability of spatially distributed soils and management information, lack of water quality simulation capabilities,
and the processing time required for analysis. Recent advances in remote sensing, availability of spatially detailed
data, and computer processing capabilities continue to support the application of distributed models. The distributed
models will remain an essential part of research designed to understand watershed system processes.
The following are the primary research needs regarding model simulation of hydrology:
• Models Based on the Soil Conservation Service (SCS) Curve Number Equation. Guidance should be
reviewed, tested, and developed on where the use of the simplified SCS Curve Number approach is
adequate for continuous hydrologic simulation. The Curve Number approach is used in models such as
GWLF and SWAT (one option) that are employed in many TMDLs; however, the Curve Number
(developed for event volume estimation) can provide biased and inappropriate estimates of hydrograph
components in certain soils and landscapes. The use of Green and Ampt or other alternative infiltration-
based approach should be provided as an alternative, similar to the optional formulation provided in
SWAT.
• Spatially Distributed Meteorologic Data. The Next Generation Weather Radar (NEXRAD) system,
operated by the U.S. Departments of Defense, Commerce, and Transportation, provides data and
processing software for weather observations throughout the United States. This system operates
approximately 159 Weather Surveillance Radar - 1988 Doppler (WSR-88D) stations throughout the United
States and select international sites, with data available through the National Climatic Data Center.
Through the transmission and reception of electomagnetic signals, the system provides improved spatial
and temporal resolution of rainfall estimates over traditional rain gage networks. This increased spatial
precision of rainfall estimates may improve hydrologic simulation capabilities. To fully use the NEXRAD
data for input, watershed models must include similar capabilities in simulating spatial variability. Grid-
based watershed models can provide more of a one-to-one linkage to spatially variable meterological data
for input. For lumped parameter models spatial data would need to be distributed by subwatershed.
Because the datasets involved are extremely large, techniques are needed to process and input data into
94
-------
either watershed-based or grid cell-based watershed modeling systems. The NEXRAD system has
functioned since 1988, but the number of stations and length of historic records for each station vary.
Stations are continually added to the system, but many of them are limited to more recent years. As a
result, use of NEXRAD data for model configuration also limits the period of simulation. Model
calibration may require simualtion of longer periods for analysis of model performance based on statistical
methods. Also, the use of NEXRAD data can confine simulation to recent years that do not indicate long-
term or critical conditions necessary for analysis of the TMDL. Therefore, rainfall gage networks are often
selected over NEXRAD data for calibration and validation of models. This limitation will diminish as
more data are collected over time.
• Meteorologic models can also be used to develop grid cell-based estimates of time variable meteorologic
inputs (http://www.waterboards.ca.gov/lahontan/TMDL/Tahoe/Tahoe Index.htm). Ultimately, input of
time varying and spatially detailed meteorologic information can support more accurate calibration and
application of watershed models, particularly in the prediction of hydrology. Hydrology is particularly
sensitive to variations in the spatial distribution of precipitation and temperature.
• Grid-Based Models. Most traditional watershed models are built on a lumped parameter network of
subwatershed units and stream reaches. The hydrologic connectivity between these components is used to
route runoff through the system. This type of configuration is appropriate for the prediction of downstream
hydrographs and is particularly well suited to systems that have a well-defined hydrologic network. Some
limitations of this formulation are the necessary "lumping" of heterogeneous spatial information (e.g., soils,
slopes, land use) into virtual land units within each subwatershed. Grid cell-based models have potential
applicability in areas with more complex or low gradient hydrologic systems, areas with highly
heterogeneous soils or land use practices, and areas with surface-to-groundwater interactions (i.e., high
water tables). Development and testing of grid cell-based models, and improvement in the development of
input data and computational efficiency, are needed to bring this class of models into practical application.
Linkage of the hydrologic processes to water quality simulation is needed to provide the water quality
analysis needed for TMDLs.
• Surface-Groundwater Models. Currently, watershed-scale simulation is largely confined to conceptual
and pseudo-distributed surface water models that are poorly linked to subsurface models or have highly
simplified representation of subsurface storage and transport. Traditional watershed models have used
groundwater components to calculate stream baseflow. The groundwater simulation was not oriented to
addressing dynamic water tables, changes in baseflow due to groundwater withdrawals, or more dynamic
interactions between surface and groundwater systems. As the realization increases that groundwater and
surface water are closely interconnected and need to be thought of as one hydrologic system, studies of
their conjunctive use and management are also increasing. If one or the other system is modeled
independently, a technique must be found to represent changes in the other system in the model, but such
techniques usually have serious limitations related to reconciling the scale and averaging across the
interface. A more accurate approach is to model the systems as a single integrated system. This approach
can account for process changes in both the groundwater and surface water systems and their mutual
interaction as such changes occur. Presently, there are watershed and groundwater models that solve
various components of the hydrologic cycle. However, there is no single model that solves the entire
hydrologic cycle in satisfactory detail. Nor are there any models that include dynamic transport and fate of
distributed pollutants in both surface and subsurface regimes.
In many areas with low gradients, hydro-modification, and high water tables, the surface and groundwater
systems are tightly and dynamically linked. Evaluation of management and source loading implications, in
these cases, requires consideration of groundwater and surface water interactions. For example, in many
areas of Florida, the management of water quality in canal systems requires concurrent and linked
evaluation of water table elevation, groundwater quality, canal and lock operations, evapotranspiration, and
rainfall-runoff processes. Although a number of detailed watershed hydrologic models and groundwater
models are available, they are not dynamically linked. Some applications have used a linkage between
HSPF using variable water table options and the USGS's MODFLOW. Few models are designed with this
capability, and the currently available models are proprietary (e.g., MIKE SHE). Continued research and
testing of linked surface and groundwater models for areas with high water tables and complex hydrology
95
-------
are a significant need. Enhancement, linkage, or new development of public domain or open source models
are also encouraged.
Among the numerical methods for solving hydrology and transport equations, the finite difference method
seems to be more popular for the ease of implementation and its relative simplicity. However, the finite
element method has potential for addressing more complex geometries and should be further investigated
for addressing hydrologic simulation.
In general, watershed hydrologic modeling research must continue to modify and improve existing algorithms to
more accurately account for physical processes, especially as affected by altered hydrology, sediment erosion,
interaction with groundwater, and watershed restoration alternatives, such as BMPs and constructed or restored
wetlands. Watershed modeling's future is very strongly linked to investigating interactions between runoff processes
and chemical and biological processes, which are crucial for water quality predictions (sediments, nutrients,
contaminants, etc.). One solution is to develop a hybrid hydrologic model, in which an adaptive selection of
kinematic, diffusive, or dynamic wave approaches can be made over all ranges of slopes in a watershed. The
hydrologic models must be linked to dynamic contaminant transport in the overland planes, transport in the rivers,
vertical transport in the unsaturated zone, and multidimensional transport in the saturated aquifer zone.
Sediment Loading
Watershed models have been extensively used to estimate overland flow or runoff sediment loads to surface
waterbodies. For example, HSPF and LSPC essentially predict a sediment mass loading time function
corresponding to the flow hydrograph for each subwatershed or watershed unit. For pervious surfaces, the empirical
formulations incorporating gross watershed characteristics, such as land use, slope and antecedent conditions, are
used to predict the mass loading, but for impervious surfaces, a particle buildup and washoff approach is used.
Other models, such as SWMM, SWAT, and GWLF, use formulations based on various versions of the USLE to
generate sediment loading as a function of soil, slope, and precipitation characteristics. Some models specialize in
sediment loading and include more detailed spatial heterogeneity such as KINEROS, GSSHA, and WEPP.
Linking or transferring the sediment mass loading time function from the watershed to the receiving water model
generally requires user definitions of how the total sediment mass loading is distributed among sediment type
classes simulated in the receiving water model. Both the empirical approach for total loading and the necessity of
user intervention to define the sediment type distribution for linking with the receiving water model pose limitations
on predictive ability. Similar to many other modeling needs, improvement in sediment calculations is needed across
the interface between watersheds and receiving waters, including dynamic changes in head-cutting of channels, rill
erosion, stream bank erosion, and riparian zones. Most often, the calibration process for linked watershed-receiving
water sediment modeling requires that the watershed component be interactively calibrated with the receiving water
model, with the watershed loading as a primary calibration parameter for the receiving water models. Because the
physics of sediment transport on impervious surfaces and saturated pervious surfaces in the watershed are the same
as those in stream channels, there is considerable opportunity to improve the ability of watershed models to predict
both the total mass and size distribution of sediment loads by the incorporation of a higher level of physics with
more quantitative information regarding soil types.
Research needs vary depending on the application type. For large multipurpose applications, often including
multiple endpoints, enhancement of traditionally used watershed models is needed. For more specialized
applications focusing on sediment management, enhancement or development of sediment modeling systems is the
focus. The following are major research needs for simulation of sediment loading:
• Enhancement of HSPF/LSPC Sediment Simulation. HSPF/LSPC have been used extensively for
TMDL development, including many applications where sediment and sediment-associated pollutants are a
concern. Improvement of the sediment simulation capabilities in these particular models would be of great
benefit to many large-scale watershed and TMDL studies. Use of kinematic and diffusive wave theory
incorporating vegetation resistance slope variations in a watershed unit would improve estimates of flow
stress responsible for surface erosion. Likewise, a better quantification of soil properties would allow the
96
-------
differing processes responsible for cohesive and noncohesive resuspension to be represented. These two
improvements would contribute significantly to the ability of a watershed model to predict sediment
loading by sediment type. HSPF/LSPC would also benefit from an alternative formulation that would
allow use of USLE-type erosion algorithms (rather than the Negev model). This would facilitate the
derivation of parameter values directly from USD A soils databases.
• Refine SWAT Modified Universal Soil Loss Equation (MUSLE). Review, validation, and refinement of
SWAT MUSLE implementation are needed. The code appears to contain a units error for MUSLE, and the
implementation for Hydrologic Response Units containing impervious land can be improved.
• Update SWMM Sediment Simulation. SWMM is also a commonly applied watershed model that would
benefit from an improved sediment simulation capability. This improvement would enhance the ability of
SWMM to be used in mixed land use watersheds. The current SWMM system uses the original USLE
formulation parameterized based on land unit characteristics. SWMM's current approach could be updated
to the more recent formulation of MUSLE with the ability to set land unit-specific parameters. Additional
improvements could be achieved by allowing for seasonal or year-to-year variations in settings.
• Overland Sediment Transport. Models for describing sediment transport in shallow overland flow need
improvement. Currently used equations were developed for channel flow conditions, which differ
significantly from shallow, overland flow conditions.
• Distributed or grid-based modeling systems. Additional research, development, and applications of grid-
based hydrologic and sediment modeling systems are needed. These models show promise for detailed
spatially heterogeneous applications, can link more effectively with receiving waters and riparian areas, and
can be used to evaluate a variety of spatially detailed management techniques. Further research is needed
to demonstrate their efficacy and demonstrate their use in TMDLs.
Pathogens
Pathogen sources, transport and behavior in water (e.g., growth and decay) are represented in general forms in
several models including HSPF, LSPC, and SWAT. Pathogen predictions, in particular fecal coliforms and E. coli,
are extremely variable and unreliable. Localized sources (e.g., wildlife) can significantly effect pathogen modeling
results. New genetic identification techniques have been used to identify specific sources present in discrete
samples; however, analysis of discrete and limited samples is insufficient to describe the dynamic loading of
pathogens from multiple sources. A complicating factor is the use of new indicator organisms for pathogen
presence including E. coli and enterococci. For example, Georgia's Environmental Protection Department recently
added E. coli in upland streams and enterococci in coastal waters to the existing fecal coliform bacteria criteria.
New indicators have additional limitations resulting from the lack of data, historic records, and modeling
experience. Several specific recommendations below suggest enhancements or alternatives to current modeling
techniques.
• Statistical Modeling of Pathogens. Guidance and additional techniques for modeling pathogens using
statistical techniques are needed. Building statistical models that associate sources or localized loading
potentials could help support evaluation of management alternatives. Simple spreadsheet tools could be
developed to facilitate analysis.
• Guidance. Examples, guidance and applications of modeling E. coli and enterococci should be developed.
An expanded dataset and compilation of available source loading and die-off characteristics would assist in
parameterizing models. Increased data collection will assist in developing calibrated applications.
• Genetic Tracer Analysis. Genetic source typing can provide a discrete representation of the sources
present at a particular location. Guidance and examples are needed on how to link genetic source typing
information with dynamic modeling applications.
• Growth and Die-off Rates. Models typically represent bacteria behavior by using a first-order decay term.
However, in many systems, bacteria appear to die-off or regrow depending on environmental conditions.
97
-------
Development of second-order equations or functional relationships that more accurately represent bacteria
growth and die-off rates would significantly improve modeling accuracy. The regrowth potential is of
particular concern in coastal areas with shellfish beds and beaches.
• Shellfish Areas. In tidally influenced areas, often located in the vicinity of shellfish beds and beaches,
specialized modeling techniques are needed to evaluate loading and associated pathogen counts. The
ability to comply with water quality standards for pathogens in tidal areas strongly correlate to the tidal
circulation and configuration of the shoreline. Areas with poor flushing potential are particularly prone to
high pathogen counts, in some cases due to highly localized sources. Some options proposed for
simulation of these tidal areas include linkage of watershed models such as HSPF and LSPC to the Tidal
Prism Model. Other techniques have included simplified loading estimates using monitoring data or
genetic tracer information in combination with receiving water models such as the Tidal Prism Model.
Additional research is needed to better characterize sources and develop protocols for linking monitoring
with models.
Nutrient Loading Simulation
Watershed models are used to develop estimates of nutrient loading to receiving waters. For watershed management
or nutrient TMDL development, the simulation might need to provide sensitivity to changes in land use practices
such as fertilizer application, tillage, and land cover management. Several existing models provide the ability to
track nutrient applications, crop uptake, and nutrient processes in soils. However these capabilities are traditionally
available only in agriculturally oriented models, such as SWAT. Nutrient loading is also an issue in mixed land use
or urbanizing areas. Comprehensive approaches are needed to evaluate changes in land use and the management of
nutrients in urban areas. In addition, because atmospheric deposition is a significant source of nitrogen, models
should provide the ability to track atmospheric sources. The following areas of research or enhancement of existing
models are recommended:
• Mixed land use watersheds. Although algorithms are available to track accumulation, uptake, and
washoff of nutrients from land areas, these techniques are not uniformly available for mixed land use
watersheds. One of the nutrient loading-related research needs includes providing and facilitating nutrient
simulation in mixed land use watersheds.
• Urban Area Nutrient Simulation. Urban areas can be significant sources of nutrients from fertilizer
application and pet waste and atmospheric deposition. Most urban models have limited representation of
the accumulation and transport of nutrients. Improving the simulation of nutrients in urban areas could
facilitate the exploration of management alternatives and the benefits of various education, street sweeping,
and infiltration or impoundment practices.
• Atmospheric Deposition. Most watershed models estimate nutrient availability through pollutant buildup
and washoff (e.g., SWMM), concentrations in runoff and soils (e.g., GWLF), or more detailed application,
transformation, and washoff (e.g., SWAT). Most models do not include atmospheric deposition as a
discrete source. For large-scale simulation of the Chesapeake Bay, HSPF includes a discrete term for
atmospheric deposition. The ability to separate atmospheric nutrients as a source is recommended for
nutrient budget development and alternatives analysis.
• Nutrient Transport Simulation. Simulation of nutrients should consider the linkage to various pathways
through overland flow, infiltration to groundwater, or across riparian buffers. There is also a need to
improve the simulation of nitrogen and phosphorus transformation in overland flow. Development of
improved nitrogen accounting models to allow better simulations of nitrogen transformations and
availability of various nitrogen species in surface-groundwater systems is needed as well. In addition,
current models could be improved by adding or facilitating the simulation of nutrient capture and
transformation in riparian areas (i.e., filter strips, riparian zones).
98
-------
Mercury
Mercury simulation continues to challenge modelers due the complex chemical processes and atmospheric source
loading. Many model applications are limited by field data to characterize toxic and mercury contamination levels
in the sediment bed, which generally are significant sources of water column contamination. Field collection of
such data is often beyond the resources of many model applications. Also, nonpoint sources and nonpermitted point
sources of contamination are difficult to determine. Improved model parameter estimation procedures offer promise
in estimating bed contamination levels and source locations and magnitude for data limited applications. EPA has
supported the development of simplified modeling systems (e.g., Mercury Tool in the TMDL Toolbox), and
continued enhancement of this system is ongoing. The most detailed modeling techniques are not in the public
domain. Detailed models are considerably more data intensive and difficult to apply. The mercury case study in
Chapter 6 reinforces the need to continue research and development in mercury modeling, including continued
improvement in detailed models, development of rates for mercury methylation, model testing, and linkages
between models. Specific recommendations for improving mercury modeling include:
• Detailed Mercury Models. A public domain or open code version of the MCM model should be made
available, and a user interface and modeling techniques to facilitate the application of the system to
decision-making should be provided. The more detailed description of chemical processes should be
included in a wider distribution of modeling systems.
• Fire Mobilization. Further research is needed in the role of fire in mobilization and transport of mercury
from the watershed. Recent TMDL development in New Mexico suggests this is a significant factor in the
watershed mercury budget in the arid West.
• Mercury Methylation. Improved methods are needed to estimate rates of mercury methylation in the
stream network (riparian wetlands, hyporheic zone, etc.). Research is needed to improve simulation of
mercury methylation in the water column (e.g., at metalimnion boundary) rather than bedded sediment,
which may be a dominant process in some turbid western lakes.
• Model Testing. New evaluation and testing of mercury transport models on detailed, radio-label studies
(e.g., METAALICUS) could assist in verifying models. To further refine models, the development and
statistical analysis of large cross-sectional databases are needed for correlation between mercury
methylation rates and other environmental variables and for correlation between methyl mercury
concentrations and biotic accumulation rates. Many assumptions are built into various models, but further
refinement and validation are needed.
• Organic Mercury Compounds. Model capabilities need to be improved to simulate other organic
mercury compounds (e.g., mercuric acetate) derived from anthropogenic sources.
• Model Linkages. Model linkages and examples of integrated atmospheric, watershed and waterbody
models for mercury are recommended. Support is also needed to develop and link air models to watershed
models.
• Mercury Snowpack Modeling. Evaluation of storage, transformation, and release of atmospheric mercury
in snowpack is needed for simulation in cold climate regions.
Other Pollutants and Toxics
Other pollutants and toxics addressed by models can include: PCBs, DDT, Dioxin, other pesticides, and heavy
metals (i.e., copper, zinc, selenium). Chlorides, although naturally occurring, can be a significant pollutant if
accumulated in toxic amounts. Various toxics, such as PCBs, DDT, and dioxin, are rarely included in watershed
loading models due to lack of active sources and limited data characterizing their availability, sources, or behavior.
Some actively used pesticides (e.g., atrazine) could be modeled as toxics; however, even these materials are rarely
modeled on a larger scale because specific data on application rates, decay, and transport behavior are limited.
Some models have the flexibility to be parameterized to include soil adsorbed and dissolved concentrations of a
variety of pollutants and toxics, so simulation is possible when needed. Generally required are better techniques for
99
-------
evaluation of sources, setup and parameterization of models, and examples for application techniques. More
information on rates, constants, and modeling techniques is needed. One frequently observed need in TMDL
modeling relates to problems associated with irrigated areas and chloride and selenium accumulations.
Chloride and Selenium
Areas of the country affected by irrigation and drainage practices may have elevated concentrations of chlorides and
selenium. These elevated concentrations have resulted in water quality impairments and the need to assess
management alternatives. Modeling of these conditions is rather complex and can require a combination of surface
and groundwater simulation, examination of irrigation and crop management practices, and leaching of pollutants
from soils. The first major limitation of these studies is the lack of reliable, public domain systems for simulation of
surface groundwater interactions. Ultimately, more robust hydrologic models should include the chemical processes
and accounting for the transport of pollutants (e.g., chloride, selenium, and others) through multiple pathways.
Management Practice Simulation
Management practices can include a combination of landscape management activities (e.g., fertilizer application),
impervious area reduction or control, and various structural management techniques (e.g., detention ponds). Point
source controls can include reduction or removal of discharges under various treatment technologies. Nontraditional
point sources can include urban areas under stormwater programs. Stormwater management, including both
structural and nonstructural management techniques, is needed to achieve water quality improvements under
stormwater program initiatives or for impaired areas to meet TMDL requirements. Implementation of TMDL
allocations will typically require a combination of point and nonpoint source control practices.
For TMDL development, models are applied to represent various levels of point and nonpoint source control
sufficient to meet water quality improvements or load allocations. Models of receiving waters (rivers, lakes, and
estuaries) can typically evaluate impoundments (e.g., ponds, reservoirs). Many models do not explicitly include
representation of management practices. For TMDLs, detailed implementation planning is not required, and the
limitations of management simulations are not an initial concern. River, lake and estuary models typically accept
input information on discharges from point sources or upstream tributaries (including point and nonpoint sources).
As discussed previously, many models represent load reduction alternatives through a generalized percent reduction
from individual sources and point source dischargers. However, a more detailed analysis of management options
requires models that support simulation of practices (or groups of practices), evaluate the implications of BMP
location, and allow for examination of management alternatives. As more TMDL studies result in implementation,
the use of models for management planning and alternatives analysis is increasing. Evaluating management
alternatives and considering financial investments will need more sophisticated BMP modeling systems.
• Large Scale Watershed Allocation Strategies. Continued development of tools that support large-scale
TMDL allocations at the subwatershed/source scale is needed. For example, the LSPC system includes
techniques for sequential allocations to multiple subwatersheds and sources and has been successfully
applied in systems with thousands of subwatersheds. Additional tools that provide optimization or cost-
related analysis and various alterative allocation techniques could facilitate development of cost-effective
and user-supported allocations.
• Comprehensive BMP Simulation Systems. For implementation planning, modeling systems are needed
that seamlessly provide support for watershed loading analysis, subwatershed and source level allocations,
and BMP placement and analysis. The majority of traditional BMPs (e.g., detention ponds, infiltration
techniques, conservation tillage) can be modeled by one or more systems. However, the practical
application of these techniques on a large-scale watershed-based application is limited. Most models that
include BMPs are not easily applied and require significant expertise in setup and adjustment of the various
BMP components, resulting in uneven levels of detail and incomplete application for management
planning. New comprehensive systems are needed, or existing watershed models should be enhanced to
provide the capability.
100
-------
BMP Simulation Techniques. Multiple techniques are available to simulate individual BMPs. Most rely
on a combination of infiltration, deposition/settling, and first-order decay. For example, SWMM includes
techniques for the simulation of detention ponds that incorporate these techniques. Agricultural models
include representation of land practices such as conservation tillage and crop rotations. Continued research
is needed into the transformation and transport of pollutants in BMPs. In particular, most BMP simulation
techniques provide limited representation of nutrient transformation. A more physically based approach
could improve understanding of the nutrient removal potential and the movement of dissolved and
adsorbed forms through one or more BMPs.
Riparian Areas. Riparian zones and stream buffers filter runoff and remove sediment and pollutants.
Riparian zones also provide shading with potential reduction of in-stream temperatures and improved
aquatic habitat. Typical watershed models have limited capabilities for simulation of the benefits of
riparian zones on water quality. The HSPF model has the capability of simulating land-to-land transfer,
which can be used for limited riparian area simulations. Specialized models such as REMM can be used to
evaluate individual riparian areas, but research is needed to provide practical techniques that can be used to
represent the benefits of riparian zones on a watershed scale.
Bioretention Simulation Techniques. Bioretention and infiltration practices are increasingly used in
watershed retrofit or new development applications. Bioretention is a commonly used practice in the
application of Low Impact Development procedures or in areas where small-scale and distributed
management techniques are employed. Several models include techniques for simulation of infiltration-
type practices; however, more detailed simulation of evapotranspiration as a function of land cover types
and nutrient/chemical processes are needed. Techniques or monitoring data are needed to evaluate the
efficacy of small-scale bioretention practices on watershed scale.
Economic Analysis of Management Strategies. TMDL allocations are typically selected to provide an
equitable distribution of the load reduction that results in meeting water quality standards. Improved
techniques are needed for evaluating the economics of allocation strategies and tradeoffs between
management techniques. Improved economic analysis tools could result in lower cost allocations and more
efficient use of resources. Analysis tools could include cost models, databases of management practice
cost, or optimization tools. Research is ongoing to develop optimization tools that support placement and
selection of BMPs and allow users to select water quantity, quality, and cost-related objectives for
optimization (Lai et al. 2003).
Demonstration of Management Success. As more TMDLs are implemented, it will be important to
document and demonstrate the reduction of loads and progress or achievement of meeting water quality
standards. To demonstrate progress, monitoring of the baseline conditions is needed with follow-up on
management practice adoption and continued monitoring of the watershed.
Management Practice Information. Datasets that compile information on location, type, and
characteristics of structural and nonstructural management practices should be developed and maintained.
For traditional point sources, such as industrial discharges, states and EPA maintain detailed databases of
permit numbers, location, and discharge monitoring records. However, for nonpoint source and stormwater
BMPs, the estimation of load reduction opportunities is often limited by a lack of consistent information on
management practices. If information is available that includes the areas served by BMPs, models or
simplified accounting techniques can be used to evaluate potential progress made toward load reduction
targets and the potential for additional management.
Hydrodynamics
Hydrodynamic modeling of surface water systems is very mature. Model physics is well established, and three-
dimensional model applications are now the norm for coastal, estuary, lake, and reservoir applications. The major
challenge in hydrodynamic modeling tends to be model run time, as grid resolutions become finer and model
simulation periods increase. Key recommendations include:
101
-------
• Models and systems that support parallelization of hydrodynamic model codes are essential to address
demands of finer spatial resolution and extended simulation periods.
• Additional software is needed to facilitate model input data preparation and display and animation of output
data. Although propriety systems have been developed, a broader distribution and support of public
domain systems are needed. The EPA TMDL Toolbox includes a public domain EFDC interface.
Continued development of user support tools is needed.
Sediment Transport
Sediment transport remains the major unsolved problem in hydraulic engineering and environmental fluid dynamics.
With respect to clean and contaminated sediment TMDL development and contaminated sediment remediation
studies at Superfund sites, predicting the source of in-stream sediment and associated sorbed contaminants and their
ultimate fate remains a particularly difficult problem. Sediment sources include erosion or resuspension of the in-
stream sediment bed, bank erosion including mass failure during high flow events, and sediment delivered via
overland flow from adjacent watersheds. Recommendations for improving source loading capabilities were
discussed above. Key recommendations for improving sediment transport capabilities are provided below.
• Laboratory and Field Research. For resuspension and erosion of in-stream beds, fairly reliable
experimentally derived relationships are available to parameterize resuspension under homogeneous bed
conditions where either cohesive or noncohesive sediment classes dominate a system. Garcia and Parker
(1991) provided a critical evaluation of the predictive ability of widely used formulations for noncohesive
sediment resuspension, and their findings remain valid today—15 years later. Recent formulations for the
resuspension of cohesive sediment are exemplified by the work of Sanford and Maa (2001) and Lick and
McNeil (2001). However, most natural surface water systems are characterized by heterogeneous sediment
bed mixtures. For example, relationships for noncohesive sediment resuspension typically fail to be
predictive when the cohesive fraction approaches 10 percent. Because laboratory results necessary to
parameterize resuspension of heterogeneous sediment mixtures are extremely limited—Gailani, et al.
(2001) being a notable exception—recourse must be made to expensive site-specific field resuspension
studies, which are well beyond the budget for many sediment modeling applications. Laboratory and field
research as well as accompanying theoretical studies are needed to develop heterogeneous bed resuspension
formulation for use in surface water models.
• Stream Bank Erosion - Simple Methods. Bank erosion is a significant source of sediment in some
stream systems and can be a source identified for load reduction in TMDLs. Techniques are needed to
evaluate the potential source as a function of local physical and hydrologic conditions. Existing watershed
models should be updated to include a stream bank erosion source to help account for the sediment sources
during calibration and for evaluation of allocation alternatives. In addition, information or techniques that
relate management actions (i.e., reduced imperviousness, riparian buffers) to the potential reduction of
stream bank and channel erosion are needed to demonstrate potential restoration.
• Laterally Averaged Bank Erosion Techniques. Bank erosion is likely the primary source of excessive
in-stream sediment levels and contributes to event-driven redistribution of contamination. Several
sediment transport models incorporating mass failure erosion of cohesive sediment banks have been
developed using the bank stability approaches (Darby and Thorne 1996b). They include the model
reported by Darby and Thorne (1996a) and the USDA-ARS CONCEPTS model (Langendoen 2000), both
based on one-dimensional longitudinal hydrodynamic and sediment transport. The CONCEPTS model
incorporates a piecewise description of the channel bed perimeter with the steeper side portions of the
perimeter representing the banks. Although appropriate for many applications, the averaged cross-sectional
approach is a simplification of bank erosion and overbank processes. Providing a more efficient approach
for application of CONCEPTS would assist in broader application of the technique. CONCEPTS has been
applied with AGNPS and is included in the EPA TMDL Toolbox. Developing a linkage between
watershed models (LSPC, HSPF) and the CONCEPTS model would provide additional utility and more
opportunities for application.
102
-------
• Multidimensional Sediment Transport. Multidimensional evaluation of cohesive and noncohesive
sediment transport is needed. This formulation can be used to evaluate contaminated sediment transport in
complex river systems. Continued development of solution techniques, sample applications, and
supporting software is needed. Further research is recommended to test and validate physically based bank
erosion formulations, such as those in the EFDC model, and to provide procedures for their application to
data limited sites.
• Turbidity. The ability to predict turbidity (an optical property) from inorganic sediment, algae, and DOC
concentrations should be improved and tested.
Nutrients and Eutrophication
Nutrient cycling and eutrophication dynamics in natural waters is fairly well understood. The major challenge in
nutrient and eutrophication modeling is calibration or estimation of kinetic coefficients and quantifying uncertainty
in model predictions. Optimization is a possible technique to address this challenge and is discussed further in
Model Defensibility. In addition, selected areas of research and improvement can broaden the applicability of
models to various algal species and eutrophication processes. Evaluation of nutrient processes in the hyporheic
zone, active areas at the interface between receiving waters and riparian or stream beds, could expand applicability
of models. For the Cahaba nutrient case study the Toolbox provided many of the needed tools for building the
modeling applications, assessing the various nutrient concentrations and evaluating the algal response. However,
the ability to model benthic algal and periphyton is still limited and relatively untested. Specific areas for additional
research in nutrient and eutrophication techniques include:
• Benthic Algae. Benthic algae can be a significant source of impairment in streams. Some limited
modeling of benthic algal response in streams has been developed (e.g., QUAL2K), and users have
customized CE-QUAL-W2 and WASP to estimate benthic algal growth. However, simpler, empirical
methods would be useful additional tools for analysis of streams. To date, these simple methods do not
adequately account for effects of scour, which appear to be a dominant control in many streams. Biggs'
New Zealand work showed that including "days of accrual" as a measure of scour frequency significantly
improved ability to predict benthic algal density from nutrient concentrations; however, the work in this
wet climate does not appear directly transferable to Mediterranean-type hydrologic regimes (Biggs 1988).
• Macrophyte. Macrophyte processes and hydrodynamics interaction simulation technology are needed for
riverine and lake systems. The ability to simulate macrophyte growth and submerged aquatic vegetation as
a function of nutrient loads is needed for many TMDLs. A more rigorous and predictive formulation
should be developed to account for the interactions between macrophytes and hydrodynamics. By
predicting not only the mass of macrophyte growth but also the height and volume, the effect of
macrophytes on dissolved oxygen, pH, and water circulation patterns can be evaluated.
• Evolution of Macrophyte/Periphyton. One potential technique to evaluating long-term growth of
macrophyte and periphyton is to use the Cellular Automata technology with traditional hydrodynamic and
water quality modeling technology to simulate evolution of macrophyte and periphyton over multiple years
(Marsili-Libelli and Giusti 2004). This technology could be extended to simulate aquatic habitat variations
over time.
• Buoyancy-compensating Algae. The ability to simulate buoyancy-compensating algae (primarily
cyanophytes) in lakes, integrated with hydrodynamic simulation, could assist in better simulation and
calibration of local conditions. Some algae move according to light and local conditions. This process can
affect accuracy of the vertical layers in the simulation model.
• Bacterial-Nutrient-Algae Interactions. Bacteria can be associated with algal processes. Storage and
regrowth of bacteria can also be linked to algal processes. Additional research and development of
simulation algorithms are needed to support linked evaluation of bacteria and nutrients.
103
-------
• Wetland and Riparian Zone Processes. Enhanced procedures are needed to address the water quality
processes that occur at the interfaces between wetland and riparian zones and receiving waters. Because of
the extensive interactions between these systems multimedia modeling capabilities are needed. For
example, EFDC includes capabilities for addressing wetting and drying of tidal wetland areas. Further
enhancement of these capabilities and the exchange of nutrients could be beneficial, especially for studies
that include preservation or enhancement of riparian areas.
• Nutrient and Eutrophication Endpoints and Targets. Closely related to modeling of nutrient and
eutrophication processes is the derivation of the specific indicators or thresholds used to evaluate
compliance with water quality standards. Some TMDLs are directly keyed to a numeric dissolved oxygen
standard. However, in many rivers and streams, the numeric measures of eutrophication effects are not
well defined. Additional research is needed to develop locally derived effects-based targets for sediment,
nitrogen, and phosphorus.
• Monitoring to Support Periphyton and Macrophyte Modeling. As new modeling techniques are
developed for periphyton and macrophytes, additional guidance is needed on monitoring techniques, field
procedures, laboratory analysis techniques, and input parameter development.
Ecological/Habitat
Ecological modeling can provide simulation of conditions that relate to key indicators of designated use support.
For example, an ecological model could predict fish propagation as a function of habitat conditions. Ecological
models can be used to evaluate response of aquatic life to elevated concentrations of toxics or the effects of
bioaccumulative substances. Other systems can evaluate response to habitat conditions, including shading,
temperature, and sediment, on fish. Application of ecological models may go beyond typical TMDL development
needs and address potential changes in water quality criteria or direct interpretation of designated uses or narrative
criteria. Key areas where new or enhanced ecological modeling techniques are needed include:
• Impervious Areas Impacts on Habitat. The ability to link changes in storm hydrographs to changes in
habitat quality and benthic biota response is needed. Impervious areas and associated hydrologic effects
are widely believed to be associated with aquatic health impairments. However, the response to changes in
imperviousness is not easily related to specific, measurable endpoints representative of aquatic health.
• Dissolved Oxygen Criteria. Evaluation of water quality criteria for dissolved oxygen may require the
assessment of the effect of low dissolved oxygen on fish. Improved ecological modeling tools could
support the development of site-specific criteria, where appropriate.
• Bioaccumulation of Toxics. In areas with contaminated sediment or excess loading of bioaccumulative
toxics, additional modeling tools could be used to evaluate bioaccumulation rates. Food chain models can
be used to evaluate bioaccumulation under various loading scenarios.
• Habitat. Techniques are needed to simulate the effects of stream channel habitat indicators on the aquatic
habit, including stream shading, fine sediment substrates, and flow frequencies. Traditional approaches use
a weight of evidence process and statistical analysis to evaluate the response of aquatic life to a variety of
habitat indicators. Modeling systems could provide enhanced multimedia analysis techniques to integrate
the various habitat indicators, predict response to land use changes, and evaluate the potential effects on
aquatic life.
Data
Modeling and environmental analysis requires data to apply models. Sufficient data are needed to verify that
models are performing appropriately and build confidence in model predictions. Typical data needs include spatial
coverages (e.g., soils, topography, land cover), water quality monitoring, point source discharges, management
activities or structures, and land use practices. Current state and local data collection studies need to continue to
104
-------
support the development of comprehensive and long-term monitoring records. As watershed model sophistication
increases, the demand for more frequent and higher resolution data will also increase to setup, calibrate, and validate
models. The challenges faced in watershed modeling have reflected the need to deal with spatial variability while
considering the linkages among climatology, hydrology, biology, and geochemistry. The more physically based
models require significant information on meteorology and chemistry. Some of the most important advances in
watershed modeling during the past decade have involved approaches that employ GIS and remotely sensed data.
The availability of GIS coverages and remotely sensed data has stimulated model development and facilitated the
representation of landscape heterogeneity. Continuous improvements in the ability to process very large distributed
sources of remotely sensed and space-based hydrologic and climatic data, combined with advanced data assimilation
algorithms, should lead to benefits in both watershed modeling and TMDL studies. With the use of remote sensing,
radar, and satellite technology, the ability to observe data over large and inaccessible areas and to map these areas
spatially is vastly improved, making it possible to develop truly distributed models. Identified below are some of
the key areas where new research can facilitate data collection and improve the quality and comprehensiveness of
data.
• Remote Sensing. Remote sensing provides a technique for developing spatial heterogeneous information.
Key areas where remote sensing might improve data gathering include Moderate Resolution Imaging
Spectroradiometer (MODIS) satellite data to evaluate algae and turbidity in larger fresh waterbodies,
techniques to reconcile satellite-based land cover and parcel-based land use information, estimation of
directly connected impervious areas (DCIA), and estimation of soil moisture content. Remote sensing
shows promise for developing detailed representation of soil moisture content that can affect infiltration
and runoff calculations. NEXRAD data can provide more spatially detailed precipitation data. Spatially
detailed information could significantly benefit development and application of distributed (grid-based)
hydrologic models.
• Specialized Monitoring Guidance. Guidance is needed on how to collect data for complex environments
such as SOD, sediment nutrient fluxes, reaeration, photosynthesis and respiration, mixing zones, diurnal
dissolved oxygen, periphyton and macrophytes.
• Geomorphic Evaluations. Guidance is needed on how best to perform geomorphic assessments to
determine sediment impairment, evaluate sources and causes, and provide insights into management
techniques. Methods are needed to collect data that can support stream sediment transport and in-stream
sediment evaluations.
• Implementation Monitoring. Guidance is needed on monitoring approaches for measuring water quality
conditions in large watersheds and evaluating the performance of management practices. Guidance should
include targeted and probability-based sampling and management practice tracking techniques.
• Testing Data. Data are needed to verify models at multiple scales. Many watershed models use individual
land units as the fundamental simulation unit. Existing models were developed and tested initially on
small-scale test plots. The USLE was developed based on an extensive national network for test plots.
Historically test plot data have been collected to support modeling of forest and agricultural lands.
However, similar datasets are not available for suburban residential lands, which comprise significant land
areas considered in many TMDLs. This information can be used to build a library of loading rates and
EMCs. Additional data gathering is needed to describe the national variability in loading from urban and
suburban land areas. An approach is needed to improve access to existing and ongoing monitoring data
and standardize data collection approaches.
Model Defensibility
Model defensibility has generally involved two components—the defensibility of the model code or computer
program and the defensibility of the model application. Defensibility of the model computer program, termed
generic defensibility, is generally inferred from previous model applications, peer reviews, and the general
acceptability of the model with respect to regulatory agencies. Although these things are certainly an important
component of generic defensibility, they provide no guarantee that the model computer program is verified.
105
-------
Verification is here defined as determining if the model program or code is error-free and makes no judgment as to
whether the physical and biogeochemical process upon which the numerical algorithms in the program are based are
correct. Errors have been found and will continue to be found in widely accepted surface water modeling systems.
The emerging discipline of computational science and engineering has developed robust methodologies for verifying
computer models (Roache 1998; Knupp and Salari 2002) and these methodologies should be applied to major
surface water modeling systems. Generic model verification addresses issues associated with the correctness and
acceptability of the physical and biogeochemical process representation upon which a model is based and the
accuracy and robustness of the numerical algorithms used in the model code. Additional guidance is needed on
verification of computer models, and fundamental testing of publicly applied models should be encouraged and
documented.
The defensibility of each model application is evaluated and documented through a process of model calibration and
validation. The model setup requires parameter estimation as part of calibration and can include additional
sensitivity and uncertainty analyses. Calibration is the adjustment of model parameters to achieve an acceptable
level of agreement between model application predictions and prototype observations. Validation is the independent
evaluation of model performance without further adjustment of model parameters. A major concern in the
demonstration of model defensibility is the lack of consistent techniques, measures, and procedures. Both watershed
and receiving water modeling studies include ample examples of calibration and validation but the measures are not
consistent, and the "goodness of fit" required for various studies is not defined. In addition, the use and accuracy of
models for predictive simulations of future conditions are not well defined. Key considerations in using models for
alternatives and forecasts need to be identified and addressed in more specific guidance. Guidance is needed to
support consistent and robust approaches for evaluating model defensibility. Specific recommendations include the
following:
• Model Performance Criteria. Various model developers and reviewers have suggested model
performance criteria for hydrologic and water quality simulation of watershed and receiving waters.
Standardized statistical tests and application procedures are needed to provide a common context for the
evaluation of model performance. Compilation of a set of typical model performance criteria and measures
would help standardize model evaluations and documentation and provide additional confidence in model
predictions by stakeholders and federal, state, and local agencies.
• Model Calibration and Validation Guidance. Additional guidance is needed on model calibration
techniques and approaches. Guidance can help to standardize approaches, statistical techniques, and
appropriate techniques for rivers, estuaries, and lakes. Guidance should define techniques performing
uncertainty analysis that put "error bars" on model predictions. New guidance could build on the initial
development of generalized modeling guidance provided by CREM (USEPA 2003).
• Optimization Techniques for Model Testing. Optimization can provide a framework for evaluating
model calibration, parameters, and measures of sensitivity and uncertainty. A general class of
optimization-based parameter estimation procedures, including generalized software packages, is available
but has found limited use in surface water modeling with respect to both nutrients and toxics. Research to
develop and make available a number of optimization-based calibration and parameter estimation systems
compatible with surface water modeling systems is essential. Uncertainty analysis using existing methods,
including Monte Carlo and Latin-Hypercube, is rarely used due to the intensity of efforts required, or, when
they are used, only limited-duration model simulations are conducted. A new class of highly efficient
stochastic response surface methods offers significant advantages for uncertainty analysis and can be
incorporated in a streamlined manner with optimization-based model calibration. One approach is to
express the level of agreement as an objective function, such as the weighted space-time squared difference
between model predictions, and observations, at multiple locations and times, with respect to a set of model
parameters. The classical minimization methods require estimation of the gradients of the objective
function sequentially as the minimum condition is approached. These gradients in turn define the
sensitivity of the model's ability to predict prototype observations with respect to model parameters over a
range of parameter sets. Thus, the optimization approach to model calibration accumulates a significant
amount of information on sensitivity.
106
-------
The primary obstacle to employing an optimization-based model calibration-sensitivity analysis approach is
obtaining a sufficient number of estimations of the objective function gradients in model parameter space. The
gradient estimation approach includes brute-force parameter perturbation, tangent-linear sensitivity, and variational
adjoint methods. The later two are extremely promising but have received relatively little attention in surface water
modeling. Given probability distributions of model parameters, the probability distributions of the response surface
can be determined with significantly fewer model simulations than are necessary for Monte Carlo or Latin-
Hypercube uncertainty analysis. It is highly recommended that further research be directed to this potentially
powerful approach that combines model calibration, parameter estimation, sensitivity analysis, and uncertainty
analysis in a unified and potentially efficient manner.
Systems Development and Supporting Tools
Development and application of models requires trained and experienced modelers. The development of shared
resources, datasets, and guidance is essential to promoting the knowledge among the user community. Listed below
are the major recommendations for guidance and tools to support the modeling community. Specific technical
guidance is also recommended in the capabilities sections included earlier.
• Linkage Tools. Standards and software tools for model interlinkages and data transfer need to be
developed and improved. Key linkages that should be supported include watershed models to receiving
water models, such as: GSSHA/HSPF/LSPC/SWMM to EFDC/WASP or CE-QUAL type models.
• Universal Database Systems. Universal databases that include water quantity, quality, biological,
physical habitat, fish, and geomorphic data should be provided to support model development, test,
calibration, and validation. One option is to link Water Resources Database (WRDB) in the EPA TMDL
Toolbox with the Ecological Data Application System (EDAS).
• Rates and Constants Manual Update. Databases and guidance are needed on key parameters and
datasets used for initial setup and parameter selection for models. The "rates and constants" manual
(Bowie et al. 1985) has provided modelers with an excellent reference document. An updated system could
include an on-line searchable database and documentation. Significant experience and data gathering has
been performed during the past 20 years. Compilation of this new information would be a great service to
the modeling community.
• Application Library. Model applications either completed or under development can provide a useful
repository of knowledge and experience. A library of completed applications, provided on-line, could link
to project reports, model input files, parameters used, and documentation of model updates. This library
could link to available on-line reviews of models, such as CREM (http ://cfpub .epa. gov/crem/)
• Modeling Guidance. Additional guidance is needed on available models, model selection, calibration,
application techniques, and allocation procedures, including optimization. Model guidance could be
delivered through on-line materials, case studies, on-line or workshop training courses, and interim
technical notes.
• Training. Additional training is needed in watershed and receiving water modeling fundamentals,
watershed modeling techniques, and the application of multiple linked modeling systems.
Integrated Modeling Systems
Flexible and adaptable integrated modeling systems are needed that can address the many technical, data, and
systems recommendations identified above. These integrated systems, if developed with a universal format and data
exchange structure, could share information and modeling modules to achieve a variety of modeling needs.
However, in the current development of systems, there are still a multitude of standards, unique data structures and
formats, and disparate systems. EPA's BASINS and Toolbox systems share information and utilities; however, the
linkage between the various systems is not yet complete.
107
-------
Encouraging open architecture and modular modeling systems could facilitate the future development and
integration of models. The EPA TMDL Toolbox (http://www.epa.gov/athens/wwqtsc/Toolbox-overview.pdf) is
designed with an open architecture that could lead to the integration of the individual model components into a
variety of modeling systems. Adoption of the most flexible modeling systems, such as Precipitation-Runoff
Modeling System (PRMS) and MMS, will require continued development, demonstration applications, and more
rapid application timeframes. Training a broader audience in the development and application of modeling using
these tools will be needed before they will be widely adopted.
The continuous improvement in the science and the description of the physical processes is inconsistently distributed
and adopted by modeling systems. By providing on-line or desktop access to modular modeling components, new
algorithms can more easily be shared among applications. For example, by building on a hydrodynamic modeling
framework, an unstructured set of modules can be developed to simulate any number of species of phytoplankton,
zooplankton, macrophytes, or higher-level organisms, including fish. With a flexible set of modules, models can be
applied at various levels of complexity to address water quality problems and ecological system analysis without the
need for users to modify the code.
Modular modeling systems provide an opportunity to address many of the specific recommendations for expanding
technical capabilities of models. If a unified framework is developed, new models or algorithms can share data
management, post-processors, and analytical tools. In an Internet environment, the maintenance of systems and user
support techniques can be more centralized and potentially more efficient.
More aggressive development of Internet-based modeling systems is highly recommended. Internet-based systems
are fully possible with emerging software for GIS, visualization, and data management now widely available and
suitable for practical use. The maintenance and application of Internet-based software can significantly reduce the
distribution and management of software systems. Internet-based systems reduce the need to address compatibility
with multiple desktop software and hardware specifications. The systems can be used without requirements for
desktop proprietary software for GIS or database management. Software updates can be provided seamlessly
through a single copy, and model runtimes can be reduced by the use of parallel processing techniques.
Future systems can provide on-line access to stakeholders to evaluate alternatives and interactively examine
assumptions. The development of on-line modeling systems has another significant potential benefit in providing
public access, transparency of technical analysis and assumptions, and interactive interfaces for community
decision-making. An on-line system can provide a dynamic representation of alternatives where users can select
criteria and see alternative predictions of results (e.g., pollutant loading, measures of ecological conditions). For
TMDLs, systems that allow users to evaluate the load allocation alternatives or implications of various user-defined
choices can encourage more involved and proactive comment and agreement with selection of preferred allocations.
One example of an on-line user interface is an evaluation of land use patterns and growth developed by the Illinois
LEAM group (http://www.rehearsal.uiuc.edu/projects/leam/KaneFinal/Model6.html') for the Kane County, Illinois,
project. This example is specific to land use planning activities but provides an example of a format and approach
that could be used with water quality modeling approaches. These more transparent and inclusive modeling
approaches can significantly expand the acceptance of models by the community at large.
Conclusions
An understanding of current trends in technology and research can help increase an understanding of how modeling
might evolve and how to support the next generation of modeling systems. However, anticipating trends is a
"crystal ball" exercise, and sometimes adoption of technology can take surprising turns. This review concludes that,
although significant progress has been made in model development, major areas of research are still needed to
expand the capabilities, defensibility, and application of models. Research has an opportunity to capitalize on the
emergence of new data management and processing technologies (e.g., GIS, graphic user interfaces, data collection
techniques (e.g., remote sensing), and the burst in enhanced performance of modern-day computers and Internet
communications.
The research needs identified are diverse and consistent with TMDL and management needs that encompass a wide
range of sources, pollutants, and processes. The diversity of the needs is indicative of the current status of the
108
-------
TMDL program and environmental management applications. Over the past 10 years, development of analytical
tools has emphasized simulation of dominant pollutants (i.e., nutrients, sediment, metal, pathogens). More recently,
the emphasis has shifted to addressing a more diverse group of listed waters and areas with specialized problems.
Although some of these concerns may affect only a small group of waters, the analytical needs are still relevant.
Continued model development should encourage building linkages and multimedia models that address air, surface
water, and groundwater interfaces more accurately. Linkage of meteorlogic, atmospheric, and watershed models
can support the evaluation of potential long-term climate changes. Multimedia-linked models encourage a holistic
and integrated approach to water quality management that can result in improved decision-making. Grid-based
models show particular promise as a framework for linking multiple media in a more physically based approach. As
development of TMDLs continues, the most significant emerging need is the evaluation of management options and
selection and determination of optimal solutions. Ultimately, more comprehensive systems are needed that can
evaluate management options and solutions at multiple scales.
New and more flexible modeling systems and tools could support a diverse set of technical needs. Common
databases and GIS systems can support multiple solution techniques at various scale and levels of complexity.
Technical innovation can be encouraged by providing integrated systems and work environments that are flexible
and modular. These integrated systems can provide the commonly needed tools and support integration of new
solution techniques, source representation, and algorithms. In particular for BMP simulation, a flexible modeling
environmental that can incorporate new solution techniques will be beneficial. Providing integrated system
platforms can help minimize duplication of effort (shared on-line data management, data display, shared resources),
while maximizing resources for more fundamental development and research of key components.
The vision of a consistent, Web-based framework so far has proved elusive and difficult to implement. However,
continued rapid expansion of broadband accesses and Internet-based GIS and data management technology make
this vision more realistic. The use of Internet-based technologies has emerged as a viable and practical medium for
management of data, analysis techniques and tools to support TMDL and more generalized watershed analyses.
Development of a standardized Internet-based framework could provide significant cost saving for the management
and application of models. In addition, a standardized and open framework, with clearly defined linkage
capabilities, could encourage research and continual testing and update of new components.
Guidance and consistent metrics and methods for assessing defensibility of models and model predictions are critical
needs to support the reliable application of models and maintain user confidence in applications. Continued
emphasis on high-quality data collection and guidance on data collection methods that can support parameter
estimation and testing are also critical needs.
Future development of models and the supporting infrastructure of data and guidance will improve our ability to
support informed environmental decision-making, help improve understanding of the physical systems in our world,
and ultimately provide information to support the effective restoration and protection of the nation's waters.
109
-------
110
-------
References
ADEM. 2004a. Nutrient Target Development in Support of Nutrient TMDLs for the Cahaba River Watershed.
Alabama Department of Environmental Management, Water Division, Water Quality Branch, Montgomery, AL.
ADEM. 2004b. Draft Nutrient Total Maximum Daily Loads (TMDLs) for the Cahaba River Watershed. Alabama
Department of Environmental Management, Water Division, Water Quality Branch, Montgomery, AL.
Benoit, J.M., C.C. Gilmour, R.P. Mason and A. Heyes. 1999. Sulfide controls on mercury speciation and
bioavailability to methylating bacteria in sediment pore waters. Environmental Science and Technology 33(6):
951-957.
Bicknell, B.R., J.C. Imhoff, J.L. Kittle, Jr., A.S. Donigian, Jr., and R.C. Johanson. 1996. Hydrological Simulation
Program -FORTRAN, User's Manual for Release 11. U.S. Environmental Protection Agency, Office of Research
and Development, Environmental Research Laboratory, Athens, GA.
Biggs, B.J.F. 1988. Artificial substrate exposure times for periphyton biomass estimates in rivers. New Zealand
Journal of Marine and Freshwater Research 22: 507-515.
Bowie, G.L., W.B. Mills, D.B. Porcella, C.L. Campbell, J.R. Pagenkopf, G.L. Rupp, K.M. Johnson, P.W.H. Chan,
and S.A. Gherini. 1985. Rates, Constants, and Kinetics Formulations in Surface Water Quality Modeling. 2nd
Edition. EPA/600/3-85/040. U.S. Environmental Protection Agency, Office of Research and Development,
Environmental Research Laboratory, Athens, GA.
Darby, S.E., and Thorne, C.R. 1996a. Numerical simulation of widening and bed deformation of straight sand-bed
rivers. I. Model development. Journal of Hydraulic Engineering 122(4): 184-193.
Darby, S.E., and Thorne, C.R. 1996b. Development and testing of riverbank-stability analysis. Journal of
Hydraulic Engineering 122(8): 443-454.
Driscoll, C.T., C. Yan, C.L. Schofield, R. Munson, and J. Holsapple. 1994. The mercury cycle and fish in the
Adirondack lakes. Environmental Science and Technology 28(3): 136a-143a.
Dunne, T., and L.B. Leopold. 1978. Water in Environmental Planning. W.H. Freeman and Co., San Francisco, CA.
Gailani, J.Z., Jin, L., McNeil, J., and Lick, W. 2001. Effects of bentonite on sediment erosion rates. DOER
Technical Notes Collection (ERDC TN-DOER-N9). U.S. Army Corps of Engineers, Engineer Research and
Development Center, Vicksburg, MS.
Garcia, M., and G. Parker. 1991. Entrainment of bed sediment into suspension. Journal of Hydraulic Engineering
117:414-435.
Gilmour, C.C., G.S. Riedel, M.C. Ederington, J.T. Bell, J.M. Benoit, G.A. Gill, and M.C. Stordal. 1998.
Methylmercury concentrations and production rates across a trophic gradient in the northern Everglades.
Biogeochemistry 40 (2/3): 327-345.
Haith, D.A., R. Mandel, andR.S. Wu. 1992. GWLF, Generalized Watershed Loading Functions, Version 2.0, User's
Manual. Cornell University, Department of Agricultural and Biological Engineering, Ithaca, NY.
Ill
-------
Haith, D.A., and D.E. Merrill. 1987. Evaluation of a daily rainfall erosivity model. Transactions of the American
Society of Agricultural Engineers 30(1): 90-93.
Hewett, S.W., and B.L. Johnson. 1992. A generalized bioenergetics model of fish growth for microcomputers.
University of Wisconsin Sea Grant Institute, Madison, WI. UW Sea Grant Tech. Rep.WIS-SG-92-250. 79 pp.
Kalin, L., and M. Hantush. 2003. Evaluation of Sediment Transport Models and Comparative Application of Two
Watershed Models. U.S. Environmental Protection Agency, Office of Research and Development, National Risk
Management Research Laboratory, Cincinnati, OH.
Knupp, P., and K. Salari. 2002. Verification of Computer Codes in Computational Science and Engineering. CRC
Press.
Lai, F-h., Fan C., Shoemaker, L., and Fields, R. 2003. Development of a Decision-support Framework for
Placement of BMPs in Urban Watersheds. Water Environment Federation Specialty Conference TMDL 2003,
Nov 15-19, 2003, Chicago, IL.
Langendoen, EJ. 2000. CONCEPTS - Conservational channel evolution and pollutant transport system: Stream
corridor version 1.0. Research Report No. 16. U.S. Department of Agriculture, Agricultural Research Service,
National Sedimentation Laboratory, Oxford, MS.
Lick, W., and J. McNeil. 2001. Effects of sediment bulk properties on erosion rates. Science of the Total
Environment 266: 41-48.
Lindberg, S.E., R.R. Turner, T.P. Meyers, G.E. Taylor Jr., and W.H. Schroeder. 1991. Atmospheric concentrations
and deposition of Hg to a deciduous forest at Walker Branch watershed, Tennessee, USA. Water, Air, and Soil
Pollution 56: 577-594.
Marsili-Libelli, S., and E. Giusti. 2004. Cellular Automata Modelling ofSeagrass in the Orbetello Lagoon. IEMSS
2004 International Congress, "Complexity and Integrated Resource Management", Osnabrueck, Germany, June
2004.
Maxwell, T., 1999. A Parsi-model approach to modular simulation. Environmental Modeling and Software 14: 511-
517.
Maxwell, T., and R. Costanza. 1994. Spatial ecosystem modeling in a distributed computational environment. In
Toward sustainable development: Concepts, methods, and policy, ed. J.v.d. Berg and J.v.d. Straaten, pp. 111-138.
Island Press.
Maxwell, T., and R. Costanza. 1997. An open geographic modelling environment Simulation Journal 68(3): 175-
185.
NRC. 2001. Assessing the TMDL Approach to Water Quality Management. International Standard Book Number
0-309-07579-3. National Research Council, Division on Earth and Life Studies, Water Science and Technology
Board. National Academy Press, Washington, DC.
Pitt, R.A., A. Maestre, R. Morquecho. The National Stormwater Quality Database (NSQD, version 1.1). University
of Alabama, Department of Civil and Environmental Engineering, Tuscaloosa, AL.
http://unix.eng.ua.edu/~rpitt/Research/ms4/Paper/Mainms4paper.html
Roache, P.J. 1998. Verification and Validation in Computational Science and Engineering. Hermosa Publishers.
Sanford, L.P., and J.P.-Y. Maa. 2001. A unified erosion formulation for fine sediment. Marine Geology 179: 9-23.
Smith, R.A., Schwarz, G.E., and Alexander, R.B. 1997. Regional interpretation of water-quality monitoring data
Regional interpretation of water-quality monitoring data. Water Resources Research 33(12): 2781-2798.
112
-------
Tetra Tech. 1999. Dynamic Mercury Cycling Model for Windows 95/NT™, A Model for Mercury Cycling in Lakes,
D-MCM Version 1.0, User's Guide and Technical Reference. Electric Power Research Institute, Palo Alto, CA.
USEPA. 2003. Draft Guidance on the Development, Evaluation, and Application of Regulatory Environmental
Models. U.S. Environmental Protection Agency, Office of Research and Development, Office of Science Policy,
Council for Regulatory Environmental Modeling, Washington, DC. November 2003.
http://www.epa.gov/osp/crem/library/CREM%20Guidance%20Draft%2012 03.pdf
USEPA. 2002. Establishing Total Maximum Daily Load (TMDL) Wasteload Allocations (WLAs) for Storm Water
Sources and NPDES Permit Requirements Based on Those WLAs. Memorandum from Robert H. Wayland, III,
Director, Office of Wetlands, Oceans and Watersheds, and James A. Hanlon, Director, Office of Wastewater
Management, U.S. Environmental Protection Agency, Washington, DC. November 22, 2002.
USEPA. 2001. Protocol for Developing Pathogen TMDLs. EPA 841-R-00-002. U.S. Environmental Protection
Agency, Office of Water, Washington, DC. 132 pp.
USEPA, 2000. Nutrient Criteria Technical Guidance Manual: River and Streams. EPA 822-B-00-002. U.S.
Environmental Protection Agency, Office of Water, Washington, DC.
USEPA. 1999a. Protocol for Developing Nutrient TMDLs. EPA 841-B-99-007. 135 pp.
USEPA. 1999b. Protocol for Developing Sediment TMDLs. EPA 841-B-99-004. U.S. Environmental Protection
Agency, Office of Water, Washington, DC. 132 pp.
USEPA. 1997. Compendium of tools for watershed assessment and TMDL development. EPA841-B-97-006, U.S.
Environmental Protection Agency, Office of Water, Washington, DC.
USEPA. 1991a. Guidance for Water Quality-Based Decisions: The TMDL Process. EPA 440/4-91-001. U.S.
Environmental Protection Agency, Office of Water, Washington, DC.
USEPA. 1991b. Technical Support Document for Water Quality-based Toxics Control. EPA 505/2-90-001. U.S.
Environmental Protection Agency, Office of Water, Washington, DC.
Watras, C.J., N.S. Bloom, R.J.M. Hudson, S. Gherini, R. Munson, S.A. Claas, K.A. Morrison, J. Hurley, J.G.
Wiener, W.F. Fitzgerald, R. Mason, G. Vandal, D. Powell, R. Rada, L. Rislov, M. Winfrey, J. Elder, D.
Krabbenhoft, A.W. Andren, C. Babiarz, D.B. Porcella, and J.W. Huckabee. 1994. Sources and fates of mercury
and methylmercury in Wisconsin lakes. In Mercury Pollution: Integration and Synthesis, ed. C.J. Watras and
J.W. Huckabee. Lewis Publishers, Chelsea, MI.
Wischmeier, W.H., and D.D. Smith. 1978. Predicting Rainfall Erosion Losses: A Guide to Conservation Planning.
Agricultural Handbook 537. U.S. Department of Agriculture, Washington, DC.
113
-------
114
-------
Contents
Appendix A. Model Fact Sheets
AGNPS: Agricultural Nonpoint Source Pollution
AGWA: Automated Geospatial Watershed Assessment
AnnAGNPS: Annualized Agricultural Nonpoint Source Pollution Model
AQUATOX
BASINS: Better Assessment Science Integrating point and Nonpoint Sources
CAEDYM: Computational Aquatic Ecosystem Dynamics Model
CCHE1D
CE-QUAL-ICM/TOXI
CE-QUAL-R1
CE-QUAL-RIV1
CE-QUAL-W2
CH3D-IMS: Curvilinear-grid Hydrodynamics 3D—Integrated Modeling System
CH3D-SED (& CH3D-WES): Curvilinear Hydrodynamics in Three Dimensions
DELFT3D
DIAS/IDLAMS: Dynamic Information Architecture System/Integrated Dynamic Landscape Modeling and Analysis System
DRAINMOD: A hydrological Model for Poorly Drained Soils
DWSM—Dynamic Watershed Simulation Model
ECOMSED: Estuary and Coastal Ocean Model with Sediment Transport
EFDC: Environmental Fluid Dynamics Code
EPIC: Erosion Productivity Impact Calculator
GISPLM: GIS-Based Phosphorus Loading Model
GLEAMS: Groundwater Loading Effects of Agricultural Management Systems
GLLVHT: Generalized. Longitudinal-Lateral-Vertical Hydrodynamics and Transport
GSSHA: Gridded Surface Subsurface Hydrologic Analysis
GWLF: Generalized Watershed Loading Functions
HEC-6: Scour and Deposition in Rivers and Reservoirs
HEC-HMS: Hydraulic Engineering Center Hydrologic Modeling System
HEC-RAS: Hydrologic Engineering Centers River Analysis System
HSCTM-2D: Hydrodynamic. Sediment and Contaminant Transport Model
HSPF: Hydrologic Simulation Program FORTRAN
KINEROS2: Kinematic Runoff and Erosion Model v2
LSPC: Loading Simulation Program in C++
-------
Mercury Loading Model: Watershed Characterization System—Mercury Loading Model
MIKE 11
MIKE 21
MIKE SHE
MINTEQA2: Metal Speciation Equilibrium Model for Surface and Groundwater
MUSIC: Model for Urban Stormwater Improvement Conceptualization
P8-UCM: Program for Predicting Polluting Particle Passage through Pits. Puddles, and Ponds—Urban Catchment Model
PCSWMM: Storm Water Management Model
PGC - BMP Module: Prince George's County Best Management Practice Module
QUAL2E: Enhanced Stream Water Quality Model
OUAL2K
REMM: Riparian Ecosystem Management Model
RMA-11
SED2D
SED3D: Three-Dimensional Numerical Model of Hydrodynamics and Sediment Transport in Lakes and Estuaries
SHETRAN
SLAMM: Source Loading and Management Model
SPARROW: SPAtially Referenced Regression on Watershed Attributes
STORM: Storage. Treatment Overflow. Runoff Model
SWAT: Soil and Water Assessment Tool
SWMM: Storm Water Management Model
TMDL Modeling Toolbox
TOPMODEL
WAMView: Watershed Assessment Model with an Arc View Interface
WARMF: Watershed Analysis Risk Management Framework
WASP: Water Quality Analysis Simulation Program
WEPP: Water Erosion Prediction Project
WinHSPF: An Interactive Windows Interface to HSPF
WMS: Watershed Modeling System
XP-SWMM: Stormwater and Wastewater Management Model
-------
Appendix: Model Fact Sheets
115
-------
116
-------
AGNPS: Agricultural Nonpoint Source Pollution
Contact Information
Ronald L. Bingner
U.S. Department of Agriculture
Agricultural Research Service
National Sediment Laboratory
598 McElroy Drive
Oxford, MS 38655
(662) 232-2966
rbingnertgimsa-oxford.ars.usda.gov
Download Information
Availability: Nonproprietary
Required to register prior to download, http://www.ars.usda.gov/Research/docs.htm?docid=5199
Cost: N/A
Model Overview/Abstract
AGNPS is a storm event model developed by the USDA Agricultural Research Service to estimate the pollution
loads from agricultural watersheds and to assess the effects of different management programs. AGNPS is capable
of simulating surface water runoff, nutrients, sediments, chemical oxygen demand, and pesticides from point and
nonpoint sources of agricultural pollution. The different point sources include feedlots, wastewater treatment plants,
gully erosion, and stream bank erosion. As a distributed model, AGNPS represents the spatial distribution of
watershed properties with a square-grid cells system. GIS interfaces in GRASS or TOPAZ are available to create the
model inputs. The model can be used to evaluate best management practices (BMPs).
Model Features
• Distributed, single-event model
• Point and nonpoint sources
Model Areas Supported
Watershed High or medium
Receiving Water None
Ecological None
Air None
Groundwater Medium
Model Capabilities
Conceptual Basis
AGNPS calculates surface runoff for each grid cell separately. The surface runoff calculated in each grid cell is then
routed through the watershed based on flow directions from one grid cell to the next grid cell until it reaches the
drainage outlet.
117
-------
Scientific Detail
AGNPS is a distributed parameter model developed by USD A, Agricultural Research Service (ARS) scientists and
engineers. It predicts soil erosion and nutrient transport/loadings from agricultural watersheds for real or
hypothetical storms; i.e., it's an event-based model. Its distributed model design derives from subdividing a
watershed to be simulated into a grid of square elemental areas, assumed to have uniform physical characteristics,
and then applying three lumped parameter models to each element:
• Erosion modeling is based on the USLE applied on a storm basis; thus, it uses an El-index but for single
storm events.
• Its hydrology is based on the Soil Conservation Service Curve Number technique, and it uses the Smith
algorithm for calculate peak flow rate.
• AGNPS uses another ARS-developed model named CREAMS to predict nutrient/pesticide and soil particle
size generation, transport and interaction. For sediment discharge, AGNPS uses the steady state continuity
equation, and it uses the Bagnold equation for sediment transport capacity calculation.
Outflows from one element become inputs to adjacent ones. Thus, AGNPS integrates lumped model predictions for
each element's behavior into a distributed watershed simulation.
Each AGNPS elemental area, typically about 100 m square, requires 22 parameters (coefficients) to describe its
antecedent conditions, physical characteristics (e.g., soil type and slope steepness), management practices and
rainfall. To predict NFS pollution, the USLE, SCS Curve Number hydrology and CREAMS relationships are
computed for each element as a function of time. Nineteen output parameters are computed, as a function of time,
for each watershed element.
AGNPS-GRASS developers recommend its use on watersheds up to 20,000 ha. (80 mi2) in size.
Model Framework
Subwatershed or cell-based distributed modeling framework. Modules are linked to calculate hydrology, erosion,
and nutrients loadings cell by cell. GIS interface is used to facilitate the cell-by-cell watershed characterization.
Scale
Spatial Scale
• One-dimensional, cell or subwatershed overland
• One-dimensional channel network
Temporal Scale
• Event
Assumptions
• Channels are assumed to have a triangular shape.
• Sediment eroded by rill and sheet erosion is assumed to be completely transported to stream without any
deposition in the fields.
• Surface runoff is assumed to flow through a 1 cm soil surface layer.
• Chemicals on the soil surface are assumed to be uniformly mixed with the surface layer.
• Infiltration first must pass through the surface layer.
• The initial abstraction (la) is the first increment of rainfall prior to surface runoff.
Model Strengths
• A distributed model
118
-------
• Capable of evaluating the effects of many BMPs, such as agricultural practices, ponds, grassed waterways,
irrigation, tile drainage, vegetative filter strips, and riparian buffers
Model Limitations
• Used only to simulate single event
• An empirical model
• Channels are assumed to have a triangular shape
Application History
AGNPS has been widely used for watershed studies (Hession et al., 1989; Engel et al., 1993; Mitchell et al., 1993;
Srinivasan and Engel, 1994) but mostly to evaluate land use change scenarios. Grunwald and Norton (1999) applied
AGNPS to two small watersheds in Germany to predict runoff and sediment yield and found that the application of
the model to unmonitored watersheds resulted in considerable under- and over prediction of surface runoff and
sediment yield. Other applications of AGNPS include Grunwald and Norton, (1999, 2000), Grunwald and Frede
(1999), and Chaubey et al. (1999).
Model Evaluation
SCS has conducted some model evaluation studies, which are available on the AGNPS Web site. A limited number
of studies used measured data to validate the AGNPS model. In a study by Srinivasan and Engel (1994) comparing
13 measured and simulated rainfall-runoff events, the simulated runoff volume was found to be underestimated for
all events.
Model Inputs
• Watershed delineation, cell (subwatershed) boundaries, land slope, slope direction, and reach
information—can be generated by TOPAZ, TOP AGNPS and AGFLOW
• Daily precipitation, maximum and minimum temperature, dew point temperature, sky cover, and wind
speed—can be generated by the climate data generator, GEM
• Management information, land characteristics, crop characteristics, field operation data, chemical operation
data, feedlots, and soil information—can be imported from RUSLE or NRCS sources
• If impoundment is present, then, elevation-storage power curve coefficient and exponent; elevation
discharge coefficient and exponent; permanent pool stage; runoff event water volume; and incoming mass
of sediment by particle size and its associated fall velocity
Users' Guide
Available online: http://sedlab.olemiss.edu/AGNPS.html
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS program that works on Windows 95/98
Programming language:
• Borland C
119
-------
Runtime estimates:
• Minutes
Linkages Supported
• NRCS GIS-support computer model HU/WQ to prepare input
Related Systems
AnnAGNPS
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• GIS interfaces available for creating the model inputs in TOPAZ, GRASS, and Arc/Info
References
Bingner, R. L. and F. D. Theurer. 2001. AGNPS 98: A Suite of water quality models for watershed use. In
Proceedings of the Sediment: Monitoring, Modeling, and Managing, 7th Federal Interagency Sedimentation
Conference, Reno, NV, March 25-29, 2001, pp. VII-1 - VII-8.
Bingner, R. L., F.D. Theurer, R.G. Cronshey, R.W. Darden. 2001. AGNPS 2001 Web Site.
http ://www. sedlab.olemiss.edu/AGNP S. html
Chaubey L, C.T. Haan, J.M. Salisbury, and S. Grunwald. 1999. Quantifying model output uncertainty due to spatial
variability of rainfall. J.AWRA. 35(5):1113-1123.
Engel, B. A., R. Srinivasan, J. Arnold, C. Rewerts, and S. J. Brown. 1993. Nonpoint source (NPS) pollution
modeling using models integrated with geographic information systems (GIS). Wat. Sci. Tech. 28(3-5):685-690.
Grunwald S. and L.D. Norton. 1999. An AGNPS-based runoff and sediment yield model for two small watersheds
in Germany. Trans. ASAE. 42(6):1723-1731.
Grunwald S. and L.D. Norton. 2000. Calibration and validation of a nonpoint source pollution model. Agricultural
Water Management. 45:17-39.
Grunwald S., andH.-G. Frede. 1999. Using AGNPS in German watersheds. Catena. 37(3-4):319-328.
Grunwald, S. and L. D. Norton. 1999. An AGNPS-based runoff and sediment yield model for two small watersheds
in Germany. Trans. ASAE. 42(6):1723-1731.
Hession, W. C., K. L. Huber, S. Mostaghimi, V. O. Shanholtz, and P. W. McClellan. 1989. BMP effectiveness
evaluation using AGNPS and a GIS. ASAE Paper No. 89-2566:1-18, ASAE, St. Joseph, Mich.
Srinivasan, R., and B. A. Engel. 1994. A spatial decision support system for assessing agricultural nonpoint source
pollution. Water Resour. Bull. 30(3):441-452.
Theurer, F.D., R.L. Bingner, W. Fontenot, and S.R. Kolian. 1999. Partnerships in Developing and Implementing
AGNPS 98: A suite of water quality models for watershed use. In Proceedings of the Sixth National Watershed
Conference, Austin, Texas, May 16-19, 1999.
120
-------
Ward, George H., Jr. and Jennifer Benaman. 1999. Models for TMDL application in Texas watercourses: Screening
and model review. Online Report CRWR-99-7. Center for Research in Water Resources, The University of Texas at
Austin.
Young, R. A., C. A. Onstad, D. D. Bosch, and W. P. Anderson. 1987. AGNPS, Agricultural Non-Point Source
Pollution Model - A watershed analysis tool. Conservation Research Report 35:1-80. United States Department of
Agriculture, Washington, DC.
Young, R. A., C. A. Onstad, D. D. Bosch, and W. P. Anderson. 1989. AGNPS: A nonpoint- source pollution model
for evaluating agricultural watersheds. J. Soil & Water Conserv. 44(2): 168-173.
121
-------
AGWA: Automated Geospatial Watershed Assessment
Contact Information
Mariano Hernandez or David C. Goodrich
U.S. Department of Agriculture
Agricultural Research Service
Watershed Research Center
2000 East Allen Road
Tucson, AZ 85719-1596
agwa(@,tucson.ars.ag.gov
http://www.tucson.ars.ag.gov/agwa
William G. Kepner or Darius J. Semmens
U.S. Environmental Protection Agency
P.O. Box 93478
Las Vegas, NV 89193-3478
http://www.epa.gov/nerlesdl/land-sci/agwa/index.htm
Download Information
Availability: Nonproprietary
http://www.tucson.ars.ag.gov/agwa/
Cost: None
Model Overview/Abstract
The USDA-ARS Southwest Watershed Research Center, in cooperation with the EPA Landscape Ecology Branch,
has developed the Automated Geospatial Watershed Assessment tool (AGWA) to facilitate using spatially
distributed data to prepare model input files and evaluate model results. AGWA uses widely available standardized
spatial datasets that can be obtained via the Internet. The data are used to develop input parameter files for
Kinematic Runoff and Erosion Model (KINEROS2) and Soil and Water Assessment Tool (SWAT), two watershed
runoff and erosion simulation models that operate at different spatial and temporal scales. AGWA automates the
process of transforming digital data into simulation model results and provides a visualization tool to help the user
interpret results. The utility of AGWA in joint hydrologic and ecological investigations has been demonstrated on
such diverse landscapes as southeastern Arizona, southern Nevada, central Colorado, and upstate New York.
Model Features
• User-friendly interface for generating model input using spatial data
• Flexibility in use from an event-oriented model for small watershed (>100 km2) to a continuous daily
timestep model for large complex watershed (>100 km2)
• Can be used to evaluate the impacts of land-use changes
• Available as an Arc View application or as an integrated part for BASINS version 3.1
122
-------
Model Areas Supported
Watershed High
Receiving Water Low
Ecological Low
Air None
Groundwater Low
Model Capabilities
Conceptual Basis
AGWA is a GIS-based system that integrates two watershed runoff and erosion models: an event-oriented model—
KINEROS—and a continuous daily timestep model—SWAT.
Scientific Detail
A fundamental assumption of AGWA is that the user has previously compiled the necessary GIS data layers, all of
which are easily obtained for the conterminous United States. The AGWA extension for Arc View adds the "AGWA
Tools" menu to the View window and must be run from an active view. Preprocessing of the DEM to ensure
hydrologic connectivity within the study area is required, and tools are provided in AGWA to aid in this task. Once
the user has compiled all relevant GIS data and initiated an AGWA session, the program is designed to lead the user
in steps through the transformation of GIS data into simulation results. The AGWA Tools menu is designed to
reflect the order of tasks necessary to conduct a watershed assessment, which is divided into five major steps: (1)
location identification and watershed delineation; (2) watershed subdivision; (3) land cover and soils
parameterization; (4) preparation of parameter and rainfall input files; and (5) model execution and visualization and
comparison of results.
Step 1: The user first creates a watershed outline, which is a grid based on the accumulated flow to the designated
outlet (pour point) of the study area. If a GIS coverage of the outlet location exists (such as would be the case for a
runoff gauging station), it can be used to designate the drainage outlet. Alternatively, the user has the option of using
a mouse to click on the watershed outlet. If internal gauging stations exist as a separate GIS coverage, AGWA will
use them as internal drainage pour points and generate output at each of the stations. This option is particularly
useful for calibration and validation of model results.
Step 2: A polygon shapefile is built from the watershed outline grid created in step 1. The user specifies the
threshold of contributing area for the establishment of stream channels, and the watershed is divided into model
elements required by the model of choice. From this point onward, tasks are specific to the model that will be used
(KINEROS2 or SWAT), but the same general process is followed, independent of model choice.
Step 3: The watershed created in Step 2 is intersected with soil and land cover data, and parameters necessary for the
hydrologic model runs are determined through a series of GIS analyses and look-up tables. The hydrologic
parameters are added to the polygon and stream channel tables to facilitate the generation of input parameter files.
At this point, the user can manually alter parameters for each model element if additional information is available to
guide the estimation of those values.
Step 4: Rainfall input files are built at this stage. For SWAT, the user must provide daily rainfall values for rainfall
gages within and near the watershed. If multiple gages are present, AGWA will build a Thiessen polygon map and
create an area-weighted rainfall file. For KINEROS2, users can select from a series of predefined rainfall events
dependent on the geographic location, choose to build their own rainfall file through an AGWA module, or use
NOAA Atlas II return period rainfall depth grids distributed with AGWA. Precipitation files may be created for
uniform (single gage) or distributed (multiple gage) rainfall data.
Step 5: After Step 4, all necessary input data have been prepared: The watershed has been subdivided into model
elements; hydrologic parameters have been determined for each element; and rainfall files have been created. The
user can proceed to run the hydrologic model of choice. AGWA will automatically import the model results and add
123
-------
them to the polygon and stream map tables for display. A separate module controls the visualization of model
results. The user can toggle among viewing various model outputs for both upland and channel elements, enabling
the problem areas to be identified visually. If multiple land cover scenes exist, the user can parameterize either or
both of the two models and attach the results to a given watershed. Results can then be compared on either an
absolute or percent-change basis for each model element (Miller et al., 2002a). Model results can also be overlaid
with other digital data layers to further prioritize management activities.
Model Framework
• SWAT
o Hydrologic response unit, subwatershed, and watershed
o Simple one-dimensional stream and well-mixed reservoir/lake model
• KINEROS
o Fields/planes and channels
• AGWA is based on Arc View GIS interface to process input data for the finest spatial unit of its component
models
Scale
Spatial Scale
• KINEROS: Fields and watershed with channel network.
• SWAT: Watershed with channel network, hydrological response unit, or single cell
Temporal Scale
• Different models in the AGWA have different temporal scales. KINEROS uses a variable timestep
(normally in minutes) to simulate a single storm event, and SWAT uses a daily timestep and can simulate a
watershed over 100 years
Assumptions
• Users previously have compiled the necessary GIS data layers, all of which can be obtained for the
conterminous United States.
• Users are familiar with GIS software
• Assumptions of the component models. See fact sheets for KINEROS and SWAT
Model Strengths
• Includes user-friendly interface to generate model input
• Includes more than one watershed model from which to choose
• Includes a visualization tool to display and interpret modeling results
• Can simulate watershed in different spatial (small to large watersheds) and temporal (event-based to
continuous daily) scales
Model Limitations
• Requires a large set of GIS data, making it difficult to setup the model for locations outside of the United
States
• Requires proprietary software: Arc View 3.x and the Spatial Analyst for grid operation
• May require training to use the advanced modeling options
Application History
The individual models within the model suite, such as KINEROS2 and SWAT have been used extensively for
watershed studies.
124
-------
There are several primary intended uses of AGWA. For example, AGWA can be used in a research environment as
a hydrologic modeling tool. Without a rigorous training set for calibration and validation, AGWA is well suited for
watershed assessment using hydrologic response as a metric of change. If multiple land cover scenes are available, a
relative assessment of the effects of land cover change on hydrologic response as a function of time may be
accomplished following Miller et al. (2002a).
Preliminary research during the development of AGWA was presented by Hernandez et al. (2000). In their study,
they showed that simulated runoff response is sensitive to land cover change in both the SWAT and KINEROS2
models and showed how the assumptions inherent in the look-up tables determine the direction and magnitude of
change.
Recent research by Miller, et al. (2002a) illustrated the use of AGWA in coordinated ecological and hydrologic
assessment. The authors analyzed of the ecological changes since the early 1970s within the Upper San Pedro River
Basin in southeastern Arizona and the Cannonsville Watershed in the Catskill/Delaware region of New York.
Model Evaluation
See application history and Miller et al. (2002b).
Model Inputs
KINEROS
• Overland flow element:
o Plane geometry (length, width, and slope), Manning's n, Chezy conveyance factor
o Canopy cover, interception depth, average micro topographic relief, average micro topographic
spacing
o Infiltration related parameters: saturated hydraulic conductivity (Ks), initial degree of soil
saturation, coefficient of variation of Ks, mean capillary drive, porosity, pore size distribution
index, volumetric rock fraction, thickness of soil layers (up to two layers)
o Rain splash coefficient, soil cohesion coefficient, and particle class fractions.
• Channel element
o Type (simple or compound), base flow discharge
o Channel geometry (length, width, bed slope, and bank slopes), Manning's n, and Chezy
conveyance factor
o Infiltration related parameters (same as specified in overland flow element)
o Cohesion coefficient of bed material
• Pond element
o Initial storage volume
o Volume, surface area, and discharge rating table
o Seepage rate
• Rainfall File
• External flow file (optional)
SWAT
• Land uses (MRLC and others)
• Soil (STATSGO and others)
• Topography (30 x 30 m2 DEM or other resolutions)
• Subwatersheds (derived from manual or auto delineation tools in BASINS 3.0)
• Point Source (PCS or other database)
• Climate data (daily temperature, precipitation, solar radiation, and wind speed)
• Crop and management databases
• USGS flow data (for calibration)
• Long-term watershed quality data (for model calibration)
125
-------
Users' Guide
Available online: http://www.tucson.ars.ag.gov/agwa/
Technical Hardware/Software Requirements
Computer hardware:
• IBM-PC
Operating system:
• Windows XP/2000/NT/98
Programming language:
• AGWA is in Arc View Avenue.
• AGWA requires Arc View 3.1 or later and Spatial Analyst version 1.1.
• Models SWAT and KINEROS are in FORTRAN.
Runtime estimates:
• Minutes to less than an hour
Linkages Supported
AGWA is also available as an integrated part of BASINS 3.1.
Related Systems
• Migration to ArcGIS is currently being developed.
• DotAGWA, a Web-based interface for AGWA, is also under development.
• Better Assessment Science Integrating Point and Nonpoint Sources (BASINS), version 3.1
Sensitivity/Uncertainty/Calibration
The SWAT interface allows basic model calibration and sensitivity analysis (27 key parameters). No tools were
developed for uncertainty analysis.
In numerous modeling studies, the KINEROS model has been applied on the USDA-ARS Walnut Gulch
Experimental Watershed. Goodrich et al. (1994) investigated the sensitivity of runoff production to the pattern of
antecedent moisture condition at the small watershed scale (6.31 km2). They suggested that a simple basin average
of initial moisture content will normally prove adequate and that, again, knowledge of the rainfall patterns is far
more important. Michaud and Sorooshian (1994) compared three different models at the scale of the whole
watershed, a lumped curve number model, a simple distributed curve number model, and the more complex
distributed KINEROS model. The modeled events were 24 severe thunderstorms with a rain gage density of one per
20 km2. Their results suggested that none of the models could adequately predict peak discharge and runoff volumes
but that the distributed models did somewhat better in predicting time to runoff initiation and time to peak.
Goodrich et al. (1997) used data from the entire Walnut Gulch watershed to investigate the effects of storm area and
watershed scales on runoff coefficients. They concluded that, unlike humid areas, there is a tendency for runoff
response to become more nonlinear with increasing watershed scale in this type of semi-arid watershed, as a result
of the loss of water into the bed of ephemeral channels and the decreasing relative size of rainstorm coverage with
watershed area for any individual event.
126
-------
Model Interface Capabilities
• Customized Arc View 3.x interface with the capacity to automate the model input creation
References
Goodrich, D.C., L.J. Lane, R.A. Shillito, S.N. Miller, K.H. Syed, and D.A. Woolhiser. 1997. Linearity of basin
response as a function of scale in a semi-arid watershed. Water Resources Research. 33 (12) :2951-2965.
Goodrich, D.C., TJ. Schmugge, TJ. Jackson, C.L. Unkrich, T.O. Keefer, R. Parry, L.B. Bach, and S.A. Amer.
1994. Runoff simulation sensitivity to remotely sensed initial soil water content. Water Resources Research. 30
(5):1393-1405.
Hernandez, M, S.N. Miller, D.C. Goodrich, B.F. Goff, W.G. Kepner, C.M. Edmonds, and K.B. Jones. 2000.
Modeling runoff response to land cover and rainfall spatial variability in semi-arid watersheds. Environmental
Monitoring and Assessment. 64:285-298.
Hernandez, M., W.G. Kepner, D.J. Semmens, D.W. Ebert, D.C. Goodrich, and S.N. Miller. 2003. Integrating a
Landscape/Hydrologic Analysis for Watershed Assessment. In Proceedings of the First Interagency Conference on
Research in the Watersheds, U.S. Department of Agriculture, Agricultural Research Service, October 27-30, 2003,
pp. 461-466.
Michaud, J.D., and S. Sorooshian. 1994. Comparison of simple versus complex distributed runoff models on a
midsized semiarid watershed. Water Resources Research. 30 (3):593-605.
Miller, S.N., W.G. Kepner, M.H. Mehaffey, M. Hernandez, R.C. Miller, D.C. Goodrich, F. Kim Devonald, D.T.
Heggem, and W.P. Miller. 2002. Integrated landscape assessment and hydrologic modeling for land cover change
analysis. Journal of the American Water Resources Association. Special Volume on Watershed Management and
Landscape Studies.
Miller, S.N., D.J. Semmens, R.C. Miller, M. Hernandez, D.C. Goodrich, W.P. Miller, W.G. Kepner, and D.W.
Ebert. 2002b. GIS-based Hydrologic Modeling: The Automated Geospatial Watershed Assessment Tool. In
Proceedings of the Second Federal Interagency Hydrologic Modeling Conference, Las Vegas, Nevada, July 28-
August 1, 2002. Available online: http://www.epa.gov/nerlesdl/land-sci/agwa/pdf/pubs/agwa-conference.pdf.
127
-------
AnnAGNPS: Annualized Agricultural Nonpoint Source Pollution Model
Contact Information
Ronald L. Bingner
U.S. Department of Agriculture
Agricultural Research Service
National Sediment Laboratory
598 McElroy Drive
Oxford, MS 38655
(662) 232-2966
rbingnertgjmsa-oxford.ars.usda.gov
http://msa.ars.usda.gov/ms/oxford/nsl/AGNPS.html
Download Information
Availability: Nonproprietary
(required to register prior to download)
Cost: N/A
Model Overview/Abstract
AnnAGNPS is a continuous simulation watershed-scale program developed based on the single-event model
AGNPS. AnnAGNPS simulates quantities of surface water, sediment, nutrients, and pesticides leaving the land
areas and their subsequent travel through the watershed. Runoff quantities are based on runoff curve numbers while
sediment is determined by using the Revised Universal Soil Loss Equation (RUSLE). Special components are
included to handle concentrated sources of nutrients (feedlots and point sources), concentrated sediment sources
(gullies), and added water (irrigation). Output is expressed on an event basis for selected stream reaches and as
source accounting (contribution to outlet) from land or reach components over the simulation period. The model can
be used to evaluate best management practices (BMPs).
Model Features
• Distributed, continuous simulation
• Point and nonpoint sources
• Source accounting
Model Areas Supported
Watershed High or medium
Receiving Water None
Ecological None
Air None
Groundwater Medium
Model Capabilities
Conceptual Basis
AnnAGNPS divides the watershed into homogenous drainage areas, which are then integrated together by simulated
rivers and streams, routing the runoff and pollutants from each area downstream. The hydrology of the model is
based on simple water balance approach.
128
-------
Scientific Detail
The model uses the SCS curve number technique to calculate the daily runoff and RUSLE 1.05 technology to
calculate daily sheet and rill erosion. The Hydro-geomorphic Universal Soil Loss Equation (HUSLE) is used for the
calculation of the sediment delivery ratio (yield from erosion divided by the amount delivered to the stream). The
instantaneous peak discharge of the runoff hydrograph is calculated using TR-55, which is then used to calculate the
time of concentration, Tc. The key processes and their details are given below:
• Climate - Climate data are generated using GEM and Complete_Climate
• Hydrology - Daily soil moisture balance
• Runoff-SCS curve number
• Potential evapotranspiration - Penman equation
• Subsurface flow - lateral subsurface flow using Darcy's equation or tile drain flow
• Rill and sheet erosion - RUSLE
• Sediment delivery - HUSLE
• Chemical routing - dissolved or adsorbed by mass balance approach
Model Framework
• Subwatershed or cell-based approach
• Simple one-dimensional channel routing
• GIS interface is used for watershed characterization and model parameterization
Scale
Spatial Scale
• One-dimensional grid or subwatershed overland
• One-dimensional channel network
Temporal Scale
• Daily
Assumptions
The climate generation module assumes that daily precipitation is independent of any precipitation on either the day
before or day after. The RUSLE K and C factors are assumed to not vary significantly day-to-day and thus the
minimum timestep of 15 days used in RUSLE is assumed to be appropriate.
Model Strengths
AnnAGNPS is a distributed parameter, watershed scale model that is used for continuous simulations. AnnAGNPS
can be used to study the effect of BMPs (agricultural practices, ponds, grassed waterways, irrigation, tile drainage,
vegetative filter strips and riparian buffers).
Model Limitations
All runoff and associated sediment, nutrient, and pesticide loads for a single day are routed to the watershed outlet
before the next day simulation. There is no tracking of nutrients and pesticides attached to sediment deposited in
stream reaches from one day to the next. Point sources are limited to constant loading rates (water and nutrients) for
the entire simulation period. Spatially variable rainfall is not allowed.
129
-------
Application History
There have been few application studies of the model other than the model evaluation studies discussed in the next
section. Srivastava et al. (2002) conducted a study using AnnAGNPS and genetic algorithm of optimization of best
management practices. Baginska et al. (2003) applied AnnAGNPS and PEST model to predict nutrient export from
a small catchment in Australia and found that the accuracy of predictions was moderate.
Model Evaluation
In a model evaluation study, Yuan et al. (2001) found out that the model-predicted monthly sediment yield was in
close agreement (R2 = 0.7) with the actual observed sediment yield, but the short-term individual event predictions
were not acceptable. In another study, Yuan et al. (2003) found out that AnnAGNPS predictions of monthly
loadings were poor though statistically not significantly different from observed values.
Model Inputs
• Watershed delineation, cell (subwatershed) boundaries, land slope, slope direction, and reach information -
generated by TOPAGNPS and AGFLOW
• Daily precipitation, maximum and minimum temperature, dew point temperature, sky cover, and wind
speed - can be generated by the climate data generator, GEM
• Management information, land characteristics, crop characteristics, field operation data, chemical operation
data, feedlots, and soil information - can be imported from RUSLE or NRCS sources
• If impoundment is present, then elevation-storage power curve coefficient and exponent; elevation
discharge coefficient and exponent; permanent pool stage; runoff event water volume; and incoming mass
of sediment by particle size and its associated fall velocity
Users' Guide
Available online: http://sedlab.olemiss.edu/AGNPS.html
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• Windows 98/NT/2000 and XP
Programming language:
• ANSI FORTRAN 95
Runtime estimates:
• Minutes
Linkages Supported
Different tools or models that are linked to AnnAGNPS include TOPAZ for watershed delineation, Stream Network
Temperature Model (SNTEMP), Sediment Intrusion & Dissolved Oxygen Model (SIDO), Conservation Channel
Evolution and Pollutant Transport System Model (CONCEPTS), and Stream Network Watershed Scale Model
(CCHE1D). NRCS plans to revise HU/WQ to work with AnnAGNPS.
Related Systems
AGNPS is the predecessor of AnnAGNPS.
130
-------
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Window GUI for editing input
• GIS interface in Arc View and ArcGIS for preprocessing input data
References
Baginska, B., W. Milne-Home, and P. S. Cornish. Modelling nutrient transport in Currency Creek, NSW with
AnnAGNPS and PEST. Environ. Model. & Software. 18:801-808.
Bingner, R. L., F.D. Theurer, R.G.Cronshey, R.W.Darden. 2001. AGNPS 2001 Web Site.
http://www.sedlab.oleniiss.edu/AGNPS.html.
Srivastava, P., J. M. Hamlett, P. D. Robillard, and R. L. Day. 2002. Watershed optimization of best management
practices using AnnAGNPS and a genetic algorithm. Water Resour. Res. 38(3):1021.
Yuan, Y., Bingner, R. L., and Rebich, R. A. 2001. Evaluation of AnnAGNPS on Mississippi Delta MSEA
Watersheds. Trans. ASAE. 44(5): 1183-1190.
Yuan, Y., Bingner, R. L., and Rebich, R. A. 2003. Evaluation of AnnAGNPS nitrogen loading in an agricultural
watershed. J AWRA. 39(2):457-466.
131
-------
AQUATOX
Contact Information
Marjorie Wellman
U.S. Environmental Protection Agency
Office of Water
Ariel Rios Building
1200 Pennsylvania Avenue, N. W.
Washington, DC 20460
(202) 566-0407
wellman. marj orie(g),epa. gov
http ://www. epa. gov/waterscience/models/aquatox/about. html
Download Information
Availability: Nonproprietary
Cost: N/A
Model Overview/Abstract
AQUATOX simulates multiple environmental stressors (including nutrients, organic loadings and chemicals, and
temperature) and their effects on the algal, macrophyte, invertebrate, and fish communities. Therefore, AQUATOX
can help identify and understand the cause and effect relationships between chemical water quality, the physical
environment, and aquatic life. AQUATOX can represent a variety of aquatic ecosystems spatially, including
vertically stratified lakes, reservoirs and ponds, and rivers and streams.
Model Features
• Evaluates which of several stressors is causing observed biological impairment
• Predicts effects of pesticides and other toxic substances on aquatic life
• Evaluates potential ecosystem responses to invasive species
• Explores how changes in land use or agricultural practices in a watershed might affect aquatic life, by using
the new linkage to BASINS
• Compares differences in biological responses to control alternatives
• Develops targets for nutrients in lakes and reservoirs with nuisance algal blooms
• Estimates time to recovery of fish or invertebrate communities after reducing pollutant loads
• Calculates bioaccumulation factors for organic toxic chemicals
• Estimates how long before tissue levels of toxic organics in fish will return to safe levels following removal
of contaminated sediments
• Has a large increase in the number of biotic state variables, with two representatives for each taxonomic
group or ecologic guild
• Macrophyte category includes bryophytes
• Includes a multi-age fish category with up to 15 age classes for age-dependent bioaccumulation and limited
population modeling
• Simulates maximum of 20 toxicants, with the capability for modeling daughter products due to
biotransformations
• Disaggregates stream habitats into riffle, run, and pool
• Includes mechanistic current- and stress-induced sloughing, light extinction, and accumulation of detritus
in periphyton
• Includes macrophyte breakage due to currents
132
-------
Computes chlorophyll a for periphyton and bryophytes, as well as for phytoplankton
Enters and tracks fish biomass in g/m2
Includes entrainment and washout of animals, including fish, during high flow
Has options of computing respiration and maximum consumption in fish as functions of mean individual
weight, using allometric parameters from the Wisconsin Bioenergetics Model
Includes density-dependent respiration in fish
Fish spawning can occur on user specified dates as an alternative to temperature-cued spawning
Includes detailed elimination of toxicants from biota
Includes settling and erosional velocities for inorganic sediments as user-supplied parameters
Includes uncertainty analysis that covers all parameters and loadings
Provides biotic risk graphs as an alternative means of portraying probabilistic results
Outputs limitation factors for photosynthesis along with the biotic rates
Extends BASINS, providing linkages to GIS data, and HSPF and SWAT simulations
Model Areas Supported
Watershed
Receiving Water
Ecological
Air
Groundwater
None
Medium
High
None
None
Model Capabilities
Conceptual Basis
AQUATOX predicts the fate of various pollutants, such as nutrients and organic toxicants, and their effects on the
ecosystem, including fish, invertebrates, and aquatic plants (Figure 1). It simultaneously computes important
chemical and biological processes over time. AQUATOX can predict not only the fate of chemicals in aquatic
ecosystems but also their direct and indirect effects on the resident organisms.
IK AI.H' Uu.X
133
-------
Scientific Detail
AQUATOX uses differential equations to represent changing values of state variables, normally with a reporting
timestep of one day. AQUATOX uses fourth- and fifth-order Runge-Kutta integration routines with adaptive step
size to solve the differential equations. The routine uses the fifth-order solution to determine the error associated
with the fourth-order solution; it decreases the step size (often to 15 minutes or less) when rapid changes occur and
increases the step size when there are slow changes, such as in winter. However, the step size is constrained to a
maximum of one day so that daily pollutant loadings are always detected. The reporting step, on the other hand, can
be as long as 99 days or as short as 0.1 day; the results are integrated to obtain the desired reporting time period.
The process equations contain another class of input variables—the parameters or coefficients that allow the user to
specify key process characteristics. For example, the maximum consumption rate is a critical parameter
characterizing various consumers. AQUATOX is a mechanistic model with many parameters; however, default
values are available so that the analyst has to be concerned with only those parameters necessary for a specific risk
analysis, such as characterization of a new chemical.
The system being modeled is characterized by site constants, such as mean and maximum depths. At present, one
can model small lakes, reservoirs, streams, small rivers, and ponds—and even enclosures and tanks.
Model Framework
It uses a waterbody ecosystem compartment framework to describe interactions between biotic and abiotic system
components.
The fate portion of the model, which applies especially to organic toxicants, includes partitioning among organisms,
suspended and sedimented detritus, suspended and sedimented inorganic sediments, and water; volatilization;
hydrolysis; photolysis; ionization; and microbial degradation. The effects portion of the model includes chronic and
acute toxicity to the various organisms modeled; and indirect effects such as release of grazing and predation
pressure, increase in detritus and recycling of nutrients from killed organisms, dissolved oxygen sag due to increased
decomposition, and loss of food base for animals.
Scale
Spatial Scale
• One-dimensional stream
• Two-box for reservoir, lake, and pond
Temporal Scale
• Daily
Assumptions
Aquatic system is considered to be a well-mixed condition. The model uses a two-box (epilimnion and hypolimnion)
approach for the lake.
Model Strengths
AQUATOX is an ecosystem model that predicts the fate of nutrients and organic chemicals in water bodies as well
as their direct and indirect effects on the resident organisms. Most water quality models predict only concentrations
of pollutants in water; they do not project effects of pollutants on organisms.
Model Limitations
AQUATOX represents the aquatic ecosystem by simulating the changing concentrations (in mg/L or g/m3) of
organisms, nutrients, chemicals, and sediments in a unit volume of water. It differs from population models, which
represent the changes in numbers of individuals modeling individual species at risk and modeling fishing pressure
and other age- or size-specific aspects.
134
-------
Application History
Validation studies using AQUATOX:
• Nutrient analysis on Onondaga Lake, New York
(http://www.epa.gov/waterscience/models/aquatox/validation/onondaga.pdf)
• Nutrient analysis of the Coralville Reservoir, Iowa
(http://www.epa.gov/waterscience/models/aquatox/validation/coralville.pdf)
• Bioaccumulation of PCBs in Lake Ontario
(http ://www. epa. gov/waterscience/models/aquatox/validation/ontario .pdf)
• Simulation of periphyton in Walker Branch, Tennessee
(http://www.epa.gov/waterscience/models/aquatox/periphvtonvalid.pdf)
Model Evaluation
The model and documentation have undergone successful peer review by an external panel convened by the U.S.
Environmental Protection Agency.
Model Inputs
• Loadings to the waterbody
• General site characteristics
• Chemical characteristics of any organic toxicant
• Biological characteristics of the plants and animals.
AQUATOX comes bundled with data libraries that provide default data. These data libraries are of particular
importance for the biological data, which are probably the most difficult for a user to obtain.
Users' Guide
Available online: http://www.epa.gov/waterscience/models/aquatox/users/user.pdf
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• Graphical User Interface
Programming language:
• Object-oriented Pascal using the Delphi programming system for Windows
Runtime estimates:
• Seconds to minutes
Linkages Supported
BASINS
Related Systems
None
135
-------
Sensitivity/Uncertainty/Calibration
AQUATOX provides probabilistic modeling capability to consider the implications of uncertainty in the modeling
analyses by allowing the user to specify the types of distributions and key statistics for any and all input variables.
Quantitative uncertainty analysis is based on a Monte Carlo simulation.
Model Interface Capabilities
• Pre- and postprocessors
• Data display tools
References
U.S. Environmental Protection Agency. 2004. Users Manual for AQUATOX (Release 2): Modeling Environmental
Fate and Ecological Effects in Aquatic Ecosystems Volume 1. EPA-823-R-04-001. (Computer program manual).
U.S. Environmental Protection Agency, Office of Water, Washington, DC.
Park, R. A., and J. S. Clough. 2004. Aquatox (Release 2): Modeling Environmental Fate and Ecological Effects in
Aquatic Ecosystems Volume 2: Technical Documentation. U.S. Environmental Protection Agency, Office of Water,
Washington, DC.
U.S. Environmental Protection Agency. 2000. AQUATOX for Windows: A Modular Fate and Effects Model for
Aquatic Ecosystems-Volume 3: Model Validation Reports. U.S. Environmental Protection Agency, Office of Water,
Washington, DC.
136
-------
BASINS: Better Assessment Science Integrating point and Nonpoint Sources
Contact Information
U.S. Environmental Protection Agency (Mailcode - 4305T)
Office of Science and Technology
Standards and Health Protection Division
Modeling and Information Technology Team
1200 Pennsylvania Ave., NW
Washington, DC 20460
basins(g),epa.gov
Download Information
Availability: Nonproprietary
http://www.epa.gov/waterscience/basins/index.html
Cost: N/A
Model Overview/Abstract
BASINS was developed by the EPA's Office of Water to support environmental and ecological studies in a
watershed context. BASINS, a multipurpose environmental analysis system designed for use by regional, state, and
local agencies in performing watershed and water quality-based studies, was originally introduced in 1996 with
subsequent releases in 1998 and 2001. BASINS works with a geographic information system (GIS) framework and
consists of: (1) national databases (2) assessment tools (3) a watershed delineation tool (4) classification utilities (5)
characterization reports (6) watershed loading and transport models, HSPF and Soil and Water Assessment Tool,
(SWAT); (7) a simplified GIS-based model, PLOAD, that estimates annual average nonpoint loads; (8) the
Automated Geospatial Watershed Assessment (AGWA) tool, a GIS-based hydrologic modeling tool; and (9) model
calibration tool, Parameter Estimation (PEST) tool. The system provides a user-friendly interface to conduct simple
watershed-level screening analysis or detailed water quality modeling studies. The interface also helps the user to
easily create the input files for various models.
Model Features
• User-friendly interface
• Flexibility in use from the simple watershed-level screening analysis to detailed water quality modeling
• Easy access to required GIS data layers for locations within the United States
• Linkage between GIS and the selected popular watershed and water quality models
Model Areas Supported
Watershed High
Receiving Water High
Ecological Medium
Air None
Groundwater Medium
137
-------
Model Capabilities
Conceptual Basis
BASINS is a GIS-based system that integrates a suite of watershed and water quality models with different
approaches.
Scientific Detail
BASINS is a freely available multipurpose Arc View desktop environmental analysis system for use by regional,
state, and local agencies in performing watershed and water quality-based studies. Many states and local agencies
are moving toward a focused, watershed-based approach. The BASINS system is configured to support
environmental and ecological studies in a watershed context. The system is designed to be flexible but support a
variety of scales using tools that range from simple to sophisticated.
BASINS also was conceived as a system for supporting the development of total maximum daily loads (TMDLs), as
defined in Section 303(d) of the Federal Clean Water Act. A TMDL is the sum of the allowable loads of a particular
pollutant from all contributing point and nonpoint sources. The calculation must include a margin of safety (MOS)
to ensure that the waterbody can be used for the purposes the state has designated (i.e., drinking water, fishing,
swimming, etc.). Developing TMDLs requires a watershed-based point and nonpoint source analysis for a variety of
pollutants. It also lets the modeler test different best management options for the impaired waterbody.
Traditional approaches to watershed-based assessments typically involve many separate steps for preparing data,
summarizing information, developing maps and tables, and applying and interpreting models. BASINS makes
watershed and water quality studies easier by bringing key data and analytical components together on a user's
desktop. Using the now-familiar Windows environment, an analyst can quickly access national environmental data,
apply some assessment and analysis tools, run several calculations and processes through hundreds of nonpoint
source loadings, and obtain results in the form of maps, charts, graphs, and reports from a choice of water quality
models in a matter of minutes.
Model Framework
BASINS consists of a GIS-based framework that links to environmental databases, characterization tools, and
watershed models that simulate watershed and subwatershed processes with 1-D streams.
• BASINS is based on the Arc View GIS interface.
• BASINS includes many Arc View modular extensions
• BASINS integrates environmental data and watershed and water quality models into a coherent system for
assisting in TMDL development and solving environmental problems.
The following components are new in BASINS 3.x:
• Extensions with an Extension Manager
• Web-based online help
Scale
Spatial Scale
• Watershed scale
• One-dimensional waterbody
Temporal Scale
• Different models in the BASINS suite have different temporal scales—HSPF: user-defined timestep,
typically hourly, continuous simulation from days to years; SWAT: daily timestep, continuous simulation
for months to years; PLOAD: Export coefficient model, annual; and KINEROS: single-storm event, part of
AGWA, variable timestep typically in minutes
138
-------
Assumptions
• Uses a lumped approach for the hydrologic response unit
Model Strengths
• Includes a large dataset for the nation
• Includes more than one watershed or water quality model from which to choose
• Has associated automatic downloading of data from Internet
• Has good customer support and e-mail listing service
Model Limitations
• Requires a large set of GIS data, making it difficult to setup the model for locations outside of the United
States
• Requires proprietary software: Arc View 3.x, and the Spatial Analyst for grid operation
• May require training to use the advanced modeling options
Application History
Previous versions of BASINS have been applied to many TMDL studies across the United States. The individual
models within the model suit like SWAT and HSPF have been used extensively for watershed and water quality
studies.
Model Evaluation
BASINS system and most of its components have been used for many TMDL developments. There are many peer-
reviewed publications available for the system and individual models.
Model Inputs
The BASINS system requires many GIS data layers with specific attributes combined with them. For locations
within the United States, all the essential datasets can be downloaded from EPA's Web site based on 8-digit
watershed or HUC in the lower 48 states, Alaska, Hawaii, and Puerto Rico with the U.S. Virgin Islands. The input
requirements for individual watershed and water quality models available in the system vary among the models. For
example, the HSPF model requires hourly precipitation depths, while SWAT requires only daily precipitation
depths.
Users' Guide
Available online: http://www.epa.gov/waterscience/basins/bsnsdocs.html
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• Windows XP/2000/NT/98
Programming language:
• BASINS system and PLOAD are developed in Arc View 3.X.
• Models including HSPF, SWAT, and KINEROS are in FORTRAN
139
-------
Runtime estimates:
• Minutes
Linkages Supported
BASINS system is a suite of tools and models linked together.
Models
• Updated the Hydrological Simulation Program-Fortran (HSPF) to version 12 and created a Windows
interface for the HSPF model (WinHSPF) to replace the NonPoint Source Model (NPSM) in the previous
versions of BASINS.
• The Automated Geospatial Watershed Assessment (AGWA) tool features the USDA-ARS models
KINEROS and SWAT.
• The Kinematic Runoff and Erosion Model (KINEROS)
• Soil and Water Assessment Tool (SWAT), developed by the USDA Agriculture Research Service.
BASINS uses the updated SWAT2000 model.
• A model called PLOAD, developed by CH2M-Hill, which uses export coefficients to estimate watershed
loading. Rosgen's Bank Erosion Hazard Index has been incorporated into PLOAD as PLOAD-BEHI
• AQUATOX receives and automatically formats output from HPSF or SWAT in order to integrate
watershed analysis with the likely effects on the aquatic biota in receiving waters.
Related Systems
Watershed Characterization System (WCS) (http://wcs.tetratech-ffx.com). TMDL Modeling Toolbox
(http ://www. epa. gov/athens/wwqtsc/html/tools. html)
Sensitivity/Uncertainty/Calibration
The new Parameter Estimation (PEST) tool in WinHSPF automates the model calibration process and allows users
to quantify the uncertainty associated with specific model predictions. This tool can also be used for uncertainty
analysis, such as Monte Carlo Analysis.
Model Interface Capabilities
• Customized Arc View 3.x interface with individual extensions organized in the following categories:
Assess, Data, Delineate, Models, Reports, and Utilities
• A scenario generator, GenScn, developed by USGS, that allows users to manage, visualize, analyze, and
compare results from WinHSPF or SWAT simulations
• A Web data extraction tool that dynamically downloads GIS data and databases from the BASINS Web site
and a variety of other sources
• A tool that automatically checks all components of the BASINS application and the last update
• Automatic watershed delineation tools based on DEM GRID
• Updated manual delineation tools based on Arc View's dynamic segmentation process
• A GRID projector, which requires Spatial Analyst extension
• An FTP process for downloading NHD layers from USGS and importing, then projecting them directly into
a BASINS project window
• A tool to archive and restore BASINS projects
References
Lahlou, M., L. Shoemaker, S. Choudhury, R. Elmer, A. Hu, H. Manguerra and A. Parker. 1998. Better Assessment
Science Integrating Point and Nonpoint Sources—BASINS Version 1.0. EPA-823-B-98-006. (Computer program
manual). U.S. Environmental Protection Agency, Office of Water, Washington, DC.
U.S. Environmental Protection Agency. 2004. Better Assessment Science Integrating Point and Nonpoint Sources
Release 3.1. EPA/823/C-04/004. (Computer program manual). U.S. Environmental Protection Agency, Office of
Water, Washington, DC.
140
-------
CAEDYM: Computational Aquatic Ecosystem Dynamics Model
Contact Information
David Horn
University of Western Australia
Centre for Water Research
35 Stirling Highway
Crawley
Western Australia
Australia 6009
+61 8 9380 1684
horn@cwr.uwa.edu.au
www.cwr.uwa.edu.au
Download Information
Availability: Nonproprietary
www.cwr.uwa.edu.au/~ttfadmin/model/caedym
Cost: N/A
Model Overview/Abstract
The Computational Aquatic Ecosystem Dynamics Model (CAEDYM) is a comprehensive aquatic ecological model
that is based on the nutrient-phytoplankton-zooplankton food chain relationship. The state variables of CAEDYM
include carbon, oxygen, silica, inorganic paniculate, and other biological factors. CAEDYM can be linked with
several hydrodynamic models, such as DYRESM and ELCOM, to describe one-, two-, and three-dimensional
processes of primary production, secondary production, nutrient and metal cycling, and oxygen dynamics and the
movement of sediment.
Model Features
• Phytoplankton: up to seven groups
• Dissolved oxygen, biochemical oxygen demand ("fast" and "slow")
• Nutrients (NH4, NO3, PO4, TP, TN and internal phytoplankton N, P and C)
• Suspended solids: two groups
• Zooplankton: up to five groups
• Fish: up to nine groups including jellyfish and seagrasses/macrophytes
• Macroalgae
• Macroinvertebrates (bivalves, polychaetes and crustacean grazers)
• pH
• Metals: iron, manganese and aluminum
Model Areas Supported
Watershed None
Receiving Water Medium
Ecological High
Air None
Groundwater None
141
-------
Model Capabilities
Conceptual Basis
The waterbody is conceptualized as a network of grid points (finite difference).
Scientific Detail
CAEDYM is a detailed aquatic ecological model that simulates the nutrients, phytoplankton, zooplankton, fish, and
benthic habitat. The model requires external hydrodynamic models to provide temperature, salinity, and transport
driving forces. CAEDYM usually runs at the same timestep as the hydrodynamic models. The state variables
include dissolved oxygen, inorganic nutrients, dissolved organic nutrients, paniculate organic nutrients, and
inorganic suspended solids in both the water column and sediment layer. The water column variables also include
pH, color, one group of bacteria, seven groups of algae, five groups of zooplankton, one group of jellyfish, five
groups of fish, one group of pathogen, four groups of macroalgae, one group of seagrass, and three groups of
invertebrates.
Model Framework
• One-dimensional vertical
• One-dimensional longitudinal
• Two-dimensional longitudinal-vertical
• Three-dimensional
• Reservoir, lake, estuary, river, floodplain
Scale
Spatial Scale
• One-, two-, or three-dimensional
Temporal Scale
• User-defined timestep
Assumptions
Aquatic ecological dynamics can be described with the ordinary differential equations.
Model Strengths
• Detailed aquatic ecology, strong ecological modeling capability
• Flexible structure
Model Limitations
• No sediment diagenesis
Application History
CAEDYM is used in 59 countries around the world.
Model Evaluation
Not available
Model Inputs
• Initial concentrations of state variables
142
-------
• Inflows and concentrations in inflows and over forcing regions
• Parameter values
• Other data may be required by the hydrodynamic driver (DYRESM or ELCOM), e.g., meteorological
forcing data
Users' Guide
Computational Aquatic Ecosystem Dynamics Model CAEDYM v2.1 Science Manual
Computational Aquatic Ecosystem Dynamics Model CAEDYM v2.1 User Manual
Available online: http://www.cwr.uwa.edu.au/~ttfadmin/model/caedvm/
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• Windows 95/98/NT, Linux, DEC Unix
Programming language:
• FORTRAN 95
Runtime estimates:
• Minutes to hours
Linkages Supported
CAEDYM can be linked with the following hydrodynamic models:
• DYRESM (one-dimensional vertical for deep lakes and reservoirs)
• DYRISM (Quasi-two-dimensional Lagrangian for rivers and floodplains)
• ELCOM-2D (two-dimensional laterally averaged for narrow lakes and reservoirs)
• ELCOM (three-dimensional for any waterbody)
Related Systems
CE-QUAL-Rl, EFDC, CE-QUAL-ICM, WASP/EUTRO, CE-QUAL-RIV1, CE-QUAL-W2
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Graphic Interface written with Java
References
Hipsey, M.R., J.R. Romero, J.P. Antenucci, and D.P. Hamilton. 2004. Computational Aquatic Ecosystem Dynamics
Model CAEDYM v2. (Computer program manual). Centre for Water Research, The University of Western Australia.
Romero, J.R., M.R. Hipsey, J.P. Antenucci, and D.P. Hamilton. 2004. Computational Aquatic Ecosystem Dynamics
Model CAEDYM v2.1. (Science manual). Centre for Water Research, The University of Western Australia
143
-------
CCHE1D
Contact Information
Dalmo A. Vieira
National Center for Computational Hydroscience and Engineering
University of Mississippi
School of Engineering
University, MS 38677
(662) 915-6562
dalmo(g),ncche.olemiss.edu
http://www.ncche.olemiss.edu/ccheld/index.html
Download Information
Availability: Nonproprietary (for beta testing program)
Cost: N/A
Model Overview/Abstract
The CCHE1D model is a general one-dimensional model that simulates unsteady flows and sedimentation processes
in channel networks, including bed aggradation and degradation, bed material composition (hydraulic sorting and
armoring), bank erosion, and the resulting channel morphologic changes.
CCHE1D uses a watershed-based approach that provides straightforward integration with existing watershed
processes (rainfall-runoff and field erosion) models to produce estimations of sediment loads and morphological
changes in channel networks. CCHE1D has a GIS-based graphical interface that provides support for automated
spatial analysis, channel network digitizing, digital mapping, and visualization of modeling results.
Model Features
• Unsteady, one-dimensional channel network flow simulation
• Nonequilibrium, nonuniform sediment transport and channel morphology modeling
• Arc View-based graphical user interface (GUI)
• Channel network and subwatershed extraction
• Channel networks digitizing
• Mesh generation
• Data management
• Interface to watershed modeling programs.
Model Areas Supported
Watershed Medium
Receiving Water Medium
Ecological None
Air None
Groundwater None
144
-------
Model Capabilities
Conceptual Basis
The river channels are conceptualized as a one-dimensional channel network.
Scientific Detail
The governing equations for the one-dimensional flow are the St. Venant equations, which describe the conservation
of mass and momentum in the channel network. Preissmann's implicit scheme is applied to solve the governing
equations numerically. The sediment transport component simulates the nonequilibrium transport of nonuniform
sediment with bed load, suspended load, and wash load. The Preissmann's implicit scheme is also applied to solve
the governing equation of sediment transport. Several empirical equations such as the sediment transport capacity,
bed-material porosity, mixing layer thickness, and settling velocities of sediment particles are used to simulate the
sediment transport.
Model Framework
• Horizontal one-dimension model
• Channel network
Scale
Spatial Scale
• One-dimensional channel network
Temporal Scale
• User-defined timestep, typically minutes.
Assumptions
• Laterally and vertically averaged
Model Strengths
The model is able to simulate unsteady flow, including hydraulic structures, and various sediment processes,
including bed aggradation and degradation, bed material composition (hydraulic sorting and armoring), bank
erosion, and the resulting channel morphologic changes. The model also provides interfaces and preprocessing tools
for preparing input data.
Model Limitations
• Only a single outlet is allowed in the dendritic channel networks.
• Flow must be primarily subcritical in all reaches. Local supercritical and transcritical flows without
hydraulic jumps in isolated cross sections are handled through the hybrid dynamic/diffusive wave model.
• Tidal flow conditions are not tested.
• The model cannot be applied to dam-break flows.
Application History
Applications examples: East Fork River, Wyoming; Danjiangkou Reservoir, China; Goodwin Creek Watershed,
Mississippi
Model Evaluation
Application and test of CCHE1D can be found in various technical reports, journals, and conference papers from
http://www.ncche.olemiss.edu/ccheld/ccheld publications.html.
145
-------
Model Inputs
• Geometric data, including cross section, reach length, channel roughness, and channel junctions
• Inflow and outflow
• Sediment properties, inflow sediment data, bed material data, bank material data
• Hydraulic structure data, including bridge crossing, culverts, drop structures, and measuring flumes
• Watershed data
Users' Guide
Available online: http://www.ncche.olemiss.edu/ccheId/ccheId documents.html
Technical Hardware/Software Requirements
Computer hardware:
• PC-Intel or compatible
Operating system:
• Windows95 or newer
Programming language:
• FORTRAN, C, and Avenue
Runtime estimates:
Linkages Supported
• Linked to TOPAZ for spatial analysis
• Linked to AGNPS and SWAT for watershed processes
Related Systems
None
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• ArcViewS.x as the GUI
References
Wu, W.M, D.A. Vieira, 2002, One-Dimensional Channel Network Model CCHE1D version 3.0, Technical Manual,
University of Mississippi, University, MS
Vieira, D.A., W.M Wu. 2002. Users Manual for One-Dimensional Channel Network Model CCHE1D version 3.0,
(Computer program manual). University of Mississippi, University, MS
146
-------
CE-QUAL-ICM/TOXI
Contact Information
Carl F. Cerco
U.S. Army Corps of Engineers
Engineer Research and Development Center
Waterways Experiment Station
Environmental Laboratory
ATTN: CEERD-EP-W
3 909 Halls Ferry Road
Vicksburg, MS 39180
(601)634-4207
cercoc(@,wes.army.mil
http://www.wes.army.mil/el/elmodels
Download Information
Availability: Nonproprietary
Cost: N/A
Model Overview/Abstract
The CE-QUAL-ICM water quality model was initially developed as one component of a model package employed
to study eutrophication processes in Chesapeake Bay. The ICM/TOXI model is the toxic chemical model and has
routines from EPA's WASP (Water Quality Analysis Simulation Program).
Model Features
• Water quality modeling
• Toxics model
• Eutrophication
• Sediment diagenesis model
Model Areas Supported
Watershed None
Receiving Water High
Ecological Low
Air None
Groundwater None
Model Capabilities
Conceptual Basis
There are two distinctly different development pathways to ICM: a eutrophication model (ICM) and an organic
chemical model (ICM/TOXI). The model employs an unstructured grid system, which facilitates linkage to a variety
of hydrodynamic models.
147
-------
Scientific Detail
• ICM stands for "integrated compartment model," which is analogous to the finite-volume numerical
method. The model computes constituent concentrations resulting from transport and transformations in
well-mixed cells that can be arranged in arbitrary one-, two-, or three-dimensional configurations. Thus, the
model employs an unstructured grid system.
• The release version of the eutrophication model computes 22 state variables, including physical properties;
multiple forms of algae, carbon, nitrogen, phosphorus, and silica; and dissolved oxygen. Recently, two size
classes of zooplankton, two benthos compartments (deposit feeders and filter feeders), submerged aquatic
vegetation (roots and shoots biomass), epiphytes, and benthic algae were added, although this version of
the code is not generally released to the public.
• Each state variable may be individually activated or deactivated.
• One significant feature of ICM, eutrophication version, is a diagenetic sediment submodel. The sub-model
interactively predicts sediment-water oxygen and nutrient fluxes. Alternatively, these fluxes may be
specified based on observations.
• The ICM/TOXI model resulted from incorporating the toxic chemical routines from EPA's WASP (Water
Quality Analysis Simulation Program) model into the transport code for ICM, incorporating a more
detailed benthic sediment model, and enhancing linkages to sediment transport models.
• ICM/TOXI includes physical processes, such as sorption to DOC and three solid classes, volatilization, and
sedimentation. It also includes chemical processes such as ionization, hydrolysis, photolysis, oxidation, and
biodegradation.
• ICM/TOXI can simulate temperature, salinity, three solids classes, and three chemicals (total chemical for
organic chemicals and trace metals). Each species can exist in five phases (water, DOC-sorbed, and sorbed
to three solids types) via local equilibrium partitioning.
Model Framework
The model consists of a main program, an INCLUDE file, and subroutines. Both the main program and subroutines
perform read and write operations on numerous input and output files. The model does not compute hydrodynamics.
Flows, diffusion coefficients, and volumes must be specified externally and read into the model. Hydrodynamics
may be specified in binary or ASCII format and may be obtained from a hydrodynamics model such as the CH3D-
WESorEFDC.
Scale
Spatial Scale
• One-, two-, or three-dimensional
Temporal Scale
• User-defined timestep. The timestep may be varied through the auto-stepping option or at discrete, user-
specified intervals.
Assumptions
The model assumes that the dynamics of each physical, chemical, and biological component can be described by the
principle of conservation of mass.
Model Strengths
• The model has a predictive diagenetic sediment submodel that interactively predicts sediment-water
oxygen and nutrient fluxes.
• The ICM/TOXI model has toxic chemical routines, which further enhance linkage with sediment transport
models.
148
-------
Model Limitations
• The model does not compute hydrodynamics. Flows, diffusion coefficients, and volumes must be specified
externally and read into the model.
• The user must provide processors, which prepare input files and process output for interpretation and
presentation.
Application History
The ICM eutrophication model has been applied to a variety of sites, including Chesapeake Bay, Inland Bays of
Delaware, New York Bight, Newark Bay, New York-New Jersey Harbors and Estuaries, Lower Green Bay, Los
Angeles-Long Beach Harbors, Cache River wetland, San Juan Bay and Estuaries, Florida Bay, and Lower St. Johns
River (on-going).
The WASP toxic chemical model on which ICM/TOXI is based has been applied to a wide variety of sites. CE-
QUAL-ICM also has been linked to the CH3D-WES and EFDC hydrodynamic models.
Model Evaluation
Not available
Model Inputs
• Initial conditions
• Time sequences of boundary conditions (inputs from watershed sources and discharges)
• Reservoir geometry
• Physical coefficients
• Biological and chemical reaction rates
• Time sequences of meteorological data used to compute temperature.
Users' Guide
Available online: http://www.wes.army.mil/el/elpubs/pdf/trel95-15.pdf
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS. Operates on a variety of platforms, including 486 PC, Silicon Graphics and Hewlett Packard
workstations, and Cray Y-MP and C-90 mainframes
Programming language:
• FORTRAN 77
Runtime estimates:
• Minutes to hours
Linkages Supported
CH3D-WES, EFDC
149
-------
Related Systems
Surface Water Modeling System (SMS)
Sensitivity/Uncertainty/Calibration
Highly accurate for simulation of reservoir systems with adequate monitoring data and application experience.
Model Interface Capabilities
Not available
References
Creco, C.F. 1995. Simulation of trends in Chesapeake Bay Eutrophication. Journal of Environmental Engineering.
121(4):298-310.
Cerco, C.F., and T. Cole. 1993. Three-dimensional eutrophication model of Chesapeake Bay. Journal of
Environmental Engineering. (119): 1006-1025.
Cerco, C.F., and T. Cole. 1994. Three-dimensional eutrophication model of Chesapeake Bay. Technical Report EL-
94-4. US Army Corps of Engineers Water Experiment Station, Vicksburg, MS.
Cerco, C.F., and T. Cole. 1995. User's Guide to the CE-QUAL-ICM Three-dimensional eutrophication model,
release version 1.0. Technical Report EL-95-15. US Army Corps of Engineers Water Experiment Station,
Vicksburg, MS.
DiToro, D.M., and J.F. Fitzpatrick (Hydroqual, Inc.). 1993. Chesapeake Bay sediment flux model. Contact Report
EL-93-2. Prepared for EPA Chesapeake Bay Program, US. Army Engineers District, Baltimore, MD, and US. Army
Engineer Waterways Exp. Station by Hydroqual, Inc.
Mark, D., B. Bunch, and N. Scheffner. 1992. Combined hydrodynamic and water quality modeling of Lower Green
Bay. Miscellaneous Paper W-92-3. In Proceedings of Water Quality '92 9th Seminar, Environmental Laboratory,
Army Engineers Waterways Experiment Station, Vicksburg, MS, pp 226-233.
150
-------
CE-QUAL-R1
Contact Information
Dorothy H. Tillman
Environmental Laboratory
U.S. Army Corps of Engineers
Engineer Research and Development Center
Waterways Experiment Station
3 909 Halls Ferry Road
Vicksburg, MS 39180
(601)634-3670
tillmad(5),we s. army. mil
http ://www. wes.army. mil/el/elmodels/index. html#wqmodels
Download Information
Availability: Nonproprietary download currently for U.S. Army Corps of Engineers use only
Cost: N/A
Model Overview/Abstract
CE-QUAL-Rl is a water quality model that describes the time-variable vertical distribution of 27 water quality
variables in reservoirs. In addition, 11 variables associated in the sediment can be modeled. CE-QUAL-Rl is
spatially one-dimensional and horizontally averaged; temperature and concentration gradients are computed only in
the vertical direction. The reservoir is conceptualized as a vertical sequence of horizontal layers in which thermal
energy and materials are uniformly distributed in each layer. The mathematical structure of the model is based on
horizontal layers whose thickness depends on the balance of inflowing and outflowing waters. Variable layer
thicknesses permit accurate mass balancing during periods of inflow and outflow.
Model Features
• Temperature
• Dissolved oxygen
• Nutrients
• Metals
• Algaes
• Macrophytes
• Fish
• pH, alkalinity
• Coliforms
Model Areas Supported
Watershed None
Receiving Water Medium
Ecological Medium
Air None
Groundwater None
151
-------
Model Capabilities
Conceptual Basis
In CE-QUAL-R1, the reservoir is conceptualized as a vertical sequence of horizontal layers in which thermal energy
and materials are uniformly distributed in each layer.
Scientific Detail
The mathematical structure is based on a set of differential equations that express conservation of mass and energy
in each horizontal layer. Solution of these equations provides material or energy concentrations as functions of time
and depth. The thermal analysis portion of CE-QUAL-R1 is provided as an independent model (CETHERM-R1) to
simplify simulation of water budgets and temperature profiles. CETHERM-R1 includes the variables of temperature,
suspended solids, and total dissolved solids. The algorithms representing physical processes are the same as in CE-
QUAL-R1. The flux model calculates and lists the rates of change for all biological processes, which should aid the
user in correctly predicting variable concentrations.
Model Framework
• One-dimensional vertical waterbody
• Deep reservoirs
Scale
Spatial Scale
• One-dimensional vertical
Temporal Scale
• User-defined timestep, typically seconds
Assumptions
• Laterally and longitudinally averaged reservoir layers
• Well-mixed in each layer
Model Strengths
The model is capable of simulating physical, chemical, and biological factors, including radiation and heat transfer;
conservative substance routing; suspended solids routing and settling; dissolved oxygen through aeration,
photosynthesis, respiration, and organic decomposition; carbon, nitrogen, and phosphorus cycles; dynamics and
trophic relationships of phytoplankton and macrophytes; coliform bacteria mortality; and accumulation, dispersion,
and reoxidation of manganese, iron, and sulfide when anaerobic conditions prevail.
Model Limitations
Limitations include the one-dimensional representation of reservoirs that limits simulation to a vertical series of
well-mixed horizontal layers. This assumption cannot predict longitudinal and lateral variations in water quality and
requires the assumption of instantaneous dispersion of all inflow quantities and constituents throughout the
horizontal layers. The model assumes that the dynamics of each physical, chemical, and biological component can
be described by the principle of conservation of mass. Because the equations are not solved in closed form, minor
errors concerning the conservation of mass can occur. The dynamic calculation of nutrient flux from bottom
sediments is not included.
Application History
Application example: Eau Galle Reservoir, Wisconsin
152
-------
Model Evaluation
Not available
Model Inputs
• Initial conditions
• Time sequences of boundary conditions (inputs from watershed sources and discharges)
• Reservoir geometry
• Physical coefficients
• Biological and chemical reaction rates
• Time sequences of hydrometeorological conditions
Users' Guide
Available online: http://www.wes.army.mil/el/elmodels/pdf/ire82-l/ire82-l.pdf
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS
Programming language:
• FORTRAN
Runtime estimates:
• Minutes to hours
Linkages Supported
None
Related Systems
None
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• WES WIN is an interactive Windows package to execute CE-QUAL-R1.
• WESPLOT is for post-processing the modeling results.
References
U.S. Army Corps of Engineers, South Florida Water Management District and Kimley-Horn and Associates, Inc.
2003. Comprehensive Everglade Restoration Plan, G.I. - Water Quality Modeling, G.I.I. - Reservoir Phosphorus
Uptake Model - Interim Report. U.S. Army Corps of Engineers, South Florida Water Management District.
153
-------
Environmental Laboratory. 1990. A dynamic one-dimensional (longitudinal) water quality model for streams: User's
manual. CE-QUAL-R1. U.S. Army Corps of Engineers Waterways Experiment Station, Vicksburg, MS.
Nestler, J.M., Lt. Scheider, and B.R. Hall. 1993. Development of a simplified approach for assessing the effects of
water release temperatures on tailwater habitat downstream of Fort Peck, Garrison, and Fort Randall Dams.
Technical Report EL-93-23. U.S. Army Corps of Engineers Waterways Experiment Station, Vicksburg, MS.
Zimmerman, M.J., and M.S. Dortch. 1989. Modeling water quality of a regulated stream below a peaking
hydropower dam. Regulated Rivers: Research and Management. 4: 235-247.
154
-------
CE-QUAL-RIV1
Contact Information
Toni Toney
U.S. Army Corps of Engineers
Engineer Research and Development Center
Waterways Experiment Station
Environmental Laboratory
ATTN: CEERD-EP-W
3 909 Halls Ferry Road
Vicksburg, MS 39180
Toni.Tonevfg.erdc.usace.armv.mil
http://www.wes.armv.mil/el/elmodels/rivlinfo.html
Download Information
Availability: Nonproprietary download currently for U.S. Army Corps of Engineers use only
Cost: N/A
Model Overview/Abstract
CE-QUAL-RIVl is a coupled one-dimensional hydrodynamic and water quality model for river systems. The model
consists of two parts, a hydrodynamic code (RIV1H) and a water quality code (RIV1Q). RIV1H is applied first to
predict flows, depths, velocities, water surface elevations, and other hydraulic characteristics. RIV1Q uses the
RIV1H output file to drive the transport of the water quality variables. RIV1Q can predict variations in each of 12
state variables—temperature, carbonaceous biochemical oxygen demand (CBOD), organic nitrogen, ammonia
nitrogen, nitrate + nitrite nitrogen, dissolved oxygen, organic phosphorus, dissolved phosphates, algae, dissolved
iron, dissolved manganese, and coliform bacteria. In addition, the impacts of macrophytes can be simulated.
Model Features
• One-dimensional unsteady flow
• Nutrients, dissolved oxygen dynamics, eutrophication, metals, and bacteria
Model Areas Supported
Watershed None
Receiving Water High
Ecological Medium
Air None
Groundwater None
Model Capabilities
Conceptual Basis
In CE-QUAL-RIVl, the river is conceptualized as a series of horizontal grids in one-dimension.
155
-------
Scientific Detail
The governing equations for CE-QUAL-RIV1 consist of continuity equation, momentum equation, and constituent
fate and transport equation. The Preissmann's implicit scheme is applied to solve the governing equations for flow.
The advection of the constituent transport is solved using a fourth-order explicit scheme. The diffusion of the
constituent transport is solved using an implicit scheme. The source/sink and reaction of water quality variables are
solved separately.
Model Framework
• One-dimensional horizontal model
• River
Scale
Spatial Scale
• One-dimensional horizontal
Temporal Scale
• User-defined timestep
• Time-variable simulation of flow and water quality variables
Assumptions
• No lateral circulation
• Hydrostatic assumption
• No lateral variation of water quality variables
Model Strengths
• Provides a high-order accuracy advection scheme to deal with various flow conditions
• Is capable of simulating physical, chemical, and biological processes in rivers
Model Limitations
• Cannot model a one-dimensional tidal system
• Does not include a sediment diagenesis component
Application History
Application example: water temperature and dissolved oxygen simulation of Youghiogheny River, Pennsylvania
Model Evaluation
Not available
Model Inputs
• Initial conditions
• Boundary conditions
• Segmentation of river network
• Physical and biological parameters
Users' Guide
Available online: http://www.wes.army.mil/el/elmodels/pdf/el952/el-95-2.pdf
156
-------
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS, Windows
Programming language:
• FORTRAN
Runtime estimates:
• Minutes to hours
Linkages Supported
None
Related Systems
EFDC1D, DYNHYD5/WASP5
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
Not available
References
Environmental Laboratory. CE-QUAL-RIV1: a dynamic, one-dimensional (longitudinal) water quality model for
streams. (Computer program manual). U.S. Army Corps of Engineers.
157
-------
CE-QUAL-W2
Contact Information
Thomas M. Cole
U.S. Army Corps of Engineers
Engineer Research and Development Center
Waterways Experiment Station
Environmental Laboratory
ATTN: CEERD-EP-W
3 909 Halls Ferry Road
Vicksburg, MS 39180
(601)634-3283
tcole(@,lasher. wes.army.mil
http ://www. wes.army. mil/el/elmodels/index. html
Scott Wells
Professor
Department of Civil and Environmental Engineering
Portland State University
P.O. Box 751
Portland, OR 97207-0751
(503) 725-4276
scott(@,cecs.pdx.edu
http://www.cee.pdx.edu/w2/
Download Information
Availability: Nonproprietary
Cost: N/A
Model Overview/Abstract
CE-QUAL-W2 is a two-dimensional, longitudinal/vertical, laterally averaged, finite-difference hydrodynamic and
water quality model. Because the model assumes lateral homogeneity, it is best suited for relatively long, narrow
waterbodies exhibiting longitudinal and vertical water quality gradients. The model can be applied to rivers, lakes,
reservoirs, and estuaries. Branched networks can be modeled.
The model accommodates variable grid spacing (segment lengths and layer thicknesses) so that greater resolution in
the grid can be specified where needed. The model equations are based on the hydrostatic approximation (negligible
vertical accelerations). Eddy coefficients are used to model turbulence. The hydrodynamic timestep is calculated
internally as the maximum allowable timestep that ensures numerical stability. A third-order accurate (QUICKEST)
advection scheme reduces numerical diffusion.
The water quality portion of the model includes the major processes of eutrophication kinetics and a single algal
compartment. The bottom sediment compartment stores settled particles, releases nutrients to the water column, and
exerts sediment oxygen demand based on user-supplied fluxes; a full sediment diagenesis model is under
development.
Model Features
• Reservoir and river hydrodynamics and transport
158
-------
• Temperature simulation
• Water quality modeling
• Eutrophication
Model Areas Supported
Watershed None
Receiving water High
Ecological Low
Air None
Groundwater None
Model Capabilities
Conceptual Basis
In CE-QUAL-W2, the river or reservoir is conceptualized as a laterally averaged two-dimensional model having
horizontal segments and vertical layers.
Scientific Detail
The mathematical structure is based on a set of differential equations that express conservation of mass and energy
in each horizontal layer. Solution of these equations provides material or energy concentrations as functions of time
and depth. Some of the key model features are given below:
• Variable layer heights and segment lengths between waterbodies; surface layer can extend through multiple
layers
• Ability to model multiple waterbodies in the same computational grid, including multiple reservoirs,
steeply sloping riverine sections between reservoirs, and estuaries
• Momentum transfer between branches
• Additional reaeration algorithms for rivers
• Additional vertical turbulence algorithms for rivers
• Numerical algorithms for pipe, weir, spillway, and pump flow
• Internal weir algorithm for submerged or skimmer weirs
• The effect of hydraulic structures on gas transfer and total dissolved gas transport
• Algorithm to calculate the maximum allowable timestep and adjust the timestep to ensure hydrodynamic
stability requirements were not violated (autostepping)
• A selective withdrawal algorithm that calculates a withdrawal zone based on outflow, outlet geometry, and
upstream density gradients
• Higher-order transport scheme (QUICKEST) that reduces numerical diffusion. Leonard's ULTIMATE
algorithm, which eliminates over/undershooting in the transport solution scheme, has been added. The
QUICKEST can be used alone or in conjunction with the ULTIMATE scheme, where oscillations due to
over/undershoots are encountered.
• Step function or linear interpolation of inputs
• Ice-cover algorithm
• Internal calculation of equilibrium temperatures and coefficients of surface heat exchange or a term-by-
term accounting of surface heat exchange
• Generalized time-varying data input subroutine with input data accepted at any frequency
• Sediment/water heat exchange
• Any number of user-defined arbitrary constituents defined by a decay rate, settling rate, and temperature
rate multiplier that can include conservative tracers, coliform bacteria, water age, and contaminants
• An implicit solution for the effects of vertical eddy viscosity in the horizontal momentum equation
• Momentum transfer between branches
• Any number of user-defined phytoplankton, epiphyton, CBOB, and inorganic solids groups
• Dissolved and paniculate biogenic silica
159
-------
• Derived constituents, such as total DOC, organic nitrogen, and organic phosphorus, that are not state
variables
• Ability to output kinetic fluxes
Model Framework
The model predicts water surface elevations, velocities, and temperatures. Temperature is included in the
hydrodynamic calculations because of its effect on water density. The water quality algorithms incorporate 21
constituents in addition to temperature, including nutrient/phytoplankton/dissolved oxygen interactions during
anoxic conditions. Any combination of constituents can be simulated. The effects of salinity or total dissolved
solids/salinity on density and thus on hydrodynamics are included only if they are simulated in the water quality
module. The water quality algorithm is modular, allowing constituents to be easily added as subroutines. The model
can be applied to estuaries, rivers, or portions of a waterbody by specifying upstream or downstream head boundary
conditions. The branching algorithm allows application to geometrically complex waterbodies such as dendritic
reservoirs or estuaries. Variable segment lengths and layer thicknesses can be used, allowing specification of higher
resolution where needed. Water quality can be updated less frequently than hydrodynamics, thus reducing
computational requirements. However, water quality kinetics are not decoupled from the hydrodynamics (i.e.,
separate, standalone code for hydrodynamics and water quality where output from the hydrodynamic model is stored
on disk and then used to specify advective fluxes for the water quality computations).
Scale
Spatial Scale
• Two-dimensional in the horizontal and vertical direction, with any number of waterbodies having any
number of branches
• Previous applications have used a horizontal grid spacing of 100 to 10,000 m and a vertical grid spacing of
0.2 to 5 m
Temporal Scale
• User-defined timestep
Assumptions
• Dynamics of each physical, chemical, and biological component can be described by the principle of
conservation of mass.
• Waterbody is laterally averaged.
Model Strengths
• Can model hydrodynamics and water quality for entire river basin with dams, river, and lakes, with a
variety of hydraulic structures.
• A higher-order transport scheme (QUICKEST) that reduces numerical diffusion is included.
Model Limitations
• Well-mixed in lateral direction.
• Hydrostatic assumption for vertical momentum equation.
• Considerable technical expertise in hydrodynamics and eutrophication/water quality processes is required
to apply the model.
• Simplistic sediment oxygen demand; the model does not have a sediment diagenisis compartment.
• No zooplankton or macrophyte.
• No toxics.
• Pre-processor of the model only helps in creating the main control file and bathymetry visualization and
editing. No bathymetry grid generator or data display/preparation or post-processing capabilities.
160
-------
Application History
The model has been successfully applied to more than 100 waterbodies. Apart from the example applications
provided in the user's manual, a partial list of CE-QUAL-W2 applications can be found at
http://www.ce.pdx.edu/w2/w2app.htm.
Model Evaluation
Unknown
Model Inputs
• Initial conditions
• Time sequences of boundary conditions (inputs from watershed sources and discharges)
• Reservoir geometry
• Physical coefficients
• Biological and chemical reaction rates
• Time sequences of hydrometeorological conditions
Users' Guide
User documentation can be downloaded along with the model: http://www.ce.pdx.edu/w2 or
http ://www. wes.army. mil/el/elmodels/index. html
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS/Windows or UNIX environment
Programming language:
• FORTRAN90
Runtime estimates:
• Minutes to hours
Linkages Supported
HSPF
Related Systems
Watershed Modeling System (WMS)
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Graphical pre-processor
161
-------
References
Cole, T.M., and S. A. Wells. 2000. CE-QUAL-W2: A two-dimensional, laterally averaged, hydrodynamic, and water
quality model. Version 2.0. Instructional Report EL-95-1. U.S. Army Engineer Waterways Experiment Station,
Vicksburg, MS.
Cole, T.M., and S. A. Wells. 2003. CE-QUAL-W2: A two-dimensional, laterally averaged, Hydrodynamic and
Water Quality Model, Version 3.1. Instruction Report EL-03-1. US Army Engineering and Research Development
Center, Vicksburg, MS.
162
-------
CH3D-IMS: Curvilinear-grid Hydrodynamics 3D—Integrated Modeling
System
Contact Information
Y. Peter Sheng
Coastal & Oceanographic Engineering
University of Florida
Gainesville, FL 32611-6590
(352) 392-6177
pctc(@,coastal.ufl.edu
http://users.coastal.ufl.edu/~pete/CH3D/
Download Information
Availability: Proprietary
Cost: Contact Dr. Y. Peter Sheng at pete@coastal.ufl.edu
Model Overview/Abstract
CH3D-IMS is an integrated modeling system based on a CH3D model framework. It models circulation, wave,
sediment transport, water quality, light attenuation, and seagrass on curvilinear grids. The circulation is solved using
CH3D. Wave is modeled using SWAN framework. Four additional modules: CH3D-SED3D, CH3D-WQ3D,
CH3D-LA, and CH3D-SAV are used for calculating the sediment transport, water quality, light attenuation, and
seagrass.
Model Features
• Three-dimensional hydrodynamics
• Cohesive and noncohesive sediment transport
• Nitrogen, phosphorus, phytoplankton, zooplankton, and dissolved oxygen
• Light attenuation and seagrass kinetics
Model Areas Supported
Watershed None
Receiving Water High
Ecological High
Air None
Groundwater None
Model Capabilities
Conceptual Basis
The waterbody is conceptualized as a series of grid points on a curvilinear orthogonal coordinate system.
Scientific Detail
The details of the circulation model are presented in the fact sheet of CH3D. The governing equations for sediment
transport and water quality are solved on a nonorthogonal curvilinear coordinate on the horizontal plane. In the
vertical direction, both the o-coordinate and z-coordinate are provided. Sediment transport and water quality are
solved using the same timestep as the hydrodynamic calculation. The sediment transport processes include
163
-------
advection, turbulent mixing, settling/flocculation, deposition, and resuspension. Wave-current interaction inside the
bottom boundary layer also is considered. For water quality simulation, the nitrogen cycling models dissolved,
paniculate, organic, and inorganic nitrogen species; the phosphorus cycling models dissolved, paniculate, organic,
and inorganic phosphorus species. In addition, dissolved oxygen, phytoplankton, and zooplankton are modeled. The
seagrass module calculates the growth and decay of seagrass biomass due to light, nutrient, temperature, and
salinity.
Model Framework
• Three-dimensional model
• River, lake, reservoir, estuary, ocean
Scale
Spatial Scale
• One-dimensional, two-dimensional, and three-dimensional
Temporal Scale
• User-defined timestep
Assumptions
• Hydrostatic assumption
• Boussinesq approximation
• Reynold's stress assumption
Model Strengths
• Capable of modeling one-dimensional, two-dimensional, and three-dimensional hydrodynamics, sediment
transport, and eutrophication in various waterbodies with complex bathymetry.
• Boundary-fit curvilinear coordinate can represent the waterbody boundaries accurately.
• The o and z coordinates in the vertical direction provides flexible options for modeling waterbodies with
different bathymetry.
Model Limitations
• Not a public modeling system.
• No source code available.
• No user's manual.
Application History
The model has been applied in Chesapeake Bay, James River, Lake Okeechobee, Sarasota Bay, Tampa Bay, Indian
River Lagoon, Florida Bay, St. Johns River, Biscayne Bay, Charlotte Harbor, Gulf of Mexico, and Pinellas County
and offshore.
Model Evaluation
The model has been evaluated in many journal and conference papers.
Model Inputs
• Initial conditions
• Time sequences of boundary conditions (inputs from watershed sources and discharges)
• Reservoir geometry
164
-------
• Physical coefficients
• Biological and chemical reaction rates
• Time sequences of hydrometeorological conditions
Users' Guide
Not available
Technical Hardware/Software Requirements
Computer hardware:
• VAX, SGI, SUN, DEC, IBM, IBM-PC
Operating system:
• PC-DOS, UNIX, WINDOWS, Linux
Programming language:
• FORTRAN
Runtime estimates:
• Minutes to hours
Linkages Supported
None
Related Systems
POM, EFDC, ECOMSED, GLLVHT
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Arc View-based GUI for grid generation, pre- and post- processing
References
Sheng, Y.P. 1986. A Three-Dimensional Mathematical Model of Coastal, Estuarine and Lake Currents Using
Boundary-Fitted Grid. Technical Report No. 585. Aeronautical Research Associates of Princeton, Princeton, New
Jersey.
Sheng, Y. P. 1989. On modeling three dimensional estuarine and marine hydrodynamics. Ed. Nihoul, J. C. J.,
Elsevier Oceanography Series, 45. 35-54.
Sheng, Y.P., D.E. Eliason, X.-J. Chen, and J.-K. Choi. 1991. A Three-Dimensional Numerical Model of
Hydrodynamics and Sediment Transport in Lakes and Estuaries: Theory, Model Development, and Documentation.
Final Report. Environmental Research Laboratory, U.S. Environmental Protection Agency, Athens, GA.
Sheng, Y.P. 1994. Modeling Hydrodynamics and Water Quality Dynamics in Shallow Waters. In Proceedings of the
International Symposium on Ecology and Engineering, Taman Negara, Malaysia, November, 1994.
165
-------
CH3D-SED (& CH3D-WES): Curvilinear Hydrodynamics in Three
Dimensions
Contact Information
Phu Luong
U.S. Army Engineer Corp of Engineers
Waterways Experiment Station
Coastal and Hydraulics Laboratory
3 909 Halls Ferry Road
Vicksburg, MS 39180-6199
(601)634-4472
Phu.V.Luong(@,erdc.usace.army.mil
Download Information
Availability: Available only to Department of Defense Agencies
Cost: N/A
Model Overview/Abstract
CH3D-SED is the newly developed mobile bed version of CH3D-WES, which was developed for the Chesapeake
Bay Program. The USAGE is using it to investigate sedimentation on bendways, crossings, and distributaries on the
lower Mississippi and Atchafalaya rivers. These applications address dredging, channel evolution, and channel
training structure evaluations. CH3D-SED functions as a hydrodynamic (through the incorporation of CH3D-WES)
and sediment transport model. Physical processes affecting circulation and vertical mixing that can be modeled
include tides, wind, density effects (salinity and temperature), freshwater inflows, turbulence, and the effect of the
earth's rotation. CH3D-SED can be applied to rivers, lakes, reservoirs, estuaries, or coastal waters.
Model Features
• Hydrodynamic
• Sediment transport
• Linkage to CE-QUAL-IC water quality model
Model Areas Supported
Watershed None
Receiving Water High
Ecological None
Air None
Groundwater None
Model Capabilities
Conceptual Basis
The CH3D-SED hydrodynamic and sediment transport model is based on an extension of the stretched vertical
coordinate version of the CH3D-SED by Spasojevic and Holly (1997) to include cohesive sediment transport. The
model is capable of two- or three-dimensional operation and employs standard formulations for settling, deposition,
and resuspension.
166
-------
CH3D-WES makes hydrodynamic computations on a curvilinear or boundary-fitted platform grid. Deep navigation
channels and irregular shorelines can be modeled because of the boundary-fitted coordinates feature of the model.
Vertical turbulence is predicted by the model and is crucial to a successful simulation of stratification,
destratification, and anoxia. A second-order model based on the assumption of local equilibrium of turbulence is
employed.
A boundary-fitted, nonorthogonal, finite-difference approximation in the horizontal plane and a sigma-stretched
approximation in the vertical direction are used for the approximations of the governing equations.
Scientific Detail
The hydrodynamic model solves the depth-averaged Reynolds approximation of the momentum equation for
velocity and the depth-averaged conservation of mass equation for water surface elevation. The three-dimensional
velocity field is determined by computing the deviation from the depth-averaged velocity by solving the
conservation of mass equation in conjunction with a k-e closure for vertical momentum diffusion.
Sedimentation computations are based on a two-dimensional solution of the conservation of mass equation for the
channel bed, and three-dimensional advection-diffusion equation for suspended sediment transport. The sediment
transport algorithms independently account for the movement of sediment as either bed load or suspended load, as
well as the exchange of sediment between these two modes of transport. The model is also generalized for
application to mixed-grain-size sediments, with appropriate bed material sorting and armoring routines. The
formulation to a user-specified multiple-grain-size distribution uniquely allows the simulation of erosion,
entrainment, transport, and deposition of contaminated sediments on the bed and in the water column. A
contaminated sediment associated with a given grain size can be independently accounted for by applying a small
dimensional perturbation from the reference grain size. This perturbation has negligible effects on sediment mobility
characteristics. Because each grain size specification is independently tracked, however, tracking of zones of
contaminated bed material is possible.
Model Framework
• Curvilinear Finite Difference Numerical Formulations for Hydrodynamic and Sediment Transport
Scale
Spatial Scale
• Three-dimensional
Temporal Scale
• Dynamic
Assumptions
• Based on accepted formulations of the three-dimensional, hydrostatic, hydrodynamic equations and
conservative transport equation.
Model Strengths
• Strong capabilities for hydrodynamics and sediment transport.
Model Limitations
• Considerable technical expertise in hydrodynamics is required to use the model effectively.
167
-------
Application History
CH3D-SED was applied to investigate maintenance dredging quantities for channel alignment studies on the lower
Atchafalaya River (Hall 1996). The model successfully reproduced existing sediment deposition quantities and
locations, and it was instrumental in the decision made by the local sponsor to maintain the existing channel
alignment.
The capabilities of CH3D are illustrated by its application to the Chesapeake Bay. The numerical grid employed in
the Chesapeake Bay model has 734 active horizontal cells and a maximum of 15 vertical layers, resulting in 3,992
computational cells. Grid resolution is 1.52 m vertical and approximately 10 km longitudinal and 3 km lateral. The
x, y coordinates of the grid are transformed into the c-curvilinear coordinates to allow for better handling of the
complex horizontal geometries. Velocity is also transformed so that its components are perpendicular to the c-
coordinate lines, thus allowing boundary conditions to be prescribed on a boundary-fitted grid in the same manner as
on a Cartesian grid. Major tributaries are modeled three-dimensionally in the lower reach of the bay and two-
dimensionally with constant width in the upper reach.
Cerco et al. (1993) used CH3D-WES in conjunction with CE-QUAL-ICM to predict water column and sediment
processes that affect water quality in the Chesapeake Bay. Data from 1984-1986 were used, and the linked modeling
approach was successful in predicting the spring algal bloom, onset and breakup of summer anoxia, and coupling of
organic particle deposition with sediment-water nutrient and oxygen fluxes.
Model Evaluation
Johnson et al. (1993) validated the model by applying it to six datasets. The first three datasets contained
approximately one month's worth of data each and represented a dry summer condition, a spring runoff, and a fall
wind-mixing event. The last three applications were yearlong simulations for 1984 (a wet year), 1985 (a dry year),
and 1986 (an average year). Results demonstrate that the model is a good representation of the hydrodynamics of the
Chesapeake Bay and its major tributaries.
Model Inputs
• Time-varying water-surface elevations, salinity, and temperature conditions at the ocean entrance and at
freshwater inflows at the head of all tributaries.
• Time-varying wind and surface heat exchange data at one or more locations.
• All input data, including initial conditions, bathymetry, boundary, and computational control data are input
from fixed files.
Users' Guide
Not available
Technical Hardware/Software Requirements
Computer hardware:
• Unix workstation or super computer.
Operating system:
• Unix
Programming language:
• FORTRAN
Run time estimates:
• Compute intensive
168
-------
• Run time is highly dependent on computer hardware, model domain spatial resolution, the period of
prototype conditions simulated and other options such as whether the model is simulation only
hydrodynamic or hydrodynamics and the fate and transport of dissolved and suspended material. Under this
wide range of variability, simulations could require minutes to weeks.
Linkages Supported
CE-QUAL-ICM
Related Systems
Surface Water Modeling System (SMS)
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• US Army Corps of Engineers Surface Water Modeling System (SMS)
References
Cerco, C.F. and T. Cole. 1993. Three-Dimensional Eutrophication Model of Chesapeake Bay. Journal of
Environmental Engineering. 119(6): 1006-1025.
Chapman, Raymond S., Billy H. Johnson, and S. Rao Vemulakonda. 1996. User's Guide for the Sigma Stretched
Version of CH3D-WES. Technical Report HL-96-21. U.S. Army Engineer Waterways Experiment Station,
Vicksburg, MS.
Engel, John J., Rollin H. Hotchkiss, and Brad R. Hall. 1995. Three Dimensional Sediment Transport Modeling
Using CH3D Computer Model. In Proceedings of the First International Water Resources Engineering Conference,
William H. Espey Jr. and Phil G. Combs, ed., American Society of Civil Engineers, New York, 1995, pp. 628-632.
Hall, Brad R. 1996. Quantifying Sedimentation Using a Three Dimensional Sedimentation Model. In Proceedings of
Water Quality "96, llth Seminar, Corps of Engineers Committee on Water Quality, Seattle, WA, 1996, pp. 88-93.
Johnson, B.H., K. W. Kim, R.E. Heath, B.B. Hsieh, and H.L. Butler. 1993. Validation of Three-Dimensional
Hydrodynamic Model of Chesapeake Bay. Journal of Hydraulic Engineering. 119(1):2-20.
Johnson, B.H., R.E. Heath, B.B. Hsieh, K.W. Kim, H.L. Butler. 1991. User's Guide for a Three-Dimensional
Numerical Hydrodynamic, Salinity, and Temperature Model of Chesapeake Bay. (Computer program manual).
Department of the Army, Waterways Experiment Station, Corps of Engineers, Vicksburg, MS.
Spasojevic, Miodrag and Forrest M. Holly. 1994. Three-Dimensional Numerical Simulation of Mobile-Bed
Hydrodynamics. Contract Report HL-94-2. U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
Van Rijn, Leo C. 1984a. Sediment Transport, Part I: Bed Load Transport. Journal of Hydraulic Engineering, ASCE.
110(10):1431-1456.
Van Rijn, Leo C. 1984b. Sediment Transport, Part II: Suspended Load Transport. Journal of Hydraulic Engineering,
ASCE. 110(11):1613-1641.
169
-------
DELFT3D
Contact Information
WL | Delft Hydraulics
Rotterdamseweg 185
P.O. Box 177
2600 MH Delft
The Netherlands
+31 152858585
delft3d.info(g),wlderft.nl
www.wlderft.nl/d3d
Download Information
Availability: Proprietary
Cost: Varies by user type (academic, governmental, private)
DelftSD may be purchased as an integrated package or in individual modules. The GUI will be part of any initial
purchase. WL | Delft Hydraulics will take care of the set-up and calibration of the model applications for clients.
Model Overview/Abstract
DelftSD models two- and three-dimensional flow and transport for tidal and riverine problems. DelftSD is an
integrated modeling environment for hydrodynamics, waves, sediment transport, morphology, water quality, and
ecology. Relevant applications include
• Prediction of nearshore currents and waves
• Prediction of small scale morphology
• Estimation of bathymetry
• Prediction of large-scale morphology based on historical observations
• Prediction of short- and long-term effects of structures
• Prediction of the effects of dike or dam breaches
The model can be applied to rivers, lakes, reservoirs, estuaries, and coastal waters. DeftSD can be applied to
evaluate
• Hydrodynamics—salt intrusion; river flow simulations; fresh water river discharges in bays; thermal
stratification in lakes, seas and reservoirs; cooling water intakes and waste water outlets; transport of
dissolved material and pollutants; tide and wind driven flows (i.e., storm surges); stratified and density
driven flows; and wave driven flows
• Sediment transport—transport of cohesive and non-cohesive sediments, e.g., spreading of dredged
materials to study sediment/erosion pattern.
• Contaminant transport— evaluation of over 140 substances including contaminants in 2 sediment layers.
Model Features
• Hydrodynamics
• Waves
• Sediment transport
• Morphology
• Water quality
• Particle tracking for water quality
170
-------
• Ecology
Model Areas Supported
Watershed None
Receiving Water High
Ecological None
Air None
Groundwater None
Model Capabilities
Conceptual Basis
Scientific Detail
DelftSD includes the following modules:
• Hydrodynamics module (Delft3D-FLOW)
• Wave module (Delft3D-WAVE)
• Water quality module (Delft3D-WAQ)
• Particle tracking module (Delft3D-PART)
• Ecological module (DelftSD-ECO)
• Sediment transport module (Delft3D-SED)
• Morphdynamic module (Delft3D-MOR)
Hydrodynamic
The FLOW module of DelftSD is a multi-dimensional (two- or three-dimensional) hydrodynamic (and transport)
simulation program that calculates non-steady flow and transport phenomena resulting from tidal and meteorological
forcing on a curvilinear, boundary-fitted grid. Features of the FLOW module include
• Coriolis force
• Advection-diffusion solver included to compute, for example, density gradients (due to nonuniform
temperature and salinity concentration distributions)
• Inclusion of pressure gradients terms in the momentum equation (density driven flows)
• Turbulence model to account for the vertical turbulent viscosity and diffusivity based on the eddy viscosity
concept.
• Shear stresses exerted by the turbulent flow on the bottom based on a quadratic Chezy or Manning's
formula
• Wind stresses on the water surface modeled by a quadratic friction law
• Simulation of the thermal discharge, effluent discharge, and the intake of cooling water at any location and
any depth in the computational field (advective-diffusion module)
• Facility to calculate drogue tracks
• Simulation of drying and flooding of intertidal flats (moving boundaries) for both two-dimensional and
three-dimensional cases
Sediment Transport
The SED module can be applied in all geographic regions where DelftSD is used. The module is generally used to
calculate the short-term transport of sediment and sand. In particular the SED is used when the effect of changes in
bottom topography on flow conditions can be neglected. For long-term development of the bottom topography or
coastal morphology, a separate morphological module with coupling capabilities with the FLOW and WAVE
modules can be applied. The following discusses the standard features of the SED module.
171
-------
Sedimentation
• Effects of "hindered settling" (i.e., decrease in sedimentation velocity at very high suspended solids
concentration) can be included
• Each of the paniculate fractions are treated independently (i.e., sand and silt)
• Bottom shear stresses take into account currents, waves, and effects of shipping and fisheries
Re-suspension
• The re-suspension flux is limited based on the available amount of sediment in a sediment layer for the
variable layer option. The re-suspension is unlimited if the fixed layer option is used
• Re-suspension flux is zero if the water depth becomes too small
• The effect of short waves (on the sediment transport) can only be taken into account as wind effects
Burial
• Sediment can be transferred downward from one sediment layer to an underlying layer in a process known
as "burial"
Upward sediment transport ("Digging")
• Sediment can be transferred upward to one sediment layer from an underlying layer in a process known as
digging
Contaminant Transport
The WAQ module of DLFT3D is a general water quality program capable of describing a wide range of water
quality processes. The water quality processes may be described by arbitrary linear or nonlinear functions of the
selected state variables and model parameters. Typical types of applications are
• Exchange of substances with the atmosphere (oxygen, volatile organic substances, temperature)
• Adsorption and desorption of contaminant substances (heavy metals, organic micropollutants) and ortho-
phosporous
• Deposition of particles and adsorbed substances to the bed
• Re-suspension of particles and adsorbed substances from the bed
Model Framework
• Three-dimensional curvilinear-orthogonal finite difference
Scale
Spatial Scale
• Three-dimensional
Temporal Scale
• Dynamic
Assumptions
• Based on accepted formulations of the three-dimensional hydrostatic hydrodynamic equations and
conservative transport equation
Model Strengths
• Strong capabilities for hydrodynamics
172
-------
Model Limitations
• Considerable technical expertise in hydrodynamics is required to use the model effectively
Application History
DelftSD has been used for storm surge and flood forecasting in India and typhoon surge modeling in Vietnam.
DelftSD has also been used to model other estuarine and riverine systems, such as Victoria Harbor in Hong Kong.
The DELFT3D hydrographic model was used for large-scale hydrodynamics and connected mass transport in Lake
Malawi/Nyasa/Niassa.
DELFT3D was also used to support the project, Sustainable Development of the Laguna de Bay Environment.
Model Evaluation
Although there are no specific evaluation studies available, as a policy, new versions are released only after an
internal, 6 month test and application period to ensure stable and validated products.
Model Inputs
• Time-varying water-surface elevations, salinity, and temperature conditions at the ocean entrance and at
freshwater inflows at the head of all tributaries
• Time-varying wind and surface heat exchange data at one or more locations
• All input data, including initial conditions, bathymetry, boundary, and computational control data are input
from fixed files
Users' Guide
• Must be purchased with model
Technical Hardware/Software Requirements
Computer hardware:
• PC and Unix Workstations
Operating system:
• UNIX workstations (HP, SGI and Sun) or Windows/Intel platform (Windows 95, 98 and NT4)
Programming language:
• FORTRAN
Runtime estimates:
• Computation intensive
• Run time is highly dependent on computer hardware, model domain, spatial resolution, the period of
prototype conditions simulated, and other options, such as whether the model is simulation-only
hydrodynamic or hydrodynamics and the fate and transport of dissolved and suspended material. Under this
wide range of variability, simulations could require minutes to weeks
Linkages Supported
None
173
-------
Related Systems
None
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Visualization: Delft-GPP
• Grid generator: Delft-RGFGRID
• Bathymetry generator: Delft-QUICKIN
References
Not available
174
-------
DIAS/IDLAMS: Dynamic Information Architecture System/Integrated
Dynamic Landscape Modeling and Analysis System
Contact Information
Pamela J. Sydelko
Decision and Information
Sciences Division
Argonne National Laboratory
9700 S. Cass Avenue, Bldg. 900
Argonne, IL 60439
(630) 252-6727
svdelko(@,dis.anl.gov
http://www.dis.anl.gov/idlams/index.html
Download Information
Availability: Nonproprietary
Need to contact the author. A Java version of DIAS is available for customization.
Cost: DIAS is licensed but free for government agencies.
Model Overview/Abstract
Argonne's Integrated Dynamic Landscape Analysis and Modeling System (IDLAMS) integrates data,
environmental models, land-use planning, and decision support technologies through a GIS-based framework (GIS-
IDLAMS). The IDLAMS prototype is populated with models appropriate for use by military land managers and
decision makers at Fort Riley, Kansas, and can be used to
• Simulate "what-if' scenarios for predicting future ecological conditions under a given land management
plan
• Incorporate trade-off analyses when comparing different land management alternatives
• Identify and resolve land-use issues and determine cost-effective solutions to long-term land stewardship
problems
Recently, IDLAMS has evolved to take advantage of a flexible, dynamic, and cost-effective object-oriented
approach. This new framework, called OO-IDLAMS, is built on Argonne's DIAS, a generic, object-oriented
architecture that supports distributed, dynamic representation of interlinked processes.
OO-IDLAMS provides environmental managers and decision makers with a strategic, adaptive approach to
integrated natural resources planning and ecosystem management. The OO-IDLAMS framework
• Brings together disparate data and software for integrated natural resource planning and ecosystem
management in a flexible and adaptive way
• Provides the ability to reflect the dynamics of living ecosystems, land uses, and land management activities
• Reduces the cost of simulation modeling by using and reusing existing data, models, and system
components with minimal reworking
• Allows for a comprehensive regional, landscape, or ecosystem approach
175
-------
Model Features
GIS-IDLAMS
• Vegetation dynamics simulation
• Wildlife habitat suitability modeling
• Erosion calculated using Revised Universal Soil Loss Equation (RUSLE)
• Trade-off analyses for land management alternatives
OO-IDLAMS
• Object-oriented architecture, military land management, integrated natural resource planning, adaptive
ecosystem management, run-time model interoperability, code reuse, environmental decision support
• Vegetation dynamics simulation
• Henslow's Sparrow habitat model
• Human activities simulation for military training, burning, and planting
Model Areas Supported
Watershed Medium
Receiving Water None
Ecological High, Medium
Air None
Groundwater None
Model Capabilities
Conceptual Basis
Uses GIS or object-oriented BIAS to integrate natural resource planning, ecosystem management, and simulation
models and to support environmental decision making processes.
Scientific Detail
Four major models were developed and integrated for the GIS-IDLAMS prototype: (1) a vegetation dynamics
model, (2) a set of wildlife habitat suitability models, (3) an erosion model, and (4) a scenario evaluation module.
Vegetation Dynamics Model
The Vegetation Dynamics Model is the core model for IDLAMS because the output from this model is the input for
all other connected IDLAMS models. The Vegetation Dynamics Model is a spatially explicit model that
incorporates vegetation changes due to (1) natural succession, (2) land use impacts, and (3) land management
actions.
Wildlife Models
The Wildlife Models are five submodels that represent individual wildlife species and are based on U.S. Fish and
Wildlife Service Habitat Suitability Indices (HSIs). Each submodel requires that the user input either a
vegetation/landcover map representing the current condition or a simulated landcover map generated by the
Vegetation Dynamics Model. In some submodels, additional input maps may be required.
Erosion Model
IDLAMS currently integrates RUSLE to generate an erosion status map for each current condition or simulated
vegetation/landcover map input by the user. RUSLE also requires other spatial data representing various factors
affecting erosion.
176
-------
Scenario Evaluation Module
IDLAMS uses a value-based decision-analysis process to link the ecological models with the management needs and
user requirements of the resource manager. This module is then used to perform trade-off analyses for land
management alternatives, on the basis of the results from the spatially explicit modeling, and to rank the alternatives
according to how well they meet the specified objectives.
In OO-IDLAMS, because the objective of the research was to demonstrate the advantages of this new object-
oriented architecture approach rather than to totally rebuild the old IDLAMS, the OO-IDLAMS prototype integrates
only a subset of the original IDLAMS. Models in the new OO-IDLAMS include the Vegetation Dynamics Model
and the Henslow's Sparrow Habitat Model. In addition, to demonstrate improved modularity and flexibility of OO-
IDLAMS and fully utilize the object-oriented capabilities of DIAS, the Military Training and Land Management
components, previously coded within the original Vegetation Dynamics Model, were broken out into three Course
of Action (COA) objects. A DIAS COA object is essentially a flowchart of individual steps constituting a specific
plan or action and is used in DIAS to model procedural or sequential processes. OO-IDLAMS employs an object-
oriented GIS module and provides real-time spatially oriented displays of an object's positions and/or parameters.
Model Framework
GIS-IDLAMS uses GIS software as the model integration framework. OO-IDLAMS is built on an object-oriented
architecture called the DIAS. DIAS supports distributed, dynamic representation of interlinked environmental
processes and behaviors at variable scales (spatial and temporal) of resolution and aggregation. For integrated
environmental modeling, the main components of a DIAS simulation are (1) software objects (entity objects) that
represent real-world entities such as atmosphere, fish, or groundwater; and (2) simulation models or other
applications that express the dynamic behaviors of the real-world entities (e.g., surface exchange, reproductive
cycles, and fate and transport). In DIAS simulations, external models or applications participate in a simulation
through a formalized registration process that "wraps" each model or application for use in DIAS. This "wrapping"
procedure requires a formal registration procedure that enables the DIAS entity objects to implement external
models to address behaviors. An important feature of DIAS is that the "wrapped" models and applications run in
their native languages rather than requiring translation to a common or standard system language.
Scale
Spatial Scale
• One-dimensional, grid and subwatershed overland
Temporal Scale
• Depends on models integrated in the system
Assumptions
The model assumes that vegetation dynamics caused by natural and human forces immediately affect the wildlife
habitats. The model also assumes that the main erosion type in the study area is sheet and rill erosion that can be
described by RUSLE.
Model Strengths
Argonne's flexible object-oriented approach to model integration overcomes several limitations of a GIS-based
integration framework. GIS-based systems are static in nature and do not lend themselves well to dynamic
intermodel processing. Furthermore, integration of new environmental models or data formats into a GIS-based
framework can require time-consuming and expensive reworking of the system.
OO-IDLAMS can execute external applications in their native languages (e.g., FORTRAN and C) and allows them
to dynamically interact with each other indirectly via real-world ecosystem objects that package attribute
information together with behavior (how the object acts and reacts). Because external applications do not interact
directly with one another, OO-IDLAMS provides a robust environment that easily accommodates adding and
177
-------
removing applications. The OO-IDLAMS prototype includes an object-oriented GIS (GeoViewer), but in addition,
external GIS software applications can be integrated to provide further spatial analysis functions.
OO-IDLAMS provides a graphical user interface for selecting appropriate applications and for easy input of data
and parameters. The OO-IDLAMS prototype integration framework runs under both UNIX (Solaris) and Windows
NT platforms.
Model Limitations
Although DIAS provides an excellent framework for the integration of multiple models (even models at different
spatial and temporal scales), it does not solve the more basic ecological and environmental research issues related to
model integration. These issues include, but are not limited to (1) the ecological implications of multiple-scale
modeling and simulation and (2) the impacts of data aggregation and disaggregations. However, DIAS can be used
as an excellent workbench from which to explore and investigate these issues. In addition, further development of
the DIAS architecture should include the application of uncertainty analysis functionality to models within the DIAS
suite and a multidisciplinary/multiagency approach to object design and development.
Technically, DIAS has the following additional limitations:
• Programming languages (C++, SmallTalk, Java, etc.) must be used for a DIAS model integration
• Substantial skills are needed for full technical utilization of the system (customization).
Application History
Object-Oriented Integrated Dynamic Landscape Analysis and Modeling System (OO-IDLAMS) is one example of
the application of DIAS, other examples include
Healthcare Management Simulator - This DIAS application was developed to simulate all major aspects of
healthcare delivery - clinical/physiological, procedural, logistical, and financial - at a level of detail appropriate to
the need.
Integrated Ocean Software Architecture - This application of DIAS is an object-based virtual maritime environment
within which existing models are employed to simulate the transition of wind-generated waves in the deep water, to
waves in the near shore environment, then to surf height and currents. The application allows for the use of several
different model combinations according to the complexity of the shoreline and near-shore sea bottoms.
Model Evaluation
Reviewed by Cooperative Research Centre for Catchment Hydrology, Australia
http://science.csumb.edu/~fwatson/publications/B2Report%20001201.pdf
Model Inputs
• Landcover map
• Successional timestep map
• Area impacted by training
• Area impacted by burning
• Area impacted by planting
• Number of maneuver impact miles
• RUSLE parameter values (GIS- IDLAMS)
Users' Guide
Not available, need to contact the authors
178
-------
Technical Hardware/Software Requirements
Computer hardware:
• PC or SUN workstation
Operating system:
• UNIX (Solaris) or Windows NT platforms
Programming language:
• SmallTalk, C, Java, and FORTRAN
Runtime estimates:
• Not available, depends on models integrated in the system
Linkages Supported
None
Related Systems
DIAS
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Windows interface with menus and tool buttons
• Map displaying capacity
References
Sydelko, PJ, KA Majerus, JE Dolph, and TN Taxon. 2000. An Advanced Object-Based Software Framework for
Complex Ecosystem Modeling and Simulation. In Proceedings of the 4th International Conference on Integrating
GIS and Environmental Modeling (GIS/EM4): Problems, Prospects and Research Needs, Banff, Alberta, Canada,
September 2-8, 2000.
Campbell, P, and JR Hummel. The Dynamic Information Architecture System: An Advanced Simulation Framework
for Military and Civilian Application. http://www.dis.anl.gov/DIAS/papers/SCS/SCS.html
Christiansen, JH. 2000. A flexible object-based software framework for modeling complex systems with interacting
natural and societal processes. In Proceedings of the 4th International Conference on Integrating GIS and
Environmental Modeling (GIS/EM4): Problems, Prospects and Research Needs, Banff, Alberta, Canada, September
2-8, 2000.
179
-------
Appendix A: Model Fact Sheets
DRAINMOD: A Hydrological Model for Poorly Drained Soils
Contact Information
R. Wayne Skaggs
Biological and Agricultural Engineering Department
NC State University
Box 7625
Raleigh, NC 27695
(919)515-6739
skaggs(@,eos.ncsu.edu
http://www.bae.ncsu.edu/people/facultv/skaggs
Download Information
Availability: Proprietary
http://www.bae.ncsu.edu/soil water/drainmod
Cost (if applicable): $250
Model Overview/Abstract
DRAINMOD is a hydrological model developed at North Carolina State University (NCSU) by Dr. Wayne Skaggs
at the Department of Biological and Agricultural Engineering. The model was initially developed in 1980 for
analyzing field-scale watershed management scenarios for poorly drained soils, but in the last 2 decades it has been
updated and used on both field- and watershed-scale watershed management sites. The model has been updated
several times to extend its capabilities. The latest version, DRAINMOD Version 5.1, has the original DRAINMOD
hydrology model with the addition of DRAINMOD-N (nitrogen submodel) and DRAINMOD-S (salinity submodel)
into a Windows-based program. The DRAINMOD hydrology model simulates the hydrology of poorly drained,
high water table soils on an hourly or daily basis for long periods of weather data. Hourly rainfall is used to compute
the infiltration using a modified Green-Ampt approach and remaining excess rainfall is considered runoff. The water
balance is made using a one-dimensional vertical column, which is further used as inputs to the one-dimensional
nitrogen fate and transport module. In addition to organizing the hydrology, nitrogen, and salinity components of
DRAINMOD, the graphical user interface allows easy preparation of input datasets, running simulations, and
displaying model outputs.
Model Features
• Field and watershed-scale analysis
• Simulates the hydrology of poorly drained, high water table soils
• Simulates nitrogen and salinity transport
• Models the performance of water table management systems
• The graphical user interface
Model Areas Supported
Watershed Medium to High
Receiving Water None
Ecological None
Air Low
Groundwater Low
180
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
DRAINMOD is based on one-dimensional water balance in the soil profile and uses long duration weather data to
simulate the performance of drainage systems. The model was developed specifically for shallow water table, poorly
drained soils. The model avoids the complex numerical methods by using approximate methods to quantify the
hydrologic components, including subsurface drainage, subirrigation, infiltration, evapotranspiration (ET), and
surface runoff. For example, subsurface drainage is computed using the Houghoudt equation, which assumes an
elliptical water table shape. The change in water table depth is based on the assumption that the soil water profile
above the water table is drained to equilibrium with the water table.
Hourly rainfall is used to compute infiltration using a modified Green-Ampt method. Excess rainfall fills surface
storage and any remaining rainfall is considered runoff. Evapotranspiration is computed from potential
evapotranspiration as limited by soil water availability. Actual evapotranspiration is the amount that can be supplied
from the water table plus the amount available from the unsaturated zone.
DRAINMOD version 5.1 is also linked to a one-dimensional nitrogen cycling model (Breve et al., 1997a and b).
The model uses the water balances and fluxes from hydrology as inputs to a one-dimensional advective-dispersive-
reactive equation for nitrogen fate and transport. The model uses nitrate as the main pool for the simplification of the
nitrogen cycle. The nitrogen balance considers fertilizer dissolution, mineralization of organic nitrogen,
denitrification, and plant uptake using first order rate equations.
Scientific Detail
The major process of the model is a water balance for the soil profile. The water balance for a time increment of At
is expressed as,
AV = D + ET + DS - F
Where AV is the change in the volume (cm), D is lateral drainage (cm) from the section, ET is evapotranspiration
(cm), DS is deep seepage (cm), and F is infiltration (cm) entering the section during time interval At. All of the
right-hand side terms are computed in terms of the water table elevation, soil water content, soil properties, site and
drainage system parameters, and atmospheric conditions.
The amount of runoff and storage on the surface is computed from a water balance at the soil surface for each time
increment as,
P = F + AS + RO
Where P is the precipitation (cm), F is infiltration (cm), AS is the change in volume of water stored on the surface
(cm), and RO is runoff (cm) during time interval At.
Model Framework
• Version 5.1 includes a user interface written in Visual Basic 6.0 and combines the different versions of
DRAINMOD (Hydrology, Nitrogen, and Salinity). Following are the model components:
o Precipitation (hourly data)
o Infiltration (the Green-Ampt equation)
o Surface drainage (the average depth of depression storage)
o Subsurface drainage (the rate of subsurface water movement into drain ditches)
o Subirrigation (lateral flow)
o Evapotranspiration
o Soil water distribution
o Rooting depth
181
-------
Appendix A: Model Fact Sheets
Scale
Spatial Scale
• Watershed and field level
Temporal Scale
• Hourly and daily
Assumptions
• Assumes an elliptical water table shape
• Assumes subsurface drains are parallel
• Drains the soil water profile above the water table to equilibrium with the water table
Model Strengths
• Simulates surface and subsurface water flows in response to water management systems in soils with high
water tables
• Compares and evaluates the effectiveness of the design over a range of weather scenarios
Model Limitations
• Supports only parallel subsurface drains
• Should not be applied to situations that are widely different than conditions for which it was developed,
without further testing
Application History
DRAINMOD reference report (Skaggs, 1980) presented these four sets of model application examples:
1. Combination surface-subsurface drainage systems
2. Subirrigation and controlled drainage
3. Irrigation of waste water on drained lands
4. Effect of root depth on the number and frequency of dry days
In addition to hydrology of DRAINMOD, the nitrogen and salinity portions of the model have been tested by Breve
et al. (1997b), Kandil et al. (1995), and Merz and Skaggs (1998). Other field testing is currently ongoing for these
versions of the model.
Model Evaluation
The reliability of DRAINMOD has been tested for a wide range of soil, crop, and weather conditions. Results of
tests in North Carolina (Skaggs, 1982), Ohio (Skaggs et al., 1981), Louisiana (Fouss et al., 1987), and Virginia
(McMahan et al., 1988) indicate that the model can be used to reliably predict water table elevations and drain flow
rates.
Model Inputs
• Soil property inputs
• Hydraulic conductivity
• Soil water characteristics
• Drainage volume - water table depth relationship
• Upward flux
• Green-Ampt equation parameters
• Crop input data
182
-------
Appendix A: Model Fact Sheets
• Drainage system parameter
• Surface drainage
• Effective drain radius
Users' Guide
Available online: http://www.bae.ncsu.edu/soil water/drainmod/index.htm
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• Microsoft Windows 95/98/2000/N17XP.
Programming language:
• Visual Basic and FORTRAN
Runtime estimates:
• Minutes to less than 1 hour
Linkages Supported
The original DRAINMOD hydrology model has been modified to include submodels on the fate and transport of
nitrogen in the soil and salinity. Following are the models developed at the Biological and Agricultural Engineering
Department at NCSU.
• DRAINLOB: DRAINMOD-based field-scale forest hydrologic model
• DRAINMOD-S: DRAINMOD-based field-scale model for predicting salinity on arid/semi-arid lands
• DRAINMOD-N: DRAINMOD-based field-scale model for predicting Nitrogen from agricultural lands
• WATGIS: A GIS-based lumped parameter watershed-scale hydrology and water quality model.
DRAINMOD and DRAINMOD-N models coupled with a delivery ratio routine to route drainage water and
nutrients to the watershed outlet
• DRAINMOD-GIS: A GIS-based lumped parameter watershed-scale hydrology and water quality model.
DRAINMOD/DRAINMOD-N models coupled with a simplified water and nutrient fate and transport
submodels
• DRAINMOD-W: A watershed-scale model based on DRAINMOD and DRAINMOD-N field-scale
submodels with a finite difference canal routing model and a finite element solute transport submodel
Related Systems
DRAINMOD supports several numerical models as listed in Linkage Supported section.
Sensitivity/Uncertainty/Calibration
The results for the sensitivity analyses conducted for different soils and water management systems of North
Carolina are presented in DRAINMOD reference report (Skaggs, 1980). This report indicated that hydraulic
conductivity (K) is a very sensitive parameter in this model. Similarly, SEW-30 (Sum of Excess Water for water
table depths greater than 30 cm) was more sensitive to errors in K and PET than to any of the other input parameters.
The number of dry days (days in which actual ET is less than PET) were less dependent on K than either working
days (a day is counted as a working day if the drained or water free pore space is greater than a threshold value) or
183
-------
Appendix A: Model Fact Sheets
SEW-30 for all cases evaluated. Dry days were also reported to be quite sensitive to errors in the water content at the
lower limit (wilting point).
Model Interface Capabilities
• A graphical user interface for assisting the user in developing input datasets
• Options to enable the user to analyze the model results graphically
• No GIS interface available for the latest version of the model, but watershed and drainage basin scale
versions are currently under development, which utilize ARC Info and ARCView to assist in model setup
and analysis of results
References
Breve, M. A., R. W. Skaggs, J. E. Parsons and J. W. Gilliam. 1997a. DRAINMOD-N: A nitrogen model for
artificially drained lands. Trans. ASAE. 40(4): 1067-1075.
Breve, M. A., R. W. Skaggs, J. W. Gilliam, J. E. Parsons, A. T. Mohammad, G. M. Chescheir and R. O. Evans.
1997b. Field testing of DRAINMOD-N. Trans. ASAE. 40(4):1077-1085.
Fouss, J. L., R. L. Bengston and C. E. Carter. 1987. Simulating subsurface drainage in the lower Mississippi Valley
withDRAINMOD. Trans. ASAE. 30(6): 1679-1688.
Kandil, H. M., R. W. Skaggs, S. A. Dayem and Y. Aiad. 1995. DRAINMOD-S: A water management model for
irrigated arid lands, crop yield and applications. Irrigation and Drainage Systems. 9(3):239-258.
McMahon, P. C., S. Mostaghimi and F. S. Wright. 1988. Simulation of corn yield by a water management model for
a Coastal Plain soil in Virginia. Trans. ASAE. 31(3):734-742.
Merz, R. D. and R. W. Skaggs. 1998. Application of DRAINMOD-S to Determine Drainage Design Criteria for
Irrigated Semi-Arid Lands. In Proceedings of Seventh International Drainage Symposium Drainage in the 21st
Century: Food Production and the Environment, Ed. Larry C. Brown, American Society of Agricultural Engineers,
Orlando, Florida, March 8-10, 1998, pp. 347-354.
Skaggs, R.W. 1980. DRAINMOD Reference Report - Methods for design and evaluation of drainage-water
management systems for soils with high water tables. USDA SCS, South National Technical Center Texas.
Skaggs, R. W., N. R. Fausey and B. H. Nolte. 1981. Water management model evaluation for North Central Ohio.
Trans. ASAE. 24(4): 922-928.
Skaggs, R. W. 1982. Field evaluation of a water management model, DRAINMOD. Trans. ASAE. 25(3): 666-674.
184
-------
Appendix A: Model Fact Sheets
DWSM - Dynamic Watershed Simulation Model
Contact Information
Deva K. Borah
Illinois State Water Survey
2204 Griffith Drive
Champaign, Illinois 61820
borah@uiuc. edu
Download Information
Availability: Non-Proprietary
Cost: Free
Model Overview/Abstract
Developed by the Illinois State Water Survey, the DWSM (Borah et al., 2002; Borah and Bera, 2003) uses
physically based governing equations to simulate surface and subsurface storm water runoff, propagation of flood
waves, soil erosion, and entrainment and transport of sediment and agricultural chemicals in agricultural watersheds
during severe rainfall events. The model has three major components: (1) DWSM-Hydrology (Hydro) simulating
watershed hydrology, (2) DWSM-Sediment (Sed) simulating soil erosion and sediment transport, and (3) DWSM-
Agricultural chemical (Agchem) simulating agricultural chemical (nutrients and pesticides) transport. Each
component has routing schemes developed using approximate analytical solutions of the physically based equations
preserving the dynamic behaviors of water, sediment, and the accompanying chemical movements within a
watershed.
Model Features
• Agricultural watershed, nonpoint sources
• Distributed, single event model
• Spatially varying rainfall inputs
• Individual hyetograph for each overland
• Rainfall excess, surface and subsurface overland flow
• Surface erosion and sediment transport
• Agrochemical mixing and transport
• Channel erosion and deposition and routing of flow, sediment, and agrochemical and flow routing through
reservoirs
• Detention basins, alternative ground covers, and tile drains
Model Areas Supported
Watershed Medium
Receiving Water Low
Ecological None
Air None
Groundwater None
185
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
The watershed is divided into subwatersheds, specifically, into one-dimensional overland elements, channel
segments, and reservoir units. An overland element is represented as a rectangular area with the same area as in the
field, width equal to the adjacent (receiving) channel length, length equal to area divided by the width, and
representative slope, soil, cover, and roughness based on physical observations of these characteristics in the
element. A channel segment is represented with a straight channel having the same length as in the field and having
a representative cross-sectional shape, slope, and roughness based on physical observations and measurements. A
reservoir unit is represented with a stage-storage-discharge relation (table) developed based on topographic data and
discharge calculations using outlet measurements and established relations.
Scientific Detail
The DWSM-Hydro: Hydrologic Simulations
The overland elements are the primary sources of runoff in which rainfall turns into surface runoff after losing first
to interception at canopies and ground covers, then to infiltration through the ground surface and depression storage
above it. The rainfall available for surface runoff is the rainfall excess. A portion of the infiltrated water flows
laterally towards downstream as subsurface flow sometimes in accelerated mode in the presence of tile drains. Two
overland elements contribute surface and subsurface flows into one channel segment laterally from each side of the
channel. The excess rainfall is routed over the overland elements beginning at their upstream edges (ridges), at
which flows are zeros, to their downstream edges, coinciding with the receiving channel banks. Similarly,
subsurface water from infiltration is routed through the soil matrix underneath the overland elements beginning at
their upstream edges (ridges), at which flows are assumed zeros, to their downstream edges, coinciding with the
receiving channel banks. Currently, the tile drain flows from overland elements having tile drains are lumped with
the subsurface flow through the soil matrix using an effective lateral saturated hydraulic conductivity concept. The
channel segments carry the receiving waters from overland elements and upstream channel segments towards the
downstream side of the watershed and ultimately to the watershed outlet. During its journey, the runoff water may
be intercepted by reservoirs, which release it again to downstream channels at reduced rates after temporary storage.
The DWSM-Sed: Soil Erosion and Sediment Transport Simulations
Similar to the hydrologic component, soil erosion and sediment transport are simulated along with water through the
overland elements and stream segments. The eroded soil or sediment is divided into number of particle size groups.
Agricultural watersheds having extensive aggregates, the sediment is divided into five size groups: sand, silt, clay,
small aggregate, and large aggregate. Each size group is dealt individually during the simulation of each of the
processes, and total response, in the form of sediment concentration and discharge is obtained by integrating the
responses from all the size groups.
The model computes soil erosion due to raindrop impact. The eroded (detached) soil is added to an existing detached
(loose) soil depth from where entrainment to runoff takes place with sufficient velocity and shear (capacity). Erosion
due to flow shear stress and deposition depends on sediment transport capacity of the flow and the sediment load
(amount of sediment already carried by the flow). Sediment transport capacity is computed using established
formulas. If the capacity is higher than the sediment load, erosion takes place and the flow picks up more materials
from the bed. If the loose soil volume at the bed is sufficient, sediment entrainment takes place from the detached
soil depth. Otherwise, the flow erodes additional soil from the parent bed material. If the sediment transport capacity
is lower than the sediment load, the flow is in a deposition mode and the potential rate of deposition is equal to the
difference of the two. The actual rate of deposition is computed by taking into account particle fall velocities.
Deposited sediment is added to the loose soil volume. If the sediment transport capacity and the sediment load are
equal, an equilibrium condition is assumed where there is neither erosion nor deposition. All the above processes are
interrelated and must satisfy locally the conservation of sediment mass principle expressed by the sediment
continuity equation. The continuity equation is solved to keep track of erosion, deposition, and sediment discharges
along the flow segments.
186
-------
Appendix A: Model Fact Sheets
The DWSM-Agchem: Agricultural Chemical Transport Simulations
The agricultural chemical transport component of DWSM involves simulations of mixing of nutrients and pesticides
and transport of these chemicals with surface runoff in dissolved form, and with sediment in adsorbed form in each
of the flow segments. Similar to the hydrologic and soil erosion-sediment transport components, these processes are
simulated along with water and sediment through the overland elements and stream or channel segments.
The model assumes equilibrium between dissolved and adsorbed phases of the chemicals, governed by a linear
adsorption isotherm. The soil profile is divided into small homogeneous soil increments. The model routes
infiltrating rainwater and solutes through the soil increments and computes resulting water contents and chemical
concentrations. When runoff begins, exchange of chemicals from a mixing soil layer of the soil profile, containing
the chemicals in dissolved form, with surface runoff are simulated using the concept of non-uniform mixing of
runoff with the mixing soil layer. Exchange of chemicals in adsorbed forms with the eroded and deposited sediments
is computed based on preference factors of the individual size groups. The entrained chemicals are routed along
slope lengths in dissolved form with surface runoff and in adsorbed form with the transported sediment using
solutions of the continuity (mass conservation) equations.
Model Framework
The watershed is divided into subwatersheds, specifically, into one-dimensional overland elements, channel
segments, and reservoir units. Overland elements are spatially distributed. Modules are linked to calculate
hydrology, erosion and sediment transport, and agricultural chemical transport.
Scale
Spatial Scale
• Watersheds of sizes ranging from few acres to several hundred square kilometers are divided into
hydrologic units defined by topographical or natural boundaries and further divided into overland, channel,
and reservoir segments.
Temporal Scale
• Several days of storm events divided into constant time intervals ranging from few minutes to few hours
Assumptions
• Uses kinematic wave equation assuming that all the acceleration and pressure gradient terms in the
momentum equation can be ignored
• Assumes equilibrium between dissolved and adsorbed phases of the chemicals, governed by a linear
adsorption isotherm.
• Assumes all sediment are trapped in reservoir and no downstream discharge
Model Strengths
• Is a distributed model (overland elements)
• Detailed and physically-based model yet relatively efficient, provides a balance between the simple
(lumped) and complicated (computationally intensive) models
• Suitable for flat Midwestern watersheds with extensive tile-drained lands
• Is capable of evaluating the effects of some BMPs, such as detention basins, alternative ground covers, and
tile drains
Model Limitations
• Only used to simulate single event
• Only for agricultural watersheds
• One-dimensional channel simulation (no ecological processes)
187
-------
Appendix A: Model Fact Sheets
Application History
The DWSM-Hydro was applied and tested on the Upper Sangamon River basin draining a
2,400-km2 agricultural watershed to Lake Decatur in Illinois (Borah et al, 1999, 2000). Lake
Decatur is a public water supply reservoir for the City of Decatur having a history of high nitrate nitrogen
(nitrate-N) concentration, periodically exceeding 10 milligrams per liter (mg/L), and violating state and national
drinking water standards. The lake also has a high sedimentation rate that is gradually reducing its water supply
capacity. The goal is to use the model to evaluate the benefits of applying alternative land use and BMPs in reducing
soil erosion, and sediment and agricultural chemical discharges into the lake and help solve the water quality and
sedimentation problems.
The DWSM-Hydro & Sed were applied and tested (calibrated and validated) on the Big Ditch watershed using data
monitored during 1998 spring and early summer storm events (Borah et al., 1999; Borah et al., 2001). Big Ditch is a
tributary to the Upper Sangamon River draining 100-km2 agricultural lands. In this application, scaling effects on
model parameters and water and sediment discharges resulting from watershed divisions with alternative subdivision
sizes were investigated. The DWSM-Agchem was also applied and tested on the Big Ditch watershed (Xia et al.,
2001) and preliminary results indicated that the model needed improvements in simulating agricultural chemicals,
especially nitrate-N.
The DWSM-Hydro was applied to the Court Creek watershed in Illinois (Borah and Bera, 2000a,b). This 251-km2
watershed is part of the Illinois multi-agency Pilot Watershed and Conservation Reserve Enhancement Programs.
The DWSM-Hydro was calibrated and validated using storm data monitored and reported earlier by the ISWS
(Roseboom et al., 1982, 1986). The model was then run for design storms and high, moderate, and low runoff
potential areas of the watershed were identified and ranked. The committee is currently using these results to plan
their initial restoration programs within the watershed. It has been realized that the design storms with Soil
Conservation Service's (SCS) rainfall distributions generated unrealistically high flows for BMP design purposes.
Therefore, rankings of overland elements and channel segments have been revised using a historical storm occurred
in the springtime. Rankings are based on unit-width peak flows and unit-width sediment yields for the overland
elements and on peak flows and sediment yields for the channel segments.
Model Evaluation
The model developers has conducted many studies to evaluate the model (see Model Application History)
Model Inputs
• Physical data representing the watershed, initial moisture, soil and agricultural chemicals and
meteorological data representing the rainfall events
Users' Guide
Contact the authors for documentation
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS or Windows system with FORTRAN complier
188
-------
Appendix A: Model Fact Sheets
Programming language:
• FORTRAN
Runtime estimates:
• > Minutes
Linkages Supported
None
Related Systems
None
Sensitivity/Uncertainty/Calibration
Sensitive to runoff curve number, saturated hydraulic conductivity, Manning's roughness coefficient, flow
detachment coefficient, chemical partition coefficient, chemical mixing parameter, sediment particle size
distribution, and rainfall intensities and their temporal distributions
Calibration parameters include runoff curve number or saturated hydrological conductivity, Manning's roughness
coefficient, flow detachment coefficient, chemical partition coefficient, and chemical mixing parameter
Model Interface Capabilities
Not available.
References
Borah, O.K. and M. Bera. 2000a. Hydrologic Modeling of the Court Creek Watershed. Contract Report 2000-04,
Illinois State Water Survey, Champaign, IL.
Borah, D.K. and M. Bera. 2000b. Watershed modeling with state and local partners in Illinois. In Proceedings of the
2000 Joint Conference on Water Resources Engineering and Water Resources Planning & Management, ed. R.H.
Hotchkiss and M. Glade, ASCE-EWRI, Reston, VA: CD-ROM.
Borah, D.K. and M. Bera. 2003. Watershed-scale hydrologic and nonpoint-source pollution models: Review of
mathematical bases. Transactions of the ASAE. 46: 1553-1566
Borah, D.K., M. Bera, S. Shaw, and L. Keefer. 1999. Dynamic Modeling and Monitoring of Water, Sediment,
Nutrients, and Pesticides in Agricultural Watersheds during Storm Events. Contract Report 655, Illinois State Water
Survey, Champaign, IL.
Borah, D.K., R. Xia, and M. Bera. 2000. Hydrologic and water quality model for tile drained watersheds in Illinois.
ASAE Paper No. 002093, St. Joseph, Mich.: ASAE.
Borah, D.K, R. Xia, and M. Bera. 2001. Hydrologic and Sediment Transport Modeling of Agricultural Watersheds.
In Proceedings of the World Water & Environmental Resources Congress, ed. D. Phelps and G. Sehlke, ASCE-
EWRI, Reston, VA: CD-ROM.
Borah, D. K., R. Xia, and M. Bera. 2002. DWSM - A dynamic watershed simulation model. Chapter 5 in
Mathematical Models of Small Watershed Hydrology and Applications, ed. V. P. Singh and D. K. Frevert, pp. 113-
166. Highlands Ranch, CO: Water Resources Publications.
189
-------
Appendix A: Model Fact Sheets
Roseboom, D., R.L. Evans, J. Erickson, and L.G. Brooks. 1982. An Inventory of Court Creek Watershed
Characteristics that may Relate to Water Quality in the Watershed. Contract Report 322, Illinois State Water
Survey, Champaign, IL.
Roseboom, D., R.L. Evans, J. Erickson, L.G. Brooks, and D. Shackleford. 1986. The Influences of Land Uses and
Stream Modifications on Water Quality in the Streams of the Court Creek Watershed. ILENR/RE-WR-86/16,
Illinois Department of Energy and Natural Resources, Springfield, IL.
Xia, R., D.K. Borah, and M. Bera. 2001. Modeling Agricultural Chemical Transport in Watersheds. In Proceedings
of the World Water & Environmental Resources Congress, ed. D. Phelps and G. Sehlke, ASCE-EWRI, Reston, VA:
CD-ROM.
190
-------
Appendix A: Model Fact Sheets
ECOMSED: Estuary and Coastal Ocean Model with Sediment Transport
Contact Information
HydroQual, Inc.
One Lethbridge Plaza
Mahwah, NJ, 07430
(201)529-5151
Ecom support(@,hvdroqual. com
http://www.hvdroqual.com/ehst ecomsed.html
Download Information
Availability: Nonproprietary Version Available
Register to download files: http://www.hvdroqual.com/ehst ecomsed.html
Cost: N/A
Model Overview/Abstract
ECOMSED is a three-dimensional hydrodynamic and sediment transport model. The hydrodynamic module solves
the conservation of mass and momentum equations with a 2.5-level turbulent closure scheme on a curvilinear
orthogonal grid in horizontal plane and o-coordinate in the vertical direction. Water circulation, salinity, and
temperature are obtained from the hydrodynamic module. The sediment transport module computes the sediment
settling and resuspension processes for both cohesive and noncohesive sediments under the impact of waves and
currents. The hydrodynamic component is same as the ECOM3D/POM model.
Model Features
• Three-dimensional hydrodynamics
• Cohesive and noncohesive sediment transport
• Sediment-bound and dissolved tracer transport
• Wind-waves-generated shear stress
Model Areas Supported
Watershed None
Receiving Water High
Ecological None
Air None
Groundwater None
Model Capabilities
Conceptual Basis
The waterbody is conceptualized as a series of grid points on a curvilinear orthogonal coordinate system.
Scientific Detail
The governing equations of the hydrodynamic component in ECOMSED are the continuity equation, Reynold's
equations, heat and salinity transport equations on curvilinear-orthogonal grid on the horizontal plane and
o-coordinate in the vertical direction. It uses a 2.5-level turbulence closure scheme that solves the transport of
191
-------
Appendix A: Model Fact Sheets
turbulent kinetic energy and turbulent macroscale. The governing equation of sediment transport is an advection-
dispersion equation that uses the hydrodynamic results. The hydrodynamic governing equations are solved using a
mode-splitting technique. The external mode that contains fast moving gravity wave is solved with small timesteps
to ensure stability, whereas the internal mode uses large timesteps to save the computation time. Finite difference of
the differential equation is applied on a staggered C grid in space, and the three-time-level leap-frog scheme is
applied for the timestepping. Three schemes, including central difference, upwind difference, and the
multidimensional positive definite advection scheme are provided in the model to solve the advection term in the
transport equations. The sediment transport component uses the same grid, structure, and computational framework
as the hydrodynamic component to simulate the settling, deposition, and resuspension of both cohesive and
noncohesive sediments. The Grant-Madson wave-current model is incorporated in ECOMSED to account for wind-
wave-generated shear stress.
Model Framework
• Three-dimensional model
• River, lake, reservoir, estuary, ocean
Scale
Spatial Scale
• One-, two-, and three-dimensional
Temporal Scale
• User-defined timestep
Assumptions
• Hydrostatic assumption
• Boussinesq approximation
• Reynold's stress assumption
Model Strengths
• ECOMSED is capable of modeling one-, two-, and three-dimensional hydrodynamics in various water
bodies with complex bathymetry.
• The boundary-fit curvilinear coordinate can represent the waterbody boundaries accurately with fewer grids
than Cartesian coordinate.
• The vertical o-coordinate represents the bathymetry without assuming rectangular bottom boundary.
Model Limitations
• The vertical o-coordinate may cause significant pressure gradient error at areas with sharp bottom elevation
change.
• Timestep and grid size need to be chosen carefully to balance the computation time and model resolution
and ensure model stability.
Application History
ECOMSED has been applied to Chesapeake Bay, New York Bight, Delaware Bay, Delaware River, Gulf Stream
Region, Massachusetts Bay, Georges Bank, the Oregon Continential Shelf, New York Harbor, and Onondaga Lake.
Model Evaluation
The hydrodynamic component of ECOMSED is based on Princeton Ocean Model, which has been tested and
applied by various users. The theory and model testing history can be found in journal and conference papers.
192
-------
Appendix A: Model Fact Sheets
Model Inputs
• Initial conditions
• Bathymetry and waterbody boundaries
• Physical coefficients
• Water surface elevations or flow rate at open boundary
• Time sequences of hydrometeorological conditions
Users' Guide
Available online (after registering): http://www.hvdroqual.com/ehst ecomsed.html
Technical Hardware/Software Requirements
Computer hardware:
• PC, workstation, and mainframe
Operating system:
• PC-DOS, Unix, Windows
Programming language:
• FORTRAN
Runtime estimates:
• Minutes to hours
Linkages Supported
Linked with HydroQual's RCA model.
Related Systems
POM, EFDC, CH3D, GLLVHT
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
Input file in text format.
References
HydroQual, Inc. 2002. A Primer for ECOMSED version 1.3. (Computer program manual). HydroQal, Inc., Mahwah,
NJ.
193
-------
Appendix A: Model Fact Sheets
EFDC: Environmental Fluid Dynamics Code
Contact Information
John M. Hamrick
Tetra Tech, Inc.
10306 Eaton Place, Suite 340
Fairfax, VA 22030
(703) 385-6000
j ohn. hamrick(@,tetratech-ffx.com
Virginia Institute of Marine Science
School of Marine Science
The College of William and Mary
Gloucester Point, VA 23052
(804) 642-7000
Download Information
Availability: Nonproprietary
Cost: N/A
Model Overview/Abstract
The EFDC model is single-source-code three-dimensional modeling system having hydrodynamic, water quality-
eutrophication, sediment transport, and toxic contaminant transport components transparently linked together. The
model can execute in a fully coupled mode, simultaneously simulating hydrodynamics and sediment and
contaminant transport, or in a transport-only mode, using saved hydrodynamic transport information. The EFDC
model uses a finite difference spatial representation and is capable of reduced dimension execution in one-
dimensional network and two-dimensional (horizontal or vertical plane) modes. Water column transport includes
three-dimensional advection and vertical and horizontal turbulent diffusion. Shear dispersion may be included for
two-dimensional horizontal applications. A water quality-eutrophication model, functionally equivalent to CE-
QUAL-IC, is also incorporated into EFDC (Hamrick and Wu, 1997; Park, et al., 1995). The model can be applied to
rivers, lakes, reservoirs, estuaries, wetlands, and coastal regions.
Model Features
• General purpose three-dimensional hydrodynamic and transport model
• Model simulates tidal, density, and wind-driven flow; salinity; temperature; and sediment transport
• Two built-in, fully coupled water quality/eutrophication submodels are included in the code, as well as a
toxicant transport and fate model
Model Areas Supported
Watershed Low
Receiving Water High
Ecological High
Air None
Groundwater Low
194
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
The EFDC model solves the vertically hydrostatic, free-surface, variable-density, turbulent-averaged equations of
motion and transport equations for turbulence intensity and length scale, salinity, and temperature in a stretched,
vertical coordinate system and in horizontal coordinate systems that may be Cartesian or curvilinear-orthogonal.
Equations describing the transport of suspended sediment, toxic contaminants, and water quality state variables are
also solved. Multiple size classes of cohesive and noncohesive sediments and associated deposition and
resuspension processes and bed geomechanics are simulated. Toxics are transported in both the water and sediment
phases in the water column and bed. The built-in 21-state-variable water quality model is based on the CE-QUAL-
ICM reaction kinetic. A 10-state-variable reduced water quality model is functionally equivalent to WASPS. Other
model features include drying and wetting, hydraulic structure representation, vegetation resistance, and Lagrangian
particle tracking. The model also accepts radiation stress fields from wave refraction-diffraction models, which
allows simulation of longshore currents and sediment transport.
Scientific Detail
The EFDC model framework includes methods for computing hydrodynamics, mixing zone dilution, eutrophication,
sediment transport, and toxic contaminant transport.
Mixing Zone
A Lagrangian buoyant jet near-field dilution and mixing-zone model is embedded within the far field solution
allowing representation of the local distribution of contaminated sediment near point sources.
Hydrodynamic
EFDC uses a finite difference scheme with three time levels and an internal-external mode splitting procedure to
achieve separation of the internal shear or baroclinic mode from the external free-surface gravity wave or barotropic
mode. An implicit external mode solution is used with simultaneous computation of a two-dimensional surface
elevation field by a multicolor successive overrelaxation procedure. The external solution is completed by
calculation of the depth-integrated barotropic velocities using the new surface elevation field. Various options can
be used for advective transport in EFDC. These include the "centered in time and space" scheme, and the "forward
in time and upwind in space" scheme.
Sediment Transport
The sediment transport component simulates a user specified number of size classes of cohesive and noncohesive
sediment. Sediment settling is represented by concentration and ambient-flow-turbulence-dependent formulations to
represent hindered settling of noncohesive sediment and approximately represent aggregation and disaggregation of
cohesive sediment. Water column-bed sediment and sorbed contaminant exchange is represented by deposition and
erosion fluxes. For noncohesive sediment, the net flux is represented as dependent on the bed stress, the near bottom
and bed surface sediment concentration, and the critical Shield's parameter. For cohesive sediment, deposition and
erosion fluxes are dependent on the bed stress, critical deposition and erosion stresses, and the shear strength of the
bed. The sediment bed is represented by a time varying number of layers. Sediment in each layer is characterized by
mass per unit area, void ratio, and shear strength. The void ratio, of the layers is specified or determined by a bed
consolidation model with shear strength being determined as a function of void ratio.
Contaminant Transport
Vertical transport of sediment and sorbed contaminants between bed layers is implicitly represented by sediment
particle displacement in response to layer thickness variations dynamically determined by the consolidation model.
Transport of dissolved contaminants between the water column and bed and between bed layers includes pore water
advection, dynamically determined by the bed consolidation model and pore water diffusion. An arbitrary number of
toxic contaminants can be simultaneously transported. The simple contaminant processes option includes constant
coefficient equilibrium partitioning, volatilization, and lumped first-order decay with unique coefficients for the
water column and sediment bed. The complex contaminant processes option allows for solids-concentration-
195
-------
Appendix A: Model Fact Sheets
dependent partitioning and specification of ambient-environment-dependent volatilization, hydrolysis, photolysis,
oxidation, and biodegradation reactions specific to the contaminants being simulated.
Ecosystem
The model is based on the CE-QUAL-ICM and incorporates a predictive sediment process or diagenesis model
(DiToro and Fitzpatrick, 1993). The eutrophication model is directly coupled to the hydrodynamic model and is
capable of two and three-dimensional spatial resolution. Water column state variables include up to three algae
classes represented in carbon equivalent units, ammonia, nitrite-nitrate, organic nitrogen, orthophosphate or
inorganic phosphorous, organic phosphorus, organic carbon, chemical oxygen demand, dissolved oxygen, available
and unavailable silica, and total active metal, which is used as a sorption site. Organic carbon, nitrogen, and
phosphorous are subdivided into three classes: dissolved, labile paniculate, and refractory paniculate. Model
variables in the sediment bed include paniculate organic carbon, nitrogen, and phosphorous, each in three reaction
rate classes; paniculate and available silica; sulfide or methane; ammonia; nitrate; inorganic phosphorus; bed-water
column fluxes of ammonia, nitrate, inorganic phosphorous and silica; sediment oxygen demand; and release of
chemical oxygen demand. The model's formulation allows direct determination of organic carbon levels in the water
column and sediment bed.
Model Framework
• Three-dimensional curvilinear-orthogonal finite difference
Scale
Spatial Scale
• Three-dimensional
Temporal Scale
• Dynamic
Assumptions
• Based on accepted formulations of the three-dimensional hydrostatic hydrodynamic equations and
conservative transport equation
Model Strengths
• Completely integrated three-dimensional hydrodynamics, water quality/eutrophication, and sediment-
contaminant transport
Model Limitations
• Requires considerable technical expertise in hydrodynamics to use the model effectively
• Requires expertise in eutrophication processes to use the water quality component
Application History
The EFDC model has been used for modeling studies in the estuaries of the Chesapeake Bay System, the Indian
River Lagoon and Lake Okeechobee in Florida, the Peconic Bay System in New York, Stephens Passage in Alaska,
and Nan Wan Bay in Taiwan. The model has also been used to simulate large-scale wetlands flow and transport in
the Everglades. The EFDC model has been applied extensively for circulation, discharge dilution, and water
quality/eutrophication studies (Hamrick, 1992b; Tetra Tech, 1994, 1995, 1998). The model has also been applied for
estuarine-cohesive sediment transport simulation (Yang, 1996), and coastal noncohesive sediment transport (Zarillo
and Surak, 1995), and heavy metals and organic contaminant transport (Schock and Hamrick, 1998).
196
-------
Appendix A: Model Fact Sheets
Model Evaluation
• Not available
Model Inputs
• Open boundary water surface elevation
• Wind and atmospheric thermodynamic conditions
• Open boundary salinity and temperature
• Volumetric inflows
• Inflowing concentrations of sediment and water quality state variables
• Input file templates are included with the source code and the user's manual to aid in input data preparation
Users' Guide
• Contact the developer at john.hamrick(g),tetratech-ffx.com.
Technical Hardware/Software Requirements
Computer hardware:
• PC, Macintosh, Unix workstation, Super computer
Operating system:
• Windows, Mac OS, Unix, Linux
Programming language:
• FORTRAN
Runtime estimates:
• Can be computation intensive but is highly optimized and has faster documented runtime than similar
software systems
• Runtime is highly dependent on computer hardware, model domain spatial resolution, the period of
prototype conditions simulated, and other options, such as whether the model is simulation-only
hydrodynamic or hydrodynamics and the fate and transport of dissolved and suspended material; under this
wide range of variability, simulations could require minutes to weeks
Linkages Supported
• Linkages to WASP, CE-QUAL-ICM, RCA, and a generic output for food chain and risk assessment model
Related Systems
GridEFDC, EFDCexplorer, EFDCview, EPA TMDL Toolbox
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Public domain GUI including EFDCexplorer and EFDCview
197
-------
Appendix A: Model Fact Sheets
References
Hamrick, J.M. 1992. A three-dimensional environmental fluid dynamics computer code: theoretical and
computational aspects. SRAMSOE #317. The College of William and Mary, Gloucester Point, VA.
Hamrick, J. M. 1992. Estuarine environmental impact assessment using a three-dimensional circulation and
transport model. In Proceedings of the 2nd International Conference Estuarine Coastal Modeling, ed. M. L.
Spaulding et al., American Society of Civil Engineers, New York, 1992, pp. 292-303.
Park, K., A. Y. Kuo, J. Shen, and J. M. Hamrick. 1995. A three-dimensional hydrodynamic-eutrophication model
(HEM3D): description of water quality and sediment processes submodels. Special Report 327. The College of
William and Mary, Virginia Institute of Marine Science, Gloucester Point, VA.
Hamrick, J. M. 1994. Linking hydrodynamic and biogeochemcial transport models for estuarine and coastal waters.
In Proceedings of the 3rd International Conference Estuarine and Coastal Modeling, ed. M. L. Spaulding et al.,
American Society of Civil Engineers, New York, 1994, pp. 591-608.
Fredricks, C., and J. M. Hamrick. 1996. The effect of channel geometry on gravitational circulation in partially
mixed estuaries. In Proceeding of Buoyancy Effects on Coastal and Estuarine Dynamics, ed. D. Aubrey and C.
Fredricks, American Geophysical Union, 1996, pp.283-300.
Hamrick, J. M. 1996. A User's Manual for the Environmental Fluid Dynamics Computer Code (EFDC). Special
Report 331. The College of William and Mary, Virginia Institute of Marine Science, Gloucester Point, VA.
Fredricks, C., and J. M. Hamrick. 1996. The effect of channel geometry on gravitational circulation in partially
mixed estuaries. Buoyancy Effects on Coastal and Estuarine Dynamics, ed. D. Aubrey and C. Fredricks, American
Geophysical Union, 283-300.
Kuo, A. Y., J. Shen, and J. M. Hamrick. 1996. The effect of acceleration on bottom shear stress in tidal estuaries.
Journal of Waterways, Ports, Coastal and Ocean Engineering. 122:75-83
Wu, T. S., J. M. Hamrick, S. C. McCutechon, and R. B. Ambrose. 1997. Benchmarking the EFDC/HEM3D surface
water hydrodymamic and eutrophication models. Next Generation Environmental Models and Computational
Methods, ed. G. Delich and M. F. Wheeler, Society of Industrial and Applied Mathematics, Philadelphia, 157-161.
Hamrick, J. M., and T. S. Wu. 1997. Computational design and optimization of the EFDC/HEM3D surface water
hydrodynamic and eutrophication models. Next Generation Environmental Models and Computational Methods, ed.
G. Delich and M. F. Wheeler, Society of Industrial and Applied Mathematics, Philadelphia, 143-156
Sucsy, P. V., F. W. Morris, M. J. Bergman, and L. D. Donnangelo. 1998. A 3-d model of Florida's Sebastian River
estuary. In Proceedings of the 5th International Conference Estuarine and Coastal Modeling, ed. M. L. Spaulding
and A. F. Blumberg, American Society of Civil Engineers, New York, 1998, pp. 59-74.
Shen, J., M. Sisson, A, Kuo, J. Boon, and S. Kim. 1998. Three-dimensional numerical modeling of the tidal York
River system, Virginia. Estuarine and Coastal Modeling. In Proceedings of the 5th International Conference
Estuarine and Coastal Modeling, ed. M. L. Spaulding and A. F. Blumberg, American Society of Civil Engineers,
New York, 1998, pp. 495-510.
Kim, S. C., L.D. Wright, J. P-Y. Maa, and J. Shen. 1998. Morphodynamic responses to extratropical meteorological
forcing on the inner shelf of the Middle Atlantic Bight: Wind waves, currents, and suspended sediment transport. In
Proceedings of the 5th International Conference Estuarine and Coastal Modeling, ed. M. L. Spaulding and A. F.
Blumberg, American Society of Civil Engineers, New York, 1998, pp. 456-466.
Shen, J. and A.Y. Kuo. 1999. Numerical investigation of an estuarine front and its associated topographic eddy.
Journal of Waterway, Port, Coastal, and Ocean Engineering. 125:127-135.
198
-------
Appendix A: Model Fact Sheets
Shen, I, J. D. Boon, and A. Y. Kuo. 1999. A modeling study of a tidal intrusion front and its impact on larval
dispersion in the James River estuary, Virginia. Estuaries. 22:681-692.
Yang, Z., T. Khangaonkar, C. DeGasperi, and K. Marshall. 2000. Three-dimensional modeling of temperature
stratification and density driven circulation in Lake Billy Chinook, Oregon. In Proceedings of the 6th International
Conference Estuarine and Coastal Modeling, ed. M. L. Spaulding and H. L. Butler, American Society of Civil
Engineers, New York, 2000, pp. 411-425.
Jin, K. R., J. M. Hamrick, and T. S. Tisdale. 2000. Application of a three-dimensional hydrodynamic model for Lake
Okeechobee. Journal of Hydraulic Engineering. 106:758-772.
Moustafa, M. Z., and J. M. Hamrick. 2000. Calibration of the wetland hydrodynamic model to the Everglades
nutrient removal project. Water Quality and Ecosystem Modeling. 1:141-167.
Wang, H. V. and S. C. Kim. 2000. Simulation of tunnel island and bridge piling effects in a tidal estuary. In
Proceedings of the 6th International Conference Estuarine and Coastal Modeling, ed. M. L. Spaulding and H. L.
Butler, American Society of Civil Engineers, New York, 2000, pp. 250-269.
Kim, S. C., J. Shen, C. S. Kim, and A. Y. Kuo. 2000. Application of VIMS HEM3D to a macro-tidal environment.
In Proceedings of the 6th International Conference Estuarine and Coastal Modeling, ed. M. L. Spaulding and H. L.
Butler, American Society of Civil Engineers, New York, 2000, pp. 238-249.
Hickey, K., I. Morin, M. Greenblatt, and G. Gong. 2000. 3D hydrodynamic model of an estuary in Nova Scotia. In
Proceedings of the 6th International Conference Estuarine and Coastal Modeling, M. L. Spaulding and H. L.
Butler, Eds., American Society of Civil Engineers, New York, 2000, pp. 1100-1111.
Boon, J. D., A. Y. Kuo, H. V. Wang, and J. M. Brubaker. 2000. Proposed third crossing of Hampton Roads, James
River, Virginia: Feature-based criteria for evaluation of model study results. In Proceedings of the 6th International
Conference Estuarine and Coastal Modeling, ed. M. L. Spaulding and H. L. Butler, American Society of Civil
Engineers, New York, 2000, pp. 223-237.
Ji, Z.-G., M. R. Morton, and J. M. Hamrick. 2001. Wetting and drying simulation of estuarine processes. Estuarine,
Coastal and Shelf Science. 53:683-700.
Hamrick, J. M., and Wm. B. Mills. 2001. Analysis of temperatures in Conowingo Pond as influenced by the Peach
Bottom atomic power plant thermal discharge. Environmental Science and Policy. 3:sl97-s209.
Jin, K. R., Z. G. Ji, and J. M. Hamrick. 2002. Modeling winter circulation in Lake Okeechobee, Florida. Journal of
Waterway, Port, Coastal, and Ocean Engineering. 128:114-125.
Ji, Z.-G., J. H. Hamrick, and J. Pagenkopf. 2002. Sediment and metals modeling in shallow river. Journal of
Environmental Engineering. 128:105-119.
Yang, Z. and J. M. Hamrick. 2002. Variational inverse parameter estimation in a long-term tidal transport model.
Water Resources Research. 38:10.
Yang, Z. and J. M. Hamrick. 2003. Variational inverse parameter estimation in a cohesive sediment transport model:
an adjoint approach Journal of Geophysical Research. 108(C2): 3055.
Wool, T. A., S. R. Davie, and H. N. Rodriguez. 2003. Development of three-dimensional hydrodynamic and water
quality models to support TMDL decision process for the Neuse River estuary, North Carolina. J. Water Resources
Planning and Management. 129:295-306.
Yang, Z. and J. M. Hamrick. In press. Optimal control of salinity boundary condition in a tidal model using a
variational inverse method. Estuarine, Coastal and Shelf Science.
199
-------
Appendix A: Model Fact Sheets
EPIC: Erosion Productivity Impact Calculator
Contact Information
Jimmy Williams
Texas A&M University - TABS
Blackland Research and Extension Center
720 E. Blackland Road
Temple, TX 76502
(254)774-6124
Williams (@,brc.tamus.edu
Avery Meinardus
Texas A&M University - TABS
Blackland Research and Extension Center
720 E. Blackland Road
Temple, TX 76502
(254)774-6110
epic(g),brc.tamus.edu
http://www.brc.tamus.edu/epic/
http://www.wiz.uni-kassel.de/model db/mdb/epic.html
Download Information
Availability: Nonproprietary
Cost: N/A
Model Overview/Abstract
EPIC assesses the effects of soil erosion on productivity and predicts the effects of management decisions on soil,
water, nutrient, and pesticide movements and their combined impact on soil loss, water quality, and crop yields for
areas with homogeneous soils and management.
Model Features
• Simulates erosion effects on water quality
• Crop management tool that examines sediment, nutrient, and pesticide transport processes
Model Areas Supported
Watershed Medium
Receiving Water None
Ecological None
Air None
Groundwater None
200
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
EPIC is a field-scale model that was developed to assess the effects of soil erosion on agricultural productivity and
water quality. It is used to examine farming practices and implementation activities.
EPIC has also been used widely for the study of global climate change. The USDA and Texas Agricultural
Experimental Station (Texas A&M) jointly developed this version of the model called Environmental Policy
Integrated Climate (EPIC).
Scientific Detail
EPIC is a continuous simulation model that has been used to examine long-term effects of various components of
soil erosion on crop production (Williams et al., 1983). EPIC is a public domain model that has been used to
examine the effects of soil erosion on crop production in over 60 different countries in Asia, South America, and
Europe. The model is used to examine soil erosion, economic factors, hydrologic patterns, weather effects, nutrients,
plant growth dynamics, and crop management. The major components in EPIC are weather simulation, hydrology,
erosion-sedimentation, nutrient cycling, pesticide fate, plant growth, soil temperature, tillage, economics, and plant
environment control.
The model requires input from GIS layers. These include soil series and weather data, although the model can
generate the necessary weather parameters. The model also requires management information that can be input from
a text file. Currently, there are many management files that exist for EPIC, and an effort is underway to catalog these
files and provide them to users. The model provides output on crop yields, economics of fertilizer use, and crop
values.
In the calculations for surface runoff, runoff volume is estimated by using a modification of the Soil Conservation
Service (SCS) curve number technique. There are two options for estimating the peak runoff rate—the modified
Rational formula and the SCS TR-55 method. The EPIC percolation component uses a storage routing technique to
simulate flow through soil layers. When soil water content exceeds field capacity, the water flows through the soil
layer. The reduction in soil water is simulated by a derived routing equation. Lateral subsurface flow is calculated
simultaneously with percolation. The evapotranspiration is calculated in four ways, using the following equations:
• Hargreaves and Samani
• Penman
• Priestley-Taylor
• Penman-Monteith
The water table height is simulated without direct linkage to other soil water processes in the root zone to allow for
offsite water effects. EPIC drives the water table up and down between input values of maximum and minimum
depths from the surface.
The EPIC precipitation model developed by Nicks is a first-order Markov chain model. Temperature and radiation
are simulated in EPIC by using a model developed by Richardson. The EPIC wind erosion model, WECS (Wind
Erosion Continuous Simulation), is used to calculate wind characteristics, including erosion due to the wind. The
relative humidity model simulates daily average relative humidity from the monthly average by using a triangular
distribution.
To simulate rainfall/runoff erosion, EPIC used six equations—the USLE, the Onstad-Foster modification of the
USLE, the MUSLE, two recently developed variations of MUSLE, and a MUSLE structure that accepts input
coefficients. The six equations are identical except for their energy components. Contaminants, such as nitrogen and
phosphorus, are used in the EPIC model. EPIC simulates the following processes involving contamination:
• Nitrate losses
• Contaminant transport due to soil water evaporation
• Organic nitrogen transport due to sediment
201
-------
Appendix A: Model Fact Sheets
• Damnification
• Mineralization
• Immobilization
• Nitrification
• Volatilization
• Soluble phosphorus loss in surface runoff
• Mineral phosphorus cycling
For climate change studies, the EPIC model appears to be the most complete model available for evapotranspiration
cover design. The most noteworthy example is the MINK (Missouri-Iowa-Nebraska-Kansas) study (Rosenberg and
Crosson, 1991). This study examined the effects of elevated CO2 (EPIC had to be modified to incorporate sensitivity
to CO2) and temperature on crop yields, soil erosion, and economics in this four state region. The MINK study also
provides general insights about the use of models for global change research.
Model Framework
• Field-scale, erosion based
Scale
Spatial Scale
• One-dimensional, agricultural field/farm scale
Temporal Scale
• Daily timestep, long-term simulations (1-4,000 years)
Assumptions
The model assumes that the dynamics of each physical, chemical, and biological component can be described by the
principle of conservation of mass.
Model Strengths
• Has been used extensively to examine the effects of soil erosion and agricultural processes.
• Describes the phosphorus cycle and differentiates between all forms of phosphorus.
• Can be used to simulate the fate of agricultural pesticides.
Model Limitations
• Cannot represent watershed subsurface flow.
• Does not simulate sediment routing in detail.
• No mention of how the model deals with tile drains.
Application History
See available literature.
Model Evaluation
EPIC has been used extensively in the United State and abroad to predict soil erosion and effects, along with the
potential costs associated with various management activities. See References for more information.
A soil loss model comparison was conducted by Bhuyan et al. 2002, which included evaluations of EPIC,
ANSWERS, and WEPP. Although the results from all three models were within the range of observed values in the
202
-------
Appendix A: Model Fact Sheets
case study, WEPP soil loss predictions were the most accurate. However, WEPP cannot be used to examine water
quality effects.
Model Inputs
• Daily timestep—long term simulations (1-4,000 years)
• Soil, weather, tillage, and crop parameter data supplied with model
• Soil profile can be divided into ten layers
• Homogeneous areas up to large fields
• Weather generation is optional
Users' Guide
Available online: http://www.wiz.uni-kassel.de/model db/mdb/epic.html
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS, UNIX
Programming language:
• FORTRAN version 5125
Runtime estimates:
• Minutes (1 sec./simulation year)
Linkages Supported
Unknown
Related Systems
APEX - small watershed scale agricultural model
Sensitivity/Uncertainty/Calibration
See references. No specific tools available.
Model Interface Capabilities
• Spatial-EPIC is a recently developed GIS-based application for EPIC
References
Williams, J.R., P.T. Dyke and C.A. Jones. 1983. EPIC: a model for assessing the effects of erosion on soil
productivity. In Analysis of Ecological Systems: State-of-the-Art in Ecological Modeling, ed. W.K. Laurenroth et al.
Elsevier, Amsterdam. pp553-572.
Jones, C.A., C.V. Cole, A.N. Sharpley, and J.R. Williams. 1984. A simplified soil and plantphosphorus model. Soil
Sci. Soc. Am. J. 48(4):800-805.
203
-------
Appendix A: Model Fact Sheets
Williams, J.R., C.A. Jones, and P.T. Dyke. 1984. A modeling approach to determining the relationship between
erosion and soil productivity. Trans. ASAE. 27:129-144.
J. Cavero, R.E. Plant, C. Sherman, J.R. Williams, J.R. Kiniry, and V.W. Benson. 1998. Application of EPIC Model
to Nitrogen Cycling in Irrigated Processing Tomatoes Under Different Management Systems. Agricultural Systems.
56(4):391-414.
S.W. Chung, P.W. Gassman, L.A. Kramer, J.R. Williams, and R. Gu. 1999. Validation of EPIC for Two Watersheds
in Southwest Iowa. J. Environ. Qual. 28:971-979.
J. Cavero, R. E. Plant, C. Sherman, D. B. Friedman, J. R. Williams, J. R. Kiniry, and V. W. Benson. Modeling
Nitrogen Cycling in Tomato-Safflower and Tomato-Wheat rotations. Agricultural System. 60:123-135.
Tharacad S. Ramanarayanan, M. V. Padmanabhan, G. N. Gajanan, Jimmy Willams. 1988. Comparison of Simulated
and Observed Runoff and Soil Loss on three Small United States Watersheds. NATO ASI Series. l(55):76-88.
J.G. Arnold, R. Srinivasan, R. S. Muttiah, J. R. Williams. 1998. Large Area Hydrologic Modeling and Assessment
Part I: Model Development. Journal of the American Water Resources Association. l(34):73-89.
J. R. Williams, J. G. Arnold. 1997. A System of Erosion-Sediment yield models. Soil Technology. 11:43-55.
Roloff, G., De Jong, R., Campbell, C.A. and Benson, V.W. 1998. EPIC estimates of soil water, nitrogen and carbon
under semi-arid temperate conditions. Can J. Soil Sci. 78:539-550.
204
-------
Appendix A: Model Fact Sheets
GISPLM: CIS-Based Phosphorus Loading Model
Contact Information
William W. Walker, Jr.
1127 Lowell Road
Concord, Massachusetts 01742-5522
(978) 369-8061
wwwalker@wwwalker. net
http ://www. wwwalker.net
Download Information
Availability: Nonproprietary
http://wwwalker.net/gisplm/index.htm
Cost: N/A
Model Overview/Abstract
GISPLM uses a spreadsheet interface to develop cost-effective strategies to reduce phosphorus loads from
watersheds. The watershed is defined as a number of subwatersheds or segments linked in a branched network.
Flows and phosphorus loads are evaluated using watershed features extracted from GIS, climatologic data, and other
local data. Phosphorus sources include runoff, farm animal populations, and point discharges. All calculations are
controlled from a Quattro Pro (version 7.0) workbook. The workbook also provides access to all input and output
screens. The workbook executes simulations by calling two FORTRAN programs, HYDRO and LOADS, which
predict surface runoff, stream flows, and phosphorus loads for each subwatershed. Flows and loads from each source
category (runoff, animal units, point sources) are totaled by model segment (subwatershed). Loads and flows totaled
by segment are routed downstream to the mouth of the watershed. Empirical models are used to estimate the
retention of phosphorus in lakes or impoundments optionally located at the downstream ends of segments. The user
defines load control options for each source category. Nonpoint source controls (BMPs) are defined for up to 12
land use categories. Point source controls for up to 3 treatment levels are defined based on effluent phosphorus
concentration and flow-dependent costs. The user specifies a target load reduction as a percentage of the load
predicted with no controls. GISPLM searches for the spatial allocation of controls, which achieves the target
reduction with minimum cost. Model results can be displayed spatially using Arc View 3.0 software.
Model Features
• Spreadsheet interface (Quattro Pro)
• Calculates surface runoff for each subwatershed
• Calculates phosphorus load through surface runoff for each subwatershed
• Flows and loads can be summarized from upstream subwatersheds to downstream outlet
• Simple empirical calculation for phosphorus retention in lakes and impoundments
• Capacity to define nonpoint source BMPs for up to 12 land use categories
• Capacity to define 3 treatment levels for point sources
• Calculates minimum treatment cost for a targeted reduction of phosphorus for the entire watershed
(optimization)
205
-------
Appendix A: Model Fact Sheets
Model Areas Supported
Watershed Medium
Receiving Water Low
Ecological None
Air None
Groundwater Very Low
Model Capabilities
Conceptual Basis
In GISPLM, a watershed is divided into many small subwatersheds to capture the spatial heterogeneity. Runoff and
phosphorus load are calculated for each subwatershed. Flow and loads can be summarized from the upstream
subwatersheds to the downstream outlet. Best management practices are specified for each source in each
subwatershed.
Scientific Detail
GISPLM is tool for developing cost-effective strategies to reduce phosphorus loads from watersheds. The watershed
is defined as a number of subwatersheds or segments linked in a branched network. Flows and phosphorus loads are
evaluated using watershed features extracted from GIS, climatologic data, and other local data. Phosphorus sources
include runoff, farm animal populations, and point discharges. All calculations are controlled from the Menu page of
the GISPLM. WB3 workbook. The Menu also provides access to all input and output screens.
HYDRO, a compiled FORTRAN program, predicts surface runoff from pervious areas for a user-defined date
interval. Calculations of daily runoff resulting from rainfall and snowmelt are driven by daily precipitation and air
temperature data. The algorithm and parameter estimates are taken from Generalized Watershed Loading Functions
or GWLF model (Haith et al, 1992). HYDRO generates a table relating unit area surface runoff from pervious areas
to SCS Runoff Curve Number. This table is later accessed by LOADS and the GISPLM workbook.
LOADS, another compiled FORTRAN program, calculates flows and phosphorus loads. The model reads watershed
data extracted from GIS databases and creates an index based on segment (subwatershed) number, model land use
code, and existing BMP code. LOADS calculates the total flow and load for each value of the index, accounting for
differences in soil group, soil origin, slope, and stream proximity. Runoff concentrations are specified as a function
of land use categories based on literature review. LOADS produces an output file containing the total area, flow,
load, impervious area, curve number, and surface runoff for each index. This file is subsequently accessed by
GISPLM workbook for subsequent processing.
The remaining calculations are performed within the GISPLM workbook. Flows and loads from each source
category (runoff, animal units and point sources) are totaled by model segment. Loads are adjusted to account for
existing phosphorus controls. Loads and flows are totaled by segment and routed downstream to the mouth of the
watershed. Empirical models (Vollenweider, 1976; Walker, 1987) are used to estimate the retention of phosphorus
in lakes or impoundments optionally located at the downstream ends of segments.
The user defines load control options for each source category. Nonpoint-source controls (BMPs) are defined for up
to 12 land use categories. Estimates of load reduction efficiency, capital cost, and annual operating cost are specified
for each BMP. Point-source controls for up to 3 treatment levels are defined based on effluent phosphorus
concentration and flow-dependent costs.
The user specifies a target load reduction as a percentage of the load predicted with no controls. GISPLM searches
for the spatial allocation of controls, which achieves the target reduction with minimum cost. Total annualized costs
are minimized. Estimates of capital and operating costs are also generated. Allocations can be constrained to provide
equal distribution of effort across source categories. Individual control measures can be specifically included or
excluded.
206
-------
Appendix A: Model Fact Sheets
Several graphical and tabular output formats are provided. These can be easily customized and manipulated within
the workbook to suit project needs. Model results can be displayed spatially using Arc View 3.0 software.
GISPLM is configured for application to the 137 km2 LaPlatte River watershed in Vermont. Guidance for
developing applications to other watersheds is provided.
Model Framework
• Watersheds are subdivided into subwatersheds with mixed land uses
• Land use, surface hydrology, and optional lakes or impoundments at the outlet of each subwatershed
Scale
Spatial Scale
• One-dimensional, subwatersheds
Temporal Scale
• Daily timestep
Assumptions
• Is a distributed model by land use but ignores the spatial location within a land use in a subwatershed
• Summarizes downstream flow and loads simply by adding the outputs from the upstream subwatersheds
• Describes BMP controls by their effectiveness and cost
Model Strengths
• Is a simple model
• Requires low level of expertise
• Can be quickly applied to evaluate phosphorus reductions due to lakes, impoundments, and BMP
implementations
• Performs optimization on the total BMP cost for the entire watershed, given a load reduction target
Model Limitations
• Simulates only phosphorus
• Does not simulate sediment and sediment phosphorus
• Results in weak simulation of nutrient fluxes because a constant concentration is used
• Includes highly simplified flow routing
• Highly simplifies groundwater inflow
• Bases BMP simulations on a single reduction effectiveness value
Application History
GISPLM was applied to the 137-km2 LaPlatte River watershed in Vermont. Guidance for developing applications to
other watersheds is provided in the users' manual.
Model Evaluation
Unknown
Model Inputs
• Daily weather data (mean temperature and precipitation)
207
-------
Appendix A: Model Fact Sheets
• Farm animal population data
• GIS data, including watershed boundaries, land use, soils (hydrologic group and slope class), streams, and
BMP locations and types
• Runoff phosphorus concentrations by land uses
• Point source flow and concentrations
• BMP cost and efficiency
Users' Guide
Available online (as part of installation package): http://wwwalker.net/gisplm
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• WINDOWS (95 or NT)
Programming language:
• Quattro Pro Macros and FORTRAN
Runtime estimates:
• Minutes
Linkages Supported
None
Related Systems
GWLF
Sensitivity/Uncertainty/Calibration
The users' manual provides simple guidance on model calibration.
Model Interface Capabilities
• Controls all calculations from a menu page of a spreadsheet workbook (GISPLM.WB3, Quattro Pro). The
workbook also provides access to all input and output screens.
• Can display model results spatially, using Arc View 3.0 software.
References
Haith, D.A., R. Mandel, R.S. Wu. 1992. GWLF - Generalized Watershed Loading Functions - Version 2.0. (User's
Manual). Department of Agricultural & Biological Engineering, Cornell University, Ithaca, New York.
Walker, W.W. 1987. Phosphorus Removal by Urban Runoff Detention Basin. In Lake and Reservoir Management
Volume 3, North American Lake Management Society, pp. 314-238.
Walker, W. W. 1987. Empirical Methods for Predicting Eutrophication in Impoundment Report 4. Applications
Manual Technical Report E-81-9. U.S. Army Corps of Engineers, Waterways Experiment Station, Vicksburg, MS.
208
-------
Appendix A: Model Fact Sheets
Walker, W. W. 1997. GISPLM User's Guide. LaPlatte River Phosphorus Modeling Project for Vermont Department
of Environmental Conservation, http://wwwalker.net/gisplm
Vollenweider, R.A. 1976. Advances in Defining Critical Loading Levels for Phosphorus in Lake Eutrophication.
Mem. 1st. Ital. Idrobiol. 33:53-83.
Vermont DEC & New York DEC. 1994. A Phosphorus Budget, Model, and Load Reduction Strategy for Lake
Champlain, Lake Champlain Diagnostic-Feasibility Study. Final Report.
209
-------
Appendix A: Model Fact Sheets
GLEAMS: Groundwater Loading Effects of Agricultural Management
Systems
Contact Information
Frank M. Davis
Southeast Watershed Research Laboratory (SEWRL)
South Atlantic Area
P. O. Box 946
Tifton, GAS 1793
(229) 391-6846
fmd@tifton.cpes.peachnet.edu
http://sacs.cpes.peachnet.edu/sewrl/
Download Information
Availability: Nonproprietary
http://sacs.cpes.peachnet.edu/sewrl/Gleams/gleams y2k update.htm
Cost: N/A
Model Overview/Abstract
Groundwater Loading Effects of Agricultural Management Systems (GLEAMS) is an extension of Chemicals,
Runoff, and Erosion from Agricultural Management Systems (CREAMS) model. GLEAMS, a continuous
simulation, field-scale model assumes that a field has homogeneous land use, soils, and precipitation. The four
major components of the model are hydrology, erosion/sediment yield, pesticide transport, and nutrients. It also
estimates surface runoff and sediment losses from the field. GLEAMS can be used to evaluate the impact of farm-
level management practices on potential pesticide and nutrient leaching within, through, and below the root zone.
GLEAMS can provide estimates of the impact management systems, such as planting dates, cropping systems,
irrigation scheduling, and tillage operations, have on the potential for chemical movement. GLEAMS can also be
useful in long-term simulations for pesticide screening of soil/management. The model tracks movement of
pesticides with percolated water, runoff, and sediment. Upward movement of pesticides and plant uptake are
simulated with evaporation and transpiration. Degradation into metabolites is also simulated for compounds that
have potentially toxic bi-products. Erosion in overland flow areas is estimated using a modified Universal Soil Loss
Equation. Erosion in chemicals and deposition in temporary impoundments such as tile outlet terraces are used to
determine sediment yield at the edge of the field.
Model Features
• Edge of field simulation model
Model Areas Supported
Watershed Low
Receiving Water None
Ecological Medium
Air None
Groundwater Medium
210
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
GLEAMS is a physically based field-scale model.
Scientific Detail
The hydrology component of GLEAMS uses a mass balance approach and represents the principal hydrologic
processes of infiltration, runoff, water application by irrigation, evapotranspiration, and soil water movement within
and through the root zone. Runoff calculation is based on the modified Soil Conservation Service (SCS) curve
number method. Percolation is calculated using storage-routing technique. Plant evapotranspiration is calculated
using either Priestley-Taylor or Penman-Monteith method. Erosion is calculated as detachment and transport
processes using USLE elements. The nutrient component of GLEAMS simulates the nitrogen and phosphorous
cycles. The pesticide component of the model calculates the daily decay based on the pesticide half-life. Based on
the partition coefficient, a portion of the pesticide is lost into runoff solution and the other part into the soil phase.
Model Framework
• Edge-of-field and bottom-of-root zone simulations of water, nutrients and pesticides
• Mainly used to simulate management systems in agricultural land
Scale
Spatial Scale
• One-dimensional field-scale
Temporal Scale
• Daily
Assumptions
• Uses a lumped parameter approach
• Assumes a spatially homogenous agricultural field
Model Strengths
• Is a simple model with few input requirements
Model Limitations
• Is limited to an agricultural field of very small size
• Is not suited for bigger watersheds
• Is not suited for urban land uses
Application History
GLEAMS is developed as an improvement over CREAMS model. Both models have sufficient application history.
http://sacs.cpes.peachnet.edu/sewrl/Gleams/glmspub.htm.
Model Evaluation
Many peer-reviewed publications are available for GLEAMS. Few studies are conducted to evaluate the accuracy of
GLEAMS and to compare with similar models like EPIC and WEPP.
http://sacs.cpes.peachnet.edu/sewrl/Gleams/glmspub.htm.
211
-------
Appendix A: Model Fact Sheets
Model Inputs
The inputs are provided separately for hydrology, erosion, pesticides, and nutrient components of the model. The
input requirements of the hydrology model include
• Daily precipitation
• Mean monthly minimum and maximum temperatures or mean daily temperature
• Mean monthly solar radiation
• Mean monthly wind movement and dew point temperature, if Penman-Monteith method is chosen for
evapotranspiration calculation
• Soil composition
The input requirements of the erosion component include
• Overland flow profile (length and slope)
• Soil properties (credibility and horizon depths)
• Overland flow channel rating-curve properties
The pesticide component's input requirements include
• Crop rotation information
• Water solubility and partitioning coefficient of pesticide
• Half-life, initial concentration, and fraction available for washoff for foliage and soil
• Crop uptake coefficient
The nutrient component's input requirements include
• Crop rotation information
• Initial soil concentration and concentrations of nutrients in rainfall and irrigation water
• Fertilizer application rate
• Crop uptake coefficient
Users' Guide
Available online:
http://www.cpes.peachnet.edu/sewrl/Gleams/gleams y2k update.htm#GLEAMS%20V3.0%20Revisions
Technical Hardware/Software Requirements
Computer hardware:
• IBM-PC
Operating system:
• PC-DOS
Programming language:
• FORTRAN
Runtime estimates:
• Minutes
Linkages Supported
None
212
-------
Appendix A: Model Fact Sheets
Related Systems
CREAMS is the predecessor of GLEAMS.
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Arc View GIS interface (see Tucker et al. 1996 and
http://www3.bae.ncsu.edu/Regional-Bulletins/Modeling-Bulletin/asae 2227-draft-extra.html')
References
Knisel, W.G., and P.M. Davis. 2000. GLEAMS: Grounchvater Loading Effects of Agricultural Management Systems.
Version 3.0. Publication No. SEWRL-WGK/FMD-050199. U.S. Department of Agriculture, Agricultural Research
Service, Southeast Watershed Research Laboratory, Tifton, GA. 191 pp.
Leonard, R. A., W. G. Knisel, and D. A. Still. 1987. GLEAMS: Groundwater loading effects of agricultural
management systems. Trans. ASAE. 30(5):1403-1418.
Tucker, M. G., D. L. Thomas, and D. D. Bosch. 1996. GLEAMS and REMM GIS-based model system: results and
sensitivity ofhydrologic components. ASAE Technical Paper No. 96-2022. ASAE, St. Joseph, MI.
More publications: http://sacs.cpes.peachnet.edu/sewrl/Gleams/glmspub.htm
213
-------
Appendix A: Model Fact Sheets
GLLVHT: Generalized, Longitudinal-Lateral-Vertical Hydrodynamics and
Transport
Contact Information
J.E. Edinger Associates, Inc.
37 West Avenue
Wayne, PA 19087
(601)293-0965
jeeai(@jeeai.com
www.jeeai.com
http://www.jeeai.com/Softwares-GEMSS.htm
Download Information
Availability: Proprietary
Cost: Unknown
Model Overview/Abstract
GLLVHT is a three-dimensional numerical model that provides solutions for rivers, lakes, estuaries, and coastal
waters. GLLVHT has five modules—hydrodynamics, water quality, sediment transport, particle tracking, and oil-
spill simulation. The hydrodynamics module provides the transport information for all the other modules. The water
quality module is a modified version of EUTRO5. The sediment transport module simulates the processes of
settling, flocculation, deposition, erosion/resuspension, bed load, armoring, bed structure, slump failure, and bed
deformation.
Model Features
• Hydrodynamic, temperature, and salinity
• Tracer simulation
• Suspended solids and sediment transport
• Nutrients, dissolved oxygen, and phytoplankton
• Coliform/bacteria
• Toxic organics and metals
Model Areas Supported
Watershed None
Receiving Water High
Ecological Medium
Air None
Groundwater None
Model Capabilities
Conceptual Basis
The waterbody is conceptualized as a series of grid points on a quasi-curvilinear coordinate system.
214
-------
Appendix A: Model Fact Sheets
Scientific Detail
The governing equations of the hydrodynamic module are developed from the horizontal momentum balance,
continuity, constituent transport, and the equation of state. The hydrostatic assumption is applied to simply the
vertical momentum balance. A zero equation turbulence scheme is utilized to close the governing equations. The
velocities and water surface elevations are obtained by solving the governing equations with a semi-implicit
numerical scheme. Transport equations are solved with an explicit scheme. A high-order difference of space is
applied for solving the transport. The transport of water quality variables and sediments are solved using the same
numerical scheme for temperature and salinity. Other processes, such as the kinetic reactions of water quality and
sediment settling and erosion, are considered as the source and sink terms in the governing equations.
Model Framework
• Three-dimensional
• River, lake, reservoir, estuary, and ocean
Scale
Spatial Scale
• One-, two-, and three-dimensional
Temporal Scale
• Variable timestep
Assumptions
• Hydrostatic assumption, Boussinesq approximation
Model Strengths
• Coupled hydrodynamic and transport model
• High-order accuracy transport scheme similar to ULTIMATE scheme
Model Limitations
• Zero equation turbulence closure scheme (Prandtl mixing length)
• Vertical Z grid (not able to follow the bathymetry)
• No sediment diagenesis
• Single phytoplankton group
Application History
Examples of application include modeling various processes such as intake entrainment; jet discharge; biochemical
oxygen demand; plume; and cooling water discharge into rivers, lakes, reservoirs, estuaries, and coastal waters, such
as Humboldt River, Nevada; Grand Lake, New Brunswick; Nechako Reservoir, British Columbia; Delaware
Estuary, Delaware; and San Diego Bay, California.
Model Evaluation
The theory and model development are published in several peer review journal papers.
Model Inputs
• Initial conditions
• Bathymetry and waterbody boundaries
215
-------
Appendix A: Model Fact Sheets
• Physical coefficients
• Water surface elevations or flow rate at open boundary (boundary conditions)
• Time sequences of hydrometeorological conditions
• Load of water quality variables
Users' Guide
GLLVHT Model, Technical Information and User's Guide
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
Microsoft Windows 95/98/2000, XP, or Microsoft Windows NT 4.0
Programming language:
• FORTRAN 90
Runtime estimates:
• Minutes to hours
Linkages Supported
EUTRO, CE-QUAL-ICM, JEEAI's GITF
Related Systems
EFDC, ECOMSED, WASP/EUTRO, CE-QUAL-ICM
Sensitivity/Uncertainty/Calibration
Not available. No specific supporting tools included.
Model Interface Capabilities
• GEMSS provides pre-processing and post-processing functions
References
J.E.Edinger Associate, Inc. GLLVHT Model, Technical Information and User's Guide. J.E.Edinger Associate, Inc.,
Wayne, PA.
216
-------
Appendix A: Model Fact Sheets
GSSHA: Gridded Surface Subsurface Hydrologic Analysis
Contact Information
WMS (including GSSHA):
Barbara Parsons
U.S. Army Corps of Engineers
Engineer Research and Development Center
Waterways Experiment Station
Coastal and Hydraulics Laboratory
Hydrologic Systems Branch
3 909 Halls Ferry Road
Vicksburg,MS39180
(601)634-2344
Barbara.A.Parsons(@,erdc.usace.army.mil
http://chl.wes.army.mil/software
GSSHA/CASC2D:
Fred L. Ogden
University of Connecticut
Dept. of Civil & Environmental Engineering, U-37
Storrs, CT 06269
(860)486-2771
ogden(g),engr.uconn.edu
http://www.engr.uconn.edu/~ogden/
Download Information
Availability:
GSSHA is included in WMS 2D Hydrology Package and can be purchased from EMS-I. Demonstration version is
free for download from the EMS-I website: http://www.ems-i.com/home.html
Cost: $2400 for WMS 7 2D Hydrology Package (includes Map/Data/Drain/Grid/GSSHA)
$1000 for GSSHA Model + 2D Grid Package (requires Map, Data, Drainage)
Model Overview/Abstract
GSSHA is a reformulation and enhancement of CASC2D developed by the Hydrologic Systems Branch of the U.S.
Army Corps of Engineers' Coastal and Hydraulics Laboratory (Downer et al., 2002). GSSHA retains the
functionality of CASC2D, but adds non-Hortonian runoff as well as many other features. Other improvements in
GSSHA that go beyond CASC2D include Richard's equation infiltration, lakes, wetlands, detention basins,
improved computational routines, full-dynamic wave channel routing with tidal influence, hydraulic structures with
rule curves, rating curves, scheduled releases, and improved erosion and sediment transport routines. CASC2D
development was initiated in 1989 at the Center for Excellence in Geosciences at Colorado State University funded
by the U.S. Army Research Office (ARO). GSSHA is also one of the surface-water hydrologic models supported by
the Watershed Modeling System (WMS).
GSSHA is a fully unsteady, physically based, distributed-parameter, square-grid, two-dimensional, hydrologic
model for simulating the response of a watershed subject to rainfall (Ogden, 2001) on an event or a continuous basis.
Major processes simulated include continuous soil-moisture accounting, precipitation distribution, snow
accumulation and melting; rainfall interception, infiltration, evapotranspiration, surface water retention, overland
flow routing, channel flow routing, unsaturated zone two-dimensional lateral flow, saturated zone groundwater flow;
overland sediment erosion, transport, and deposition; and sediment channel routing. GSSHA allows the user to
217
-------
Appendix A: Model Fact Sheets
select a grid size (typically 30-200 m) that appropriately describes the spatial variability in all watershed
characteristics. The physically based distributed model is superior in simulation of runoff process at small scales
within the watershed. As a spatially distributed model, GSSHA offers the capability of determining the value of any
hydrologic variable at any grid point in the watershed at the expense of requiring significantly more input than
traditional approaches. GSSHA can accept spatially varied hydrologic parameter input or rainfall input; however,
because of the extensive amount of data, data uncertainty may result in a non-unique calibration.
Model Features
• Grid-based hydrologic model
• Precipitation distribution and snow accumulation and melting
• Rainfall interception, infiltration, evapotranspiration, surface water retention, overland flow routing,
channel flow routing, unsaturated zone two-dimensional lateral flow, saturated zone groundwater flow
• Overland sediment erosion, transport, and deposition
• Sediment channel routing
• Lakes, wetlands, and detention basins simulation
• Hydraulic structures with rule curves, rating curves, and scheduled releases
Model Areas Supported
Watershed High
Receiving Water Medium
Ecological None
Air None
Groundwater Medium
Model Capabilities
Conceptual Basis
The watershed is divided into a uniform finite difference grid. Processes that occur before, during, and after a
rainfall event are calculated for each grid, routed through the flow direction, and integrated to produce the watershed
output.
Scientific Detail
GSSHA is physically based and solves the equations of conservation of mass and energy to determine the timing
and path of runoff in the watershed. GSSHA applies Green and Ampt, with or without a redistribution method, and
optional Richard's equation for infiltration simulation; an explicit finite-difference, two-dimensional, diffusive-wave
method for overland flow routing; and options of explicit one-dimensional, diffusive-wave, or implicit dynamic-
wave channel routing. Snowmelt is simulated based on energy balance. GSSHA applies bucket model or Richard's
equation for computing soil moisture in the vadose zone; Deardoff or Penman-Monteith with seasonal canopy
resistance method for evapotranspiration; and Darcy's Law for stream/groundwater interaction and exfiltration. The
empirical Kilinc and Richardson soil erosion model, as modified by Mien (1995), is applied in GSSHA to
determine the sediment transport from one overland flow grid cell to the next. GSSHA employs Yang (1973)
method to routing sand-size sediment in stream channels. Silt and clay size sediment is assumed to be transported
with flow; therefore, deposition or erosion of silt and clay within the channels is neglected (Ogden, 1998). A
physically based nutrient module has been incorporated into the GSSHA and can simulate N and P transformation
based on the SWAT nitrogen and phosphorus cycle kinetics.
Model Framework
• Grid-based
218
-------
Appendix A: Model Fact Sheets
Scale
Spatial Scale
• Two-dimensional grid overland
• One-dimensional channel network
Temporal Scale
• Variable timestep (seconds to minutes)
Assumptions
• Sediment discharge by means of overland flow is related to flow rate, soil credibility, and surface condition
• Deposition or erosion of silt and clay within the channels is negligible
Model Strengths
• Fully unsteady physically based distributed watershed model at a user-specified resolution
• Offers fully dynamic hydraulic channel routing
• Uses diffusive wave method to route overland flow
• Performs continuous simulation using variable timestep
Model Limitations
• Splash overland erosion is not considered
• Requires extensive input data preparation and calibration
Application History
It has been mainly used by the U.S. Army Corps of Engineers (USAGE). CASC2D was recently applied to study the
extreme flood on the Rapidan River, Virginia, on June 27, 1995 (Smith et al., 1996), for the purpose of examining
geomorphological changes. CASC2D was also used by the USAGE to evaluate the extreme urban flood event in
Trenton, New Jersey (Stock, C.A., B.S. Thesis, Princeton University, May 1997) for the purpose of recommending
stormwater management improvements. CASC2D is currently being applied to evaluate the impact of radar-rainfall
estimation errors in a study funded by ARO and in an NSF-sponsored study of the devastating flood that severely
impacted Fort Collins, Colorado, on June 28, 1997.
Model Evaluation
Studies have been conducted on GSSHA's predecessor, CASC2D. Recent experiences with CASC2D have shown
that in regions of infiltration-excess (Hortonian) runoff production, CASC2D is quite accurate at predicting runoff,
even at internal locations within the watershed (Johnson et al., 1993, Ogden et al., 1998). The continuous simulation
capability of CASC2D has been found to be particularly good for reducing the uncertainty in estimating initial soil-
moisture conditions and for improving calibration uniqueness (Ogden and Senarath 1997, Ogden et al., 1998).
CASC2D has also proven to be valuable for studying extreme runoff events. The overland erosion/sediment
transport capabilities of CASC2D were evaluated in detail by Johnson (1997). In upland areas, the method was
shown to calculate sediment yield well within the acceptable range of -50% to +200%. Compared with actual field
observations of annual sediment yield, CASC2D predictions were generally within 20% of observed values.
Model Inputs
• Rainfall (intensity, duration, and start time)
• Grid setup (grid size, number of rows and columns, and outlet grid row and column)
• For each grid (infiltration parameters, retention-interception parameters, soil properties, and canopy
parameters)
• For each channel (cross-section information, slope, Manning's n, and initial and boundary conditions)
219
-------
Appendix A: Model Fact Sheets
Users' Guide
Available online (as part of the WMS document package): http://www.ems-i.com/home.html
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS and Windows
Programming language:
• C
Runtime estimates:
• Minutes to hours
Linkages Supported
• WMS 2D Hydrology Package
Related Systems
WMS, CASC2D
Sensitivity/Uncertainty/Calibration
GHHSA provides an automated calibration procedure using the shuffled complex evolution (SCE) method.
Model Interface Capabilities
• The WMS system provides full input data preparation and output display capabilities.
• The GRASS GIS, developed by the U.S. Army Construction Engineering Research Laboratories, can be
used in the preparation of CASC2D datasets.
References
Downer, C. W., and F. L. Ogden. 2002. GSSHA-User's Manual, Gridded Surface Subsurface Hydrologic Analysis
Version 1.43 for WMS 6.1. ERDC Technical Report, Engineer Research and Development Center, Vicksburg,
Mississippi.
Downer, C.W., and F.L. Ogden. 2004. GSSHA: A model for simulating diverse streamflow generating processes. J.
Hydrol. Engrg. 9(3): 161-174.
Downer, C.W., and F.L. Ogden. 2004. Appropriate Vertical Discretization of Richards' Equation for Two-
Dimensional Watershed-Scale Modelling. Hydrological Processes. 18:1-22.
Downer, C.W., and F.L. Ogden. 2004. Prediction of runoff and soil moistures at the watershed scale: Effects of
model complexity and parameter assignment. Water Resour. Res. 39(3).
220
-------
Appendix A: Model Fact Sheets
Julien, P.Y., B. Saghafian, and F.L. Ogden. 1995. Raster-based Hydrological Modeling of Spatially-varied Surface
Runoff. Water Resources Bulletin. 31(3):523-536.
Johnson, B.E. 1997. Development of a Storm Event Based Two-Dimensional Upland Erosion Model. Ph.D. diss.,
Colorado State University, Fort Collins, CO.
Johnson, B.E., N.K. Raphelt, and J.C. Willis. 1993. Verification of Hydrologic Modeling Systems USGS Water
Resources Investigations Report 93-4018. In Proceedings of the Federal Water Agency Workshop on Hydrologic
Modeling-Demands for the 90's, June 6-9, 1993, Sec. 8.9-20.
Ogden, F.L. 1998. CASC2D Reference Manual version 1.18. (Computer program manual). University of
Connecticut, Storrs, CT. Available at http://www.engr.uconn.edu/~ogden/casc2d/.
Ogden, F.L., S.U.S. Senarath, and B. Saghafian. 1998. Use of Continuous Simulations to Improve Distributed
Hydrologic Model Calibration Uniqueness. Unpublished paper.
Ogden, F.L., and S.U.S. Senarath. 1997. Continuous Distributed Parameter Hydrologic Modeling with CASC2D. In
Proceedings of the XXVII Congress, International Association of Hydraulic Research, San Francisco, CA, Aug. 10-
15, 1997.
Smith, J.A., M.L. Baeck, and M. Steiner. 1996. Catastrophic rainfall from and upslope thunderstorm in the central
Appalachians: The Rapidan storm of June 27, 1995. Water Resources Research. 32(10):3099-3113.
Stock, C.A. 1997. B.S. Thesis, Princeton University, Princeton, New Jersey.
Yang, C.T. 1973. Incipient motion and sediment transport. Journal Hydraulics. 99(HY10): 1679-1704.
221
-------
Appendix A: Model Fact Sheets
GWLF: Generalized Watershed Loading Functions
Contact Information
Douglas A. Haith
Department of Agricultural and Biological Engineering
Cornell University
Ithaca, NY 14853
(607) 255-2802
Download Information
Availability: Nonproprietary
The original version of the model has been used for 15 years and can be obtained from Dr. Haith at Cornell
University.
A Windows interface is available at http://www.vims.edu/bio/vimsida/basinsim.html (Dai et al, 2000).
Perm State University developed an Arc View interface for GWLF (http://www.avgwlf.psu.edu/) and compiled data
for the entire State of Pennsylvania (Evans et al., 2002).
Cost: N/A
Model Overview/Abstract
GWLF is used for the simulation of mixed land use watersheds to evaluate the effect of land use practices on
downstream loads of sediment and nutrients (N, P). Typically used to evaluate long-term loadings, GWLF was
developed at Cornell University as "a compromise between the empiricism of export coefficients and the complexity
of chemical simulation models" (Haith and Shoemaker, 1987). As a loading function model, it simulates runoff and
sediment delivery using the curve number (CN) and Universal Soil Loss Equation (USLE), combined with average
nutrient concentration, based on land use. Because of the lack of detail in predictions and stream routing, the outputs
are only given monthly, although they are calculated on a daily basis.
More recently, an in-stream routing and sediment transport component has been added and linked to BasinSim
GWLF model with a generic Arc View interface that is able to utilize the national land use and soil GIS data. The
new component employs the algorithms in the Annualized Agricultural Nonpoint Source Model (AnnAGNPS) to
simulate sediment transport. The Arc View interface automatically prepares the input files for GWLF and generates
stream network that links multiple subwatersheds in a study area. Sediment transport is simulated using three
particle size classes (clay, silt, and sand). The Muskingum-Cunge method is used for flow routing. The GWLF
output is interpolated for small timesteps (daily or subdaily) to meet the requirement of the in-stream modeling
component. The new enhanced GWLF modeling system (the generic Arc View interface, BasinSim GWLF, and the
in-stream routing/transport module) is currently being applied to West Virginia TMDL projects (Tetra Tech, Inc.,
Fairfax, Virginia).
Model Features
• Calculation of water, sediment, and total and dissolved nitrogen and phosphorus from a watershed with
mixed land uses
• Low input data requirements
Model Areas Supported
Watershed Medium
Receiving Water None (original version), Low (Tetra Tech version)
222
-------
Appendix A: Model Fact Sheets
Ecological None
Air None
Groundwater Low
Model Capabilities
Conceptual Basis
GWLF was developed as "a compromise between the empiricism of export coefficients and the complexity of
chemical simulation models" (Haith and Shoemaker, 1987). As a loading function model, it simulates runoff and
sediment delivery using very simple, yet widely acceptable, algorithms, combined with average nutrient
concentration, based on land use.
Scientific Detail
Runoff is calculated by means of the SCS curve number equation. USLE is applied to simulate erosion. Rural
nutrients are estimated based on empirical concentrations of each land use, which are based on both dissolved
concentration in runoff and solid concentration in sediment. Urban nutrient loads are computed by exponential
accumulation and washoff functions. Nutrient loads from septic systems are calculated by estimating the per capita
daily load from each type of septic system considered and the number of people in the watershed served by each
type. Groundwater runoff and discharge are obtained from a lumped-parameter watershed water balance for both
shallow saturated and unsaturated zones.
Model Framework
• Watershed with mixed land uses
• Land use, surface hydrology, unsaturated soil zone, shallow saturated zone, and deep saturated zone
Scale
Spatial Scale
• One-dimensional, subwatershed overland
Temporal Scale
• Daily input
• Monthly output
Assumptions
• It is a distributed model by land use but ignores the spatial location within a land use in a subwatershed
• A unit (watershed or subwatershed) is divided into surface, unsaturated zone, and saturated zone
• Pollutant parameters values are based on data of a particular study area
Model Strengths
• It is a simple model
• It requires low level of expertise
• It can be quickly applied to evaluate potential loadings with some recognition of seasonal variability
Model Limitations
• Simplifications in stream transport and water quality simulation
• Simulation of peak nutrient fluxes is weak because a constant concentration is used
• Highly simplified flow routing
• Groundwater inflow represented using a user-defined recession coefficient
223
-------
Appendix A: Model Fact Sheets
• Stormwater storage and treatment are not considered
Application History
GWLF has been used in studies and TMDL development nationally. It is suitable for application to generalized
watershed loading, source assessment, and seasonal and interannual variability. It has been extensively used in
northeast and mid-Atlantic regions. It has been adopted by Pennsylvania as state system for TMDL development and
agricultural land management.
Model Evaluation
GWLF validations have been published in a number of peer-reviewed studies, including Haith and Shoemaker
(1987), Howarth et al. (1991), Swaney et al. (1996), Lee et al. (2000), and Schneiderman et al. (2002), who made
several modifications to the model. The algorithms on which the model is based are widely used and accepted. The
model is in the public domain and the source code is available through Cornell University.
Model Inputs
• Climate: daily precipitation and temperature data and runoff source areas
• Transport parameters: runoff curve numbers, soil loss factor, evapotranspiration cover coefficient,
groundwater recession and seepage coefficients, and sediment delivery ratio
• Chemical parameters: urban nutrient accumulation rates, dissolved nutrient concentrations in runoff, and
solid-phase nutrient concentrations in sediment
• Septic System: per capita nutrient load, system effluent, nutrient losses due to plant uptake, and people
served by septic system
• Point source discharge and concentration
Users' Guide
Complete documentation is readily available (contact Dr. Haith for a hardcopy). The manual (Haith et al., 1992) is
available as part of the BasinSim manual (Dai et al., 2000) at http://www.vims.edu/bio/vimsida/basinsim.html.
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS, Windows
Programming language:
• BASIC, Visual BASIC
Runtime estimates:
• Seconds to minutes
Linkages Supported
Tetra Tech has developed an ArcGIS interface and a stream network transport model for BasinSim (GWLF) for
Windows
224
-------
Appendix A: Model Fact Sheets
Related Systems
BasinSim for Windows (http://www.vims.edu/bio/models/basinsim.htmlX AVGWLF (http://www.avgwlf.psu.edu/)
Sensitivity/Uncertainty/Calibration
The BasinSim interface provides a plotting tool for users to compare the simulated monthly watershed flow,
sediment, and nutrients loadings with observed data.
Model Interface Capabilities
• DOS version: Option menu and graphic and text output
• Windows version: Standard Windows features
• ArcGIS interface: Perm State AVGWLF; Tetra Tech Arc View for BasinSim
References
Delwiche, L.L.D., and D.A. Haith. 1983. Loading functions for predicting nutrient losses from complex watersheds.
Water Resources Bulletin. 19(6):951-959.
Haith, D.A. 1985. An event-based procedure for estimating monthly sediment yields. Transactions of the American
Society of Agricultural Engineers. 28(6): 1916-1920.
Haith, D.A. and L.L. Shoemaker, 1987. Generalized Watershed Loading Functions for Stream Flow Nutrients.
Water Resources Bulletin. 23(3):471-478.
Haith, D.A. 1990. Mathematical models of nonpoint-source pollution. Cornell Quarterly. 25(1):26.
Haith, D.R., R. Mandel, and R.S. Wu. 1992. GWLF: Generalized Watershed Loading Functions User's Manual,
Vers. 2.0. (Computer program manual). Cornell University, Ithaca, NY.
Dai, T., R.L. Wetzel, Tyler R.L. Christensen and E.A. Lewis. 2000. BasinSim 1.0 A Windows-Based Watershed
Modeling Package User's Guide SRAMSOE #362. (Computer program manual). Virginia Institute of Marine
Science, School of Marine Science, College of William & Mary, Gloucester Point, VA. Available at
http://www.vims.edu/bio/vimsida/basinsim.html
Evans, B.M., D.W. Lehning, K.J. Corradini, G.W. Petersen, E. Nizeyimana, J.M. Hamlett, P.O. Robillard, and R.L.
Day. 2002. A Comprehensive GIS-Based Modeling Approach for Predicting Nutrient Loads in Watersheds. J.
Spatial Hydrology. 2(2). Available at http://www.spatialhvdrology.com/
Howarth, R.W., J.R. Fruci, D. Sherman. 1991. Inputs of sediment and carbon to an estuarine ecosystem: Influence of
land use. Ecological Applications. l(l):27-39.
Lee, K.-Y., T.R. Fisher, T.E. Jordan, D. L. Correll, and D. E. Weller. 2000. Modeling the hydrochemistry of the
Choptank River basin using GWLF and GIS. Biogeochem. 49: 143-173.
Parson, S.C. 1999. Development of an Internet watershed educational tool (INTERWET) for the Spring Creek
Watershed of central Pennsylvania. Ph.D. diss., The Pennsylvania State University, University Park, Pennsylvania.
Available at http://www.interwet.psu.edu/index.htm
Schneiderman, E.M., D.C. Pierson, D.G. Lounsbury, and M.S. Zion. 2002. Modeling the hydrochemistry of the
Cannonsville Watershed with Generalized Watershed Loading Functions (GWLF). J. Amer. Water Resour. Assoc.
38:1323-1347.
Swaney, D.P., D. Sherman, and R. W. Howarth. 1996. Modeling water, sediment, and organic carbon discharges in
the Hudson-Mohawk Basin: coupling to terrestrial sources. Estuaries. 19: 833-847.
225
-------
Appendix A: Model Fact Sheets
HEC-6: Scour and Deposition in Rivers and Reservoirs
Contact Information
U.S. Army Corps of Engineers
Institute for Water Resources
Hydrologic Engineering Center
609 Second Street
Davis, CA 95616-4687
(530)756-1104
hec.webmastertgiusace.army.mil
http://www.hec.usace.armv.mil/software/legacvsoftware/hec6/hec6.htm
Download Information
Availability: Nonproprietary
Cost: N/A
HEC-6 was developed for the U.S. Army Corps of Engineers, but the model program files, executables, and
documentation are available for free download by the public at the website above. However, the Hydrologic
Engineering Center (HEC) of the U.S. Army Corps of Engineers will not provide user support to non-Corps users. In
addition, public distribution of the model source code is generally discouraged by HEC.
Proprietary versions of HEC-6 are also available through venders that provide model distribution and user support
for a fee. A list of venders is included at the HEC website. Some proprietary versions of HEC-6 include enhanced
simulation capabilities that expand on limitations of HEC-6 and provide more user friendliness. For example, the
proprietary HEC-6T (MBH Software, Inc., 2002) provides additional plotting and hydraulic simulation capabilities
not available in the HEC-6 version downloadable from HEC.
Model Overview/Abstract
HEC-6 is a one-dimensional open channel flow model capable of simulating changes of river profile due to scour
and/or sediment deposition. Based upon flow records, a water surface profile is calculated that provides an energy
slope, velocity, and depth at each cross-section. These predictions are used to estimate potential sediment transport
rates at each section, which are considered with volume of flow and sediment yield from upstream sources to
determine the scour and deposition. Changes in bed elevation, which impacts channel geometry and subsequent
sediment transport potential, are also computed for each section. HEC-6 can be used to simulate both channel and
reservoir sediment deposition and can include analysis of impacts of dredging.
Model Features
• Water surface and energy profile simulation
• Sediment scour and deposition modeling
• Sediment transport modeling
• River geometry simulation
Model Areas Supported
Watershed None
Receiving Water High
Ecological None
Air None
226
-------
Appendix A: Model Fact Sheets
Groundwater None
Model Capabilities
Conceptual basis
Capability to analyze networks of streams, reservoirs, automatic channel dredging, various levee and encroachment
options, and several options for computation of sediment transport rates.
Scientific detail
HEC-6 (HEC, 1991a) simulates sediment bed and suspended load transport as a function of Einstein's Bed-Load
Function (1950) that assumes an alluvial stream with consistent sediment material between the streambed and that
moving in the stream. Based on characteristics of the stream hydraulics and the sediment material (e.g., grain size
distribution), the rate of sediment transport is calculated.
A one-dimensional energy equation (USAGE, 1959) is used to compute water surface profiles for characterization of
stream hydraulics. Flow conveyance limits, levee hydraulic assumptions, and hydraulic energy and resulting water
surface elevation are simulated in a manner similar to HEC-2 (HEC, 1991b). HEC-6 can be operated in a "fixed
bed" mode that is similar to a HEC-2 application for simulation of water surface elevation only.
Sediment transport rates can be estimated by HEC-6 for grain sizes up to 2048 mm. Different methods for sediment
transport are used by HEC-6 based on grain size and user specification. Sediment transport potential is based only
on hydraulic and sediment material characteristics. Boundary conditions for sediment loading at the river main stem,
tributaries, or inflow/outflow points can be specified to change with time.
Model Framework
The model can represent a river or reservoir system consisting of a main stem, tributaries, and local inflow/outflow
points in a one-dimensional mode. Inflowing sediment loads are related to water discharge by sediment-discharge
curves for the upstream boundaries.
Scale
Spatial Scale
• Operation unit one-dimensional
Temporal Scale
• Variable timesteps—Short timesteps must be taken during flood events when large amounts of sediment
are moving and the hydrograph is rapidly changing. Longer timesteps are used during low flow periods.
• This is discussed in further detail in the document Guidelines for the Calibration and Application of
Computer Program HEC-6 available at http://www.hec.usace.armv.mil/software/legacvsoftware/hec6/tdl3-
documentation. htm
Assumptions
• Bed material transport algorithms assume that equilibrium conditions are reached within each timestep.
• The cross section is subdivided into two parts representative of a movable and immovable bed based on
limits of the wetted perimeter and other considerations.
• The entire wetted part of the cross section is normally moved uniformly up or down; however, an option is
available to adjust the bed elevation in horizontal layers when deposition occurs.
• Irregularities of the streambed are not simulated, but Manning's n values can be specified as functions of
discharge that can be assumed to indirectly account for effects of bed forms.
227
-------
Appendix A: Model Fact Sheets
Model Strengths
• Simulates the sediment passing through each cross section and the volume of sediment deposited or
scoured at each section.
• Can be used for simulating changing sediment and hydraulic conditions
• Can be used for simulation and design of channel or reservoir dredging
Model Limitations
• Does not include capabilities for simulating the development of meanders or lateral distribution of sediment
load across a cross section.
• Does not simulate density and secondary currents.
• Designed to analyze long-term scour and/or deposition. Single flood event analyses must be performed
with caution.
• Sediment transport in diverging streams is not possible
• Flow around islands (i.e., closed loops) cannot be directly accommodated
• Only one junction or local inflow point is allowed between any two cross sections.
Application History
See available references.
Model Evaluation
Results of model testing and evaluation are reported extensively by HEC (1986, 1990a, 1990b, and 1991a).
Model Inputs
• Stream cross-sectional geometry and longitudinal elevation information
• Sediment particle characteristics
• Time series data of boundary inflows and sediment loading assumptions
Users' Guide
HEC-6, Scour and Deposition in Rivers and Reservoirs, User's Manual (HEC, 1991a). Available online:
http://www.hec.usace.armv.mil/software/legacvsoftware/hec6/hec6-documentation.htm.
Technical Hardware/Software Requirements
Computer hardware:
The minimum hardware requirements include
• 570 KB of RAM
• 20 MB of free disk space
Operating system:
PC-DOS. Two editions of the HEC-6 program are distributed in the HEC-6 package: "overlayed" and "extended
memory." While the basic programs are the same, the extended memory version runs faster and provides for up to
500 cross sections in a 10-stream network, whereas the overlayed version only allows 150 sections. The overlayed
version operates within the DOS 640K limit (570Kb RAM). The extended memory version requires a 386 (or better)
computer with 2-4MB extended memory and a math co-processor.
Programming language:
FORTRAN
228
-------
Appendix A: Model Fact Sheets
Runtime estimates:
Minutes to hours
Linkages Supported
HEC-DSS
Related Systems
HEC-6T, HEC-2
Sensitivity/Uncertainty/Calibration
HEC (1991a) provides a description of the sensitivity of simulated bed profile changes to various input data
uncertainties. A qualitative assessment of the sensitivity of model results to field data (geometry, sediment and
hydrology) is presented in the manual. HEC (1991a) also reports results of analyses of sensitivity to cross sections,
movable bed limits, roughness, bed material gradation, inflowing load, flow record, rating curve, and temperature.
Additional results of model sensitivity analyses to bed roughness is reported by HEC (1992). Apart from sensitivity,
HEC (1986 and 1991a), USACE (1992), and Gee (1984) provide detailed descriptions and guidance in calibration
and selection of hydraulic and sediment modeling parameters.
Model Interface Capabilities
HEC-DSS (HEC, 1990c) can be used for managing and displaying time series data when simulating for long time
periods.
References
Gee, Michael. 1984. Role of Calibration in the Application of HEC-6. Technical Paper No. 102. Hydrologic
Engineering Center, Davis, CA.
MBH Software, Inc. 2002. Sedimentation In Stream Networks (HEC-6T) - User Manual. (Computer program
manual). Available at http://www.mbh2o.com/docs.html
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 1986. Accuracy of Computed Water Surface
Profiles. Research Document No. 26. U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, CA.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 1990a. Computing Water Surface Profiles
with HEC-6 on a Personal Computer. Training Document No. 26. U.S. Army Corps of Engineers, Hydrologic
Engineering Center, Davis, CA..
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 1990b. HEC-2, Water Surface Profiles
User's Manual. U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, CA.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 1990c. HECDSS User's Guide and Utility
Program Manuals. CPD-45. U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, CA.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 1991a. HEC-6, Scour and Deposition in
Rivers and Reservoirs, User's Manual. U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC),
Davis, CA.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 1991b. HEC-2, Water Surface Profiles:
User's Manual. U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC), Davis, CA.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 1992. Guidelines for the Calibration and
Application of Computer Program HEC-6. Training Document No. 13. U.S. Army Corps of Engineers, Hydrologic
229
-------
Appendix A: Model Fact Sheets
Engineering Center, Davis, CA. Available at http://www.hec.usace.army.mil/software/legacvsoftware/hec6/td 13-
documentation. htm)
U.S. Army Corps of Engineers (USACE). 1992. River Hydraulics. DRAFT EM 1110-2-1415. U.S. Army Corps of
Engineers, Washington, D.C.
230
-------
Appendix A: Model Fact Sheets
HEC-HMS: Hydraulic Engineering Center Hydrologic Modeling System
Contact Information
U.S. Army Corps of Engineers
Institute for Water Resources
Hydrologic Engineering Center
609 Second Street
Davis, CA 95616-4687
(530)756-1104
hec.webmaster(@,usace.armv.mil
http://www.hec.usace.armv.mil/software/hec-hms/hechms-hechms.html
Download Information
Availability: Nonproprietary
Cost: N/A
HEC-HMS was developed for the U.S. Army Corps of Engineers, but the model program files, executables, and
documentation are available for free download by the public at the website above. However, the Hydrologic
Engineering Center (HEC) of the U.S. Army Corps of Engineers will not provide user support to non-Corps users. In
addition, public distribution of the model source code is generally discouraged by HEC.
Model Overview/Abstract
The Hydrologic Engineering Center (HEC) of the U.S. Army Corps of Engineers released the Hydrologic Modeling
System (HEC-HMS) (HEC, 2001a) for rainfall-runoff simulation as a successor to HEC-1 (HEC, 1998). This model
includes many of the watershed runoff and routing computation methods of HEC-1 but also includes many
additional capabilities, including continuous hydrograph simulation over longer periods of time, distributed runoff
computation using a grid cell representation of the watershed, a GUI, integrated hydrograph analysis tools, data
storage and management tools, and graphics and reporting packages. HEC (200 Ib) reports specific differences
between HEC-HMS and HEC-1.
HEC-HMS is specifically designed for simulation of rainfall-runoff processes of networking watershed systems. The
modeling system includes many modernized and expanded algorithms from previous HEC models, including HEC-
1, HEC-1F (HEC, 1989), PRECIP (HEC, 1989), andHEC-IFH (HEC, 1992).
Model Features
Modeling components
• Losses
• Runoff transform
• Open-channel routing
• Analysis of meteorologic data
• Rainfall-runoff simulation
• Parameter estimation
• Reservoir system simulations
User interface includes
• File management
231
-------
Appendix A: Model Fact Sheets
• Data entry and editing
• Basin mapping for model configuration and data input and access
• Tabular and graphical display of input and output data
Model Areas Supported
Watershed Low
Receiving Water Low
Ecological None
Air None
Groundwater Low
Model Capabilities
Conceptual basis
HEC-HMS is designed to simulate the rainfall-runoff processes of networked watershed systems. This model serves
as the successor to HEC-1, providing a user interface and improvements and additional capabilities for distributed
modeling and continuous simulation. It is designed to be applicable in a wide range of geographic areas for solving
the widest possible range of problems.
Scientific detail
The physical representation of watersheds or basins and rivers is configured in the model based on representation of
general hydrologic elements, including subbasins, reaches, junctions, reservoirs, diversions, sources, and sinks. The
system encompasses losses, runoff transform, open-channel routing, analysis of meteorological data, rainfall-runoff
simulation, and parameter estimation. A wide array of options is available to simulate losses, including initial and
constant rates, the SCS curve number method, and the Green-Ampt method. Runoff transform methods include the
Clark, Snyder, and SCS unit hydrograph techniques. User-specified unit hydrograph ordinates can also be used.
Open-channel routing methods include the lag method, Muskingum method, the modified Puls method, the
kinematic wave method, and the Muskingum-Cunge method. Meteorological data analysis can also be performed in
the model for precipitation and evapotranspiration and includes various historical and synthetic methods (HEC,
2001).
Model Framework
Each model run combines a basin model, meteorologic model, and control specifications with run options to obtain
results. The system connectivity and physical data describing the watershed are stored in the basin model. The
precipitation and evapotranspiration data necessary to simulate watershed processes are stored in the meteorologic
model (HEC, 2001).
Scale
Spatial Scale
• One-dimensional
Temporal Scale
• User-defined
Assumptions
Multiple assumptions are made that reduce the watershed to three separate processes—loss, transform, and
baseflow. The number of assumptions is controlled by the hydrologic methods selected by the user for simulation.
HEC (2000) reports specific assumptions for each method and algorithm used in HEC-HMS.
232
-------
Appendix A: Model Fact Sheets
Model Strengths
• Simplified methods of hydrologic simulation encourage reduced number of parameters for model
calibration.
• Capable of modeling common types of hydraulic control structures with appropriate on and off features.
• Includes a GUI with pre- and post-processing capabilities.
Model Limitations
• Cannot simulate water quality processes
• Relatively difficult to use in conjunction with other water quality models
• Cannot simulate groundwater levels
Application History
Multiple example model applications are reported by HEC (200la and 2002). Many algorithms from HEC-1, HEC-
1F, PRECIP, and HEC-IFH have been modernized and combined with new algorithms to form a comprehensive
library of simulation routines in HEC-HMS.
Model Evaluation
See available references.
Model Inputs
• Initial conditions and attributes
• Inputs from watershed sources and discharges
• Element data
• Physical coefficients
• Time sequences of hydrometeorological conditions
Users' Guide
Hydrologic Modeling System, HEC-HMS: User's Manual (HEC, 200 la). Available online:
http://www.hec.usace.armv.mil/software/hec-hms/documentation/hms user.pdf
Technical Hardware/Software Requirements
Computer hardware:
The minimum hardware requirements for a Microsoft Windows installation includes
• Intel 80486 compatible processor
• 16-MB memory to run the program individually
• 15-MB available hard-disk space
• 15-inch VGA monitor
• Microsoft compatible mouse
The minimum hardware requirements for a Unix installation includes
• 64-MB memory to run the program individually
• SuperSPARC processor
• 28-MB available hard-disk space
• 10-MB available hard-disk space per user
• 17-inch color monitor
Operating system:
• Microsoft Windows 2000, 98, and 95
233
-------
Appendix A: Model Fact Sheets
• Microsoft Windows NT 4.0
• Unix 2.5 or higher
Programming language:
FORTRAN
Runtime estimates:
Available intervals range from 1 minute to 24 hours
Linkages Supported
HEC-DSS
Related Systems
HEC-l, HEC-1F, PRECIP, HEC-IFH, HEC-RAS, HEC-DSS
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
The program features a completely integrated work environment including a database, data entry utilities,
computation engine, and results reporting tools. A GUI is also included.
References
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 1989. Water Control Software: Forecast
and Operations. (Computer program manual). U.S. Army Corps of Engineers, Hydrologic Engineering Center,
Davis, CA.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 1992. HEC-IFH Interior Flood Hydrology
Package: User's Manual. (Computer program manual). U.S. Army Corps of Engineers, Hydrologic Engineering
Center, Davis, CA.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 1998. HEC-l Flood Hydrograph Package:
User's Manual. (Computer program manual). U.S. Army Corps of Engineers, Hydrologic Engineering Center,
Davis, CA.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 2000. Hydrologic Modeling System, HEC-
HMS: Technical Reference Manual. CPD-74B. U.S. Army Corps of Engineers, Hydrologic Engineering Center,
Davis, CA. Available at http://www.hec.usace.armv.mil/software/hec-hms/documentation/hms technical.pdf.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 200la. Hydrologic Modeling System, HEC-
HMS: User's Manual. CPD-74A. U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, CA.
Available at http://www.hec.usace.armv.mil/software/hec-hms/documentation/hms user.pdf.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 200 Ib. Hydrologic Modeling System, HEC-
HMS: Differences Between HEC-HMS and HEC-l. CPD-74B. U.S. Army Corps of Engineers, Hydrologic
Engineering Center, Davis, CA. Available at http://www.hec.usace.army.mil/software/hec-
hms/documentation/hms differences.pdf.
234
-------
Appendix A: Model Fact Sheets
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 2002. Hydrologic Modeling System, HEC-
HMS: Applications Guide. CPD-74C. U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, CA.
Available at http://www.hec.usace.army.mil/software/riec-hms/documentation/hms applications.pdf.
235
-------
Appendix A: Model Fact Sheets
HEC-RAS: Hydrologic Engineering Centers River Analysis System
Contact Information
U.S. Army Corps of Engineers
Institute for Water Resources
Hydrologic Engineering Center
609 Second Street
Davis, CA 95616-4687
(530)756-1104
hec.webmaster(@,usace.armv.mil
http://www.hec.usace.armv.mil/software/hec-ras/hecras-hecras.html
Download Information
Availability: Nonproprietary
Cost: N/A
HEC-RAS was developed for the U.S. Army Corps of Engineers, but the model program files, executables, and
documentation are available for free download by the public at the website above. However, the Hydrologic
Engineering Center (HEC) of the U.S. Army Corps of Engineers will not provide user support to non-Corps users. In
addition, public distribution of the model source code is generally discouraged by HEC.
Model Overview/Abstract
The Hydrologic Engineering Center (HEC) of the U.S. Army Corps of Engineers released the River Analysis
System (HEC-RAS) (HEC, 2002a and 2002b) for one-dimensional steady and unsteady flow simulation and
sediment transport/moveable boundary conditions. HEC-RAS is designed to perform one-dimensional hydraulic
calculations for a full network of natural and constructed channels. The model expands on methods from previous
models for steady and unsteady flow simulation, including HEC-2 (HEC, 1991) and UNET (Barkau, 1992; HEC,
1997) and provides additional capabilities for simulation of bridge scour.
Model Features
Hydraulic analysis components
• Steady flow water surface profiles
• Unsteady flow simulation
• Sediment transport/moveable boundary computations
User interface includes
• File management
• Data entry and editing
• Hydraulic analyses
• Tabular and graphical display of input and output data
• Reporting facilities
Model Areas Supported
Watershed None
Receiving Water Low
Ecological None
236
-------
Appendix A: Model Fact Sheets
Air None
Groundwater None
Model Capabilities
Conceptual basis
HEC-RAS is an integrated system composed of a GUI, separate hydraulic analysis components, data storage and
management capabilities, graphics, and reporting facilities. The system ultimately contains three one-dimensional
hydraulic analysis components for (1) steady flow water surface profile computations, (2) unsteady flow simulation,
and (3) moveable boundary sediment transport computations. All three components use a common geometric data
representation and common geometric and hydraulic computation routines. In addition, the system contains several
hydraulic design features that can be invoked once the basic water surface profiles are computed.
Scientific detail
The steady flow component is capable of modeling subcritical, supercritical, and mixed flow regime water surface
profiles. The basic computational procedure is based on the solution of a one-dimensional energy equation. Energy
losses are evaluated by friction and contraction/expansion. The momentum equation is utilized in situations where
the water surface profile is rapidly varied.
The model can perform mixed flow regime calculations in the unsteady flow computations module, based on
methods adapted from UNET (Barkau, 1992; HEC, 1997). For HEC-RAS, the hydraulic calculations for cross
sections and hydraulic structures (e.g., bridges and culverts) that were developed for steady flow simulation were
incorporated in the unsteady flow module.
The sediment transport/movable boundary computations are based on methods reported by the Federal Highway
Administration (1995 and 1996). This module includes one-dimension calculation of sediment transport potential
based on grain size distribution and results of the hydraulic model. The current version of HEC-RAS (Version 3.1)
is limited to short-term analyses of scour at piers and abutments. However, the current version does not include
long-term analysis of aggregation and degradation.
Model Framework
Model supports the following analyses:
• River flow simulation
• Floodway encroachment analysis
• Bridge scour simulation
• Channel hydraulic design
Scale
Spatial Scale
• One-dimensional
Temporal Scale
• User-defined
Assumptions
Key assumptions in model development are definition of channel geometry and flow path for model configuration.
Cross-sectional assumptions include consideration of effective and ineffective flow areas, longitudinal slope,
overflow of floodplains, and considerations to channel roughness. Inherent assumptions of the model based on
theoretical formulations and requirements for model parameterization are reported by HEC (2002a and 2002b).
237
-------
Appendix A: Model Fact Sheets
Additional assumptions are related to inflows required for modeling analyses, which are determined outside of
model development or based on analysis of flow records or results of separate watershed models.
Model Strengths
• Capable of modeling common types of hydraulic control structures with appropriate on and off features.
• Has a GUI with pre- and post-processing capabilities.
Model Limitations
• Cannot simulate water quality processes and relatively difficult to use in conjunction with other water
quality models.
• Cannot simulate groundwater levels.
Application History
Multiple example model applications are reported by HEC (2002c).
Model Evaluation
See available references.
Model Inputs
• Channel geometric data and assumptions for channel roughness
• Flow data and boundary conditions
Users Guide
HEC-RAS, River Analysis System User's Manual (HEC, 2002b).
Available online: http://www.hec.usace.army.mil/software/hec-ras/hecras-document.html
Technical Hardware/Software Requirements
Computer hardware:
The minimum hardware requirements for a Microsoft Windows installation includes
• Intel-based PC or compatible machine with Pentium processor or higher
• 40-MB available hard disk space
• 32-MB of RAM if using Windows 95, 98, ME or 64-MB of RAM using Windows NT, 2000, or XP
• Color video display
Operating system:
• Microsoft Windows XP, 2000, 98, 95, or later editions
• Microsoft Windows NT 4.0
Programming language:
FORTRAN
Runtime estimates:
Available intervals range from 1 minute to 24 hours
238
-------
Appendix A: Model Fact Sheets
Linkages Supported
HEC-DSS
Related Systems
HEC-2, UNET, HEC-HMS, HEC-DSS
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Pre- and post-processors
• Data display tools
• Data preparation tools
References
Barkau, R.L. 1992. One-Dimensional Unsteady Flow Through a Full Network of Open Channels. (Computer
Program). UNET, St. Louis, MO.
Federal Highway Administration (FHWA). 1995. Evaluating Scour at Bridges. HEC No. 18, Publication No.
FHWA-IP-90-017, 3rd Edition. Federal Highway Administration Washington, DC.
Federal Highway Administration (FHWA). 1996. Channel Scour at Bridges in the United States. Publication No.
FHWA-RD-95-185. Federal Highway Administration, Washington, DC.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 1991. HEC-2, Water Surface Profiles:
User's Manual. U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, CA.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 1997. UNET, One-Dimensional Unsteady
Flow Through a Full Network of Open Channels: User's Manual. (Computer program manual). U.S. Army Corps of
Engineers, Hydrologic Engineering Center, Davis, CA.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 2002a. HEC-RAS, River Analysis System
Hydraulic Reference Manual. U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, CA. Available
at http ://www. hec .usace .army. mil/software/hec-ras/hecras-document. html
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 2002b. HEC-RAS, River Analysis System
User's Manual. U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, CA. Available at
http://www.hec.usace.armv.mil/software/hec-ras/hecras-document.html
U.S. Army Corps of Engineers, Hydrologic Engineering Center (HEC). 2002c. HEC-RAS, River Analysis System
Applications Guide. U.S. Army Corps of Engineers, Hydrologic Engineering Center, Davis, CA. Available at
http://www.hec.usace.armv.mil/software/hec-ras/hecras-document.html
239
-------
Appendix A: Model Fact Sheets
HSCTM-2D: Hydrodynamic, Sediment, and Contaminant Transport Model
Contact Information
Model Distribution Coordinator
Center for Exposure Assessment Modeling (CEAM)
U.S. Environmental Protection Agency
960 College Station Road
Athens, GA 30605-2700
(706) 355-8400
ftp://ftp.epa.gov/epa ceam/wwwhtml/ceamhome.htm
Download Information
Availability: Nonproprietary
http://www.epa.gov/ceampubl/swater/hsctm2d/index.htm
Cost: N/A
Model Overview/Abstract
HSCTM-2D is a single model that incorporates internally linked hydrodynamic, sediment transport and contaminant
transport and fate simulation components. HSCTM-2D (Hayter et al., 1998) uses a two-dimensional finite element
formulation and incorporates the RMA2 hydrodynamic model (King, 1990) and the CSTM-H sediment transport
model (Hayter and Mehta, 1986) extended to include sorptive contaminants. Horizontal water column transport
includes advection and shear dispersion. The model can be applied rivers, lakes, estuaries, and coastal waters.
Model Features
• Hydrodynamics
• Sediment transport
• Contaminant transport
Model Areas Supported
Watershed None
Receiving Water High
Ecological None
Air None
Groundwater None
Model Capabilities
Conceptual Basis
HSCTM2D is a finite element modeling system for simulating two-dimensional, vertically integrated, surface water
flow (typically riverine or estuarine hydrodynamics), sediment transport, and contaminant transport. The modeling
system consists of two modules, one for hydrodynamic modeling (HYDRO2D) and the other for sediment and
contaminant transport modeling (CS2D). One example problem is included. The HSCTM2D modeling system may
be used to simulate both short-term (less than 1 year) and long-term scour and/or sedimentation rates and
contaminant transport and fate in vertically well mixed bodies of water.
240
-------
Appendix A: Model Fact Sheets
Scientific Detail
The hydrodynamic module HYDRO2D simulates two-dimensional, depth-averaged flow of surface waters. The
governing equations, eqs. 3.1, 3.7, and3.8, are solved by the Galerkin method of weighted residuals using the finite
element method, as described in Section 4. The depth-averaged velocities in the two horizontal directions and the
flow depth are computed at each node. In addition, continuity can be checked across multiple cross sections. The
effects of bottom, internal, and surface shear stresses and the Coriolis force are simulated in HYDRO2D. Bottom
and surface stresses are due to friction; internal stresses are the results of turbulence. The module can simulate both
steady state and dynamic flows.
The cohesive sediment transport model CS2D is a time varying, two-dimensional finite element model that is
capable of predicting the horizontal and temporal variations in the depth-averaged suspended cohesive sediment
concentrations and bed surface elevations in an estuary, coastal waterway, or river (Hayter and Mehta, 1986). In
addition, it can be used to predict the steady state or unsteady transport of any conservative or nonconservative
constituent if the reaction rates are known. CS2D simulates the advection and dispersion of suspended constituents,
aggregation, and deposition to and erosion from the bed of cohesive sediments. Hayter (1983) describes a series of
experiments that were used to partially validate the model.
Model Framework
• Depth average finite element
Scale
Spatial Scale
• Two-dimensional in horizontal
Temporal Scale
• Dynamic
Assumptions
• Based on accepted formulations of the two-dimensional, depth-averaged hydrodynamic equations and
conservative transport equation
Model Strengths
• Strong capabilities for hydrodynamics and sediment transport in vertically mixed waterbodies
Model Limitations
• Two-dimensional, vertically averaged homogeneous flow
• Will not accurately simulate supercritical flow
• Based on a streamwise bottom slope that is mild and not steeper than (01:10)
Application History
The model has been applied to the Maurice River and Union Lake in New Jersey (Hayter and Gu, 1998).
Model Evaluation
None
Model Inputs
ForHYDRO2D
241
-------
Appendix A: Model Fact Sheets
• Program operation data
• Grid geometry data
• Initial conditions data
• Boundary conditions data
For CS2D:
• Program operation data
• Grid geometry data
• Nodal velocities and salinities
• Initial conditions
• Boundary conditions
• Parameters describing the erosional and depositional behavior of the cohesive sediment as well as the
structure of the bed and contaminant partition coefficients and decay rates.
Users' Guide
Available online http://www.epa.gov/ceampubl/swater/hsctm2d/USERMANU.PDF
Technical Hardware/Software Requirements
Computer hardware:
• PC and UNIX workstations
Operating system:
• Windows or UNIX
Programming language:
• FORTRAN
Runtime estimates:
• Compute intensive with slower run times than similar two-dimensional finite difference models
• Run time is highly dependent on computer hardware, model domain spatial resolution, the period of
prototype conditions simulated, and other options, such as whether the model is simulation-only
hydrodynamic or hydrodynamics and the fate and transport of dissolved and suspended material. Under this
wide range of variability, simulations could require minutes to weeks.
Linkages Supported
None
Related Systems
None
Sensitivity/Uncertainty/Calibration
No sensitivity, uncertainty, or calibration capabilities directly incorporated into model
Model Interface Capabilities
None
242
-------
Appendix A: Model Fact Sheets
References
Hayter, E.J., and AJ. Mehta. 1982. Modeling of Estuarial Fine Sediment Transport for Tracking Pollutant
Movement. UFL/COEL-82/009. Coastal and Oceanographic Engineering Department, University of Florida,
Gainesville, Florida.
Hayter, E. I, and A. J. Mehta. 1986. Modelling cohesive sediment transport in estuarial waters. Appl. Math.
Modelling. 10:294-303.
Hayter, E.J., M. Bergs, R. Gu, S. McCutcheon, S. J. Smith, and H. J. Whiteley. 1998. HSCTM-2D, a finite element
model for depth-averaged hydrodynamics, sediment and contaminant transport. Technical Report. U. S. EPA
Environmental Research Laboratory, Athens, Georgia.
Hayter, E.J., and R. Gu. 1998. Prediction of contaminated sediment transport in the Maurice River-Union Lake,
New Jersey. Paper presented at 5th International Conference on Cohesive Sediment Dynamics, May, 1998, Seoul,
Korea.
243
-------
Appendix A: Model Fact Sheets
HSCTM-2D: Hydrodynamic, Sediment, and Contaminant Transport Model
Contact Information
Model Distribution Coordinator
Center for Exposure Assessment Modeling (CEAM)
U.S. Environmental Protection Agency
960 College Station Road
Athens, GA 30605-2700
(706) 355-8400
ftp://ftp.epa.gov/epa ceam/wwwhtml/ceamhome.htm
Download Information
Availability: Nonproprietary
http://www.epa.gov/ceampubl/swater/hsctm2d/index.htm
Cost: N/A
Model Overview/Abstract
HSCTM-2D is a single model that incorporates internally linked hydrodynamic, sediment transport and contaminant
transport and fate simulation components. HSCTM-2D (Hayter et al., 1998) uses a two-dimensional finite element
formulation and incorporates the RMA2 hydrodynamic model (King, 1990) and the CSTM-H sediment transport
model (Hayter and Mehta, 1986) extended to include sorptive contaminants. Horizontal water column transport
includes advection and shear dispersion. The model can be applied rivers, lakes, estuaries, and coastal waters.
Model Features
• Hydrodynamics
• Sediment transport
• Contaminant transport
Model Areas Supported
Watershed None
Receiving Water High
Ecological None
Air None
Groundwater None
Model Capabilities
Conceptual Basis
HSCTM2D is a finite element modeling system for simulating two-dimensional, vertically integrated, surface water
flow (typically riverine or estuarine hydrodynamics), sediment transport, and contaminant transport. The modeling
system consists of two modules, one for hydrodynamic modeling (HYDRO2D) and the other for sediment and
contaminant transport modeling (CS2D). One example problem is included. The HSCTM2D modeling system may
be used to simulate both short-term (less than 1 year) and long-term scour and/or sedimentation rates and
contaminant transport and fate in vertically well mixed bodies of water.
240
-------
Appendix A: Model Fact Sheets
HSPF: Hydrologic Simulation Program FORTRAN
Contact Information
U.S. Environmental Protection Agency
Office of Water
Office of Science and Technology (430IT)
1200 Pennsylvania Avenue, N.W.
Washington, DC 20460
basins(@,epa.gov
http://www.epa.gov/ost/ftp/basins/svstem/BASINS3/gww.htm
AQUA TERRA
http://www.aquaterra.com/
Download Information
Availability: Nonproprietary
Cost: N/A
Model Overview/Abstract
HSPF is a comprehensive package for simulating watershed hydrology and water quality for a wide range of
conventional and toxic organic pollutants. With its predecessors dating back to the 1960s, HSPF is the culminating
evolution of the Stanford Watershed Model (SWM), watershed-scale Agricultural Runoff Model (ARM), and
Nonpoint Source Loading Model (NPS) into an integrated basin-scale model that combines watershed processes
with in-stream fate and transport in one-dimensional stream channels. HSPF simulates watershed hydrology, land
and soil contaminant runoff, and sediment-chemical interactions. The model can generate time series results of any
of the simulated processes. Overland sediment may be divided into three types of sediment (sand, silt, and clay) for
in-stream fate and transport. Pollutants interact with suspended and bed sediment through soil-water partitioning.
The most recent release is HSPF Version 12, which is distributed as part of the EPA BASINS (Better Assessment
Science Integrating Point and Nonpoint Sources) system.
Model Features
• Detailed watershed simulation model
• Watershed hydrology
• Runoff/sediment/pollutant generation and transport
• One-dimensional stream hydrology and transport
• Pesticide fate and transport simulation
Model Areas Supported
Watershed High
Receiving Water Medium
Ecological None
Air None
Groundwater Low
244
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
In HSPF, a subwatershed is typically conceptualized as a group of various land uses all routed to a representative
stream segment. Several small subwatersheds and representative streams may be networked together to represent a
larger watershed drainage area. Various modules are available and may be readily activated to simulate various
processes, both on land and in-stream.
Scientific Detail
Land processes for pervious and impervious areas are simulated through water budget, sediment generation and
transport, and water quality constituents' generation and transport. Hydrology is modeled as water balance of soil
(or storage) in different layers as described by the SWM methodology. Interception, infiltration, evapotranspiration,
interflow, groundwater loss, and overland flow processes are considered and are generally represented by empirical
equations. Sediment production is based on detachment and/or scour from a soil matrix and transport by overland
flow in pervious areas, whereas solids buildup and washoff is simulated for impervious areas. It includes agricultural
components for land-based nutrient and pesticide processes and a special actions block for simulating management
activities. HSPF also simulates the in-stream fate and transport of a wide variety of pollutants, such as nutrients,
sediments, tracers, dissolved oxygen/biochemical oxygen demand, temperature, bacteria, and user-defined
constituents.
Model Framework
• Hydrologic response unit, subwatershed, and watershed
• Simple one-dimensional stream and well-mixed reservoir/lake model
Scale
Spatial Scale
• One-dimensional, lumped parameters on a land use or subwatershed basis
Temporal Scale
• User-defined timestep, typically hourly
Assumptions
• Land simulation component is a distributed model by land use but ignores the spatial variation within a
land use in a subwatershed.
• For overland flow, model assumes one-directional kinematic-wave flow.
• Model also assumes subwatershed and streams as series of reservoirs while routing flows.
• The receiving waterbody assumes complete mixing along the width and depth.
Model Strengths
• One the few watershed models capable of simulating land processes and receiving water processes
simultaneously.
• Capable of simulating both peak flow and low flows.
• Simulates at a variety of timesteps, including subhourly to 1 minute, hourly or daily.
• Simulates the hydraulics of complex natural and man-made drainage networks
• Includes capabilities to address a variable water table.
• Simulates results for many locations along a reservoir or tributary.
• Includes user-defined model output options by defining the external targets block.
• Can be setup as simple or complex, depending on application, requirements, and data availability.
245
-------
Appendix A: Model Fact Sheets
Model Limitations
• Relies on many empirical relationships to represent physical processes.
• Lumps simulation processes for each land use type at the subwatershed level (i.e., does not consider the
spatial location of one land parcel relative to another in the watershed). The model approaches a distributed
model when smaller subwatersheds are used; however, this may result in increased model complexity and
simulation time.
• Requires extensive calibration.
• Requires a high level of expertise for application.
• Is limited to well-mixed rivers and reservoirs and one-directional flow.
Application History
The modeling concept had its debut in the early 1960s as the Stanford Watershed Model. During the 1970s, water
quality processes were added. A FORTRAN version was developed in the late 1970s, incorporating several related
models and software engineering design and development concepts funded by EPA's research laboratory in Athens,
GA. In the 1980s, pre- and post-processing software, algorithm enhancements, and use of the USGS binary
Watershed Data Management (WDM) system were developed jointly by the USGS and EPA. Since 1980, all model
code changes have been maintained by Aqua Terra Consultants, under contract with EPA and USGS. During the
mid to late 1990s, Tetra Tech, Inc., under contract with EPA developed the BASINS system and NPSM, resulting in
the first Windows-based interface for the HSPF model. The current supported model release is Version 12,
distributed with BASINS 3.0 as the WinHSPF model and interface. HSPF is a proven and tested continuous
simulation watershed model. It is one of the models recommended by the EPA for complex TMDL studies. The
HSPF model has been widely used and its application has been documented throughout its development cycle.
Model Evaluation
HSPF has been widely reviewed and applied throughout its long recent history (Hicks, 1985; Ross et al., 1997; and
Tsihrintzis et al., 1996). One of the largest applications of the model was to the Chesapeake Bay Watershed, as part
of the EPA's Chesapeake Bay Program's management initiative (Donigian, 1990, 1992). Tsihrintzis et al. (1994,
1995) applied HSPF in a GIS shell (using ARC/INFO) to evaluate the impact of agricultural activities, specifically
transport of sediments, nutrients, and pesticides, on streams and groundwater in Southern Florida. An extensive
HSPF bibliography has been compiled to document model development and application and is available online at
http://hspf.com/hspfbib.html.
Model Inputs
• Continuous meteorological time series records including (at a minimum)
o Rainfall
o Potential evapotranspiration
For SNOW simulation, additional required meteorological time series include
o Temperature
o Wind speed
o Solar radiation
o Dewpoint temperature
For additional simulation options, other required meteorological time series may include
o Pan evaporation
o Cloud cover
• Soils data (auxiliary dataset to guide hydrologic calibration), pollutant buildup and washoff, stream
dimensions or rating curves, and point-source loading inputs
• A large number of parameters need to be specified (some default values are available)
Users' Guide
• For model documentation, underlying theory, and parameterization, the HSPF users' manual is a
recommended source (Bicknell et al., 2001).
• A browseable Windows help file version of the manual is available at http://hspf.com/pub/hspf/HSPF.chm.
246
-------
Appendix A: Model Fact Sheets
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• DOS or Windows Operating System
Programming language:
• FORTRAN
Runtime estimates:
• Seconds to minutes or hours, depending on spatial/temporal resolution and computer performance
Linkages Supported
CE-QUAL-W2
Related Systems
WinHSPF, an interface to HSPF, is a key component of Better Assessment Science Integrating point and Nonpoint
Sources (BASINS) Version 3.0. BASINS 3.0 was developed for the U.S. Environmental Protection Agency's Office
of Water to respond to the continued needs of various agencies to perform watershed and water quality assessments,
integrating point and nonpoint sources.
LSPC, another interface to HSPF, is available through the EPA Modeling Toolbox
(http://www.epa.gov/athens/wwqtsc/index.html).
Sensitivity/Uncertainty/Calibration
HSPFParm is a free HSPF parameter database distributed with EPA's BASINS System. The software is installed
independent of the BASINS system. It provides regionalized model parameters for published applications across the
United States. It serves as a good starting point for parameter selection during model setup and calibration.
The Expert System for calibration of HSPF (HSPEXP) is an interactive program that evaluates modeled versus
observed time series using over 35 rules and some 80 conditions (Lumb, 1994). It uses Artificial Intelligence
techniques, incorporating expert advice, based on statistics and evaluation results, to recommend which parameters
should be adjusted.
The Parameter Estimation software package (PEST) is a model calibration aid that can be run in conjunction with
HSPF (Doherty, 2003). The objective function's goal is to minimize the least squares of the difference between
modeled and observed flow by varying model parameters over a range that the user defines. PEST then iterates
through a series of HSPF model runs, changing selected parameters and rerunning the model, until the objective is
satisfied.
Model Interface Capabilities
Using the HSPF Model requires at least two files: the User Control Input File (UCI) for parameters and control
specifications and a WDM file for time series. When run by itself, the UCI text file serves as the interface for the
HSPF model.
247
-------
Appendix A: Model Fact Sheets
The WinHSPF interface, first distributed with BASINS 3.0, provides an interactive interface to HSPF in a Windows
environment. WinHSPF may be used for creating a new HSPF input sequence or for modifying an existing HSPF
input sequence. The program HSPF may be run from within WinHSPF. Input sequences may be modified and saved
under another name, thus creating simulation scenarios.
References
Bicknell, B.R., J.C. Imhoff, J.L. Kittle, Jr., T.H. Jobes, and A. S. Donigian, Jr. 2001. HYDROLOGICAL
SIMULATION PROGRAM - FORTRAN, Version 12, User's Manual. (Computer program manual). AQUA TERRA
Consultants.
Doherty, John, and John M. Johnston. 2003. Methodologies for calibration and predictive analysis of a watershed
model. J. American Water Resources Association. 39(2):251-265.
Donigian, A.S. Jr., and A.S. Patwardhan. (1992). Modeling nutrient loadings from croplands in the Chesapeake Bay
Watershed. In Proceedings of water resources sessions at Water Forum '92, Baltimore, Maryland, August 2-6,
1992, pp. 817-822.
Donigian, A.S., Jr., B.R. Bicknell, L.C. Linker, J. Hannawald, C. Chang, and R. Reynolds. 1990. Chesapeake Bay
Program Watershed Model application to calculate bay nutrient loadings: preliminary Phase I findings and
recommendations. Prepared for the U. S. Chesapeake Bay Program, Annapolis, MD by AQUA TERRA consultants.
Donigian, A.S., Patwardhan, A.S., and R.M. Jacobson. 1996. Watershed Modeling of Pollutant Contributions and
Water Quality in the Le Sueur Basin of Southern Minnesota. In Proceedings of Watershed 96, Baltimore, MD, June
8-12, 1996.
Hicks, C.N. 1985. Continuous Simulation of Surface and Subsurface Flows in Cypress Creek Basin, Florida, Using
Hydrological Simulation Program - FORTRAN (HSPF). Water Resources Research Center, University of Florida,
Gainesville, FL.
Lumb, A.M., McCammon, R.B., and Kittle, J.L., Jr. 1994. Users manual for an expert system (HSPEXP) for
calibration of the Hydrologic Simulation Program—FORTRAN. U.S. Geological Survey Water-Resources
Investigations Report 94-4168. U.S. Geological Survey.
Moore, L.W., C. Y. Chew, R.H. Smith, and S. Sahoo. 1992. Modeling of Best Management Practices on North
Reelfoot Creek, Tennessee. Water Environment Research. 64(3):241-247.
Ross, M.A., P.O. Tara, J.S. Geurink, and M.T. Stewart. 1997. FIPR Hydrologic Model: Users Manual and
Technical Documentation. Prepared for Florida Institute of Phosphate Research, Bartow, FL, and Southwest Florida
Water Management District, Brooksville, FL by University of South Florida, Tampa, FL.
Scheckenberger, R.B., and A.S. Kennedy. 1994. The use of HSPF in subwatershed planning. In Current practices in
modelling the management of stormwater impacts, ed. W. James. Lewis Publishers, Boca Raton, FL. pp. 175-187.
Tsihrintzis, V.A., H.R. Fuentes and R. Gadipudi. 1996. Modeling Prevention Alternatives for Nonpoint Source
Pollution at a Wellfield in Florida. Water Resources Bulletin, Journal of the American Water Resources Association
(AWRA). 32(2):317-331.
Tsihrintzis, V., H. Fuentes, and R. Gadipudi. 1995. Modeling prevention alternatives for nonpoint source pollution
at a wellfield in Florida. Water Resources Bulletin. 32(2):317-331.
Tsihrintzis, V., H. Fuentes, and R. Gadipudi. 1994. Interfacing GIS and water quality models for agricultural areas.
Hydraulic Engineering '94, ed. G. Cotroneo andR. Rumer, ASCE,1, pp 252-256.
248
-------
Appendix A: Model Fact Sheets
KINEROS2: Kinematic Runoff and Erosion Model v2
Contact Information
U.S. Department of Agriculture
Agricultural Research Service
Southwest Watershed Research Center
2000 E.Allen Road
Tucson, Arizona 85719
(520)670-6381
http://www.tucson.ars.ag.gov/kineros/
Download Information
Availability: Nonproprietary
Cost: N/A
Model Overview/Abstract
KINEROS2 is the upgrade from an earlier version called KINEROS (Woolhiser, et al., 1990). The model is an
event-oriented, physically based model that describes the processes of interception, infiltration, surface runoff, and
erosion from small agricultural and urban watersheds (USDA, 2003). The model represents a watershed by an
abstraction into a tree-like network sequence of planes and channels and solves the partial differential equations
describing overland flow, channel flow, erosion, and sediment transport by using finite-difference techniques.
Spatial variation of rainfall, infiltration, runoff, and erosion parameters can be accommodated. The model allows
pipe flow and pond elements as well as infiltrating surfaces and includes a partially paved element to use in urban
area simulation. KINEROS can be used to determine the effects of various artificial features, such as urban
developments, small detention reservoirs, or lined channels, on flood hydrographs and sediment yield. This model is
suitable for small agricultural and disturbed urban watersheds.
Model Features
• Overland flow and channel flow
• Erosion and sediment transport
Model Areas Supported
Watershed High
Receiving Water Low
Ecological None
Air None
Groundwater None
Model Capabilities
Conceptual Basis
KINEROS2 represents a watershed as a cascade of planes and sequence of channels. Multigage rainfall input is
distributed by assigning rain gages to overland flow planes.
249
-------
Appendix A: Model Fact Sheets
Scientific Detail
KINEROS uses one-dimensional kinematic equations to simulate flow over rectangular planes and through
trapezoidal open channels, circular conduits, and small detention ponds. Runoff is routed with an implicit finite
difference solution of the kinematic wave equation. The infiltration algorithm is dynamic, physically based, and
interacting with both rainfall and surface water in transit. The infiltration capacity is simulated as a function of
infiltrated depth and allows estimation of the soil redistribution behavior by using the pore size distribution index.
Further, as an option, the effect of spatial variation in soil hydraulic conductivity can be simulated by assuming a
normal distribution and a user-defined coefficient of variation. Channel transmission losses are also included in the
model. For overland surfaces, erosion consists of two major components—erosion by splash of rainfall on bare soil
and hydraulic erosion (or deposition) due to the interplay between the shearing force of water on the loose soil bed
and the tendency of soil particle to settle under the force of gravity. The splash erosion is approximated as a function
of the rainfall rate square and a reduction factor representing the ponding water depth. The hydraulic erosion is
simulated by applying a modified Engelund and Hansen equation to calculate the shallow flow transport capacity.
Erosion and flow equations are solved numerically at each timestep and for each particle size. A four-point finite-
difference scheme is used.
Model Framework
• Fields/planes and channels
Scale
Spatial Scale
• One-dimensional, fields, subwatershed overland
• One-dimensional, channel network
Temporal Scale
• Variable timestep (normally in minutes)
Assumptions
• For overland flow and channel flow routing, it is assumed that backwater and diffusive wave attenuation is
negligible
• Normal distribution of the soil hydraulic conductivity
Model Strengths
• Contains a physically based infiltration model that allows estimation of the soil redistribution behavior and
considers heterogeneity
• Simulates both splash erosion and hydraulic erosion
• Can represent rainfall spatial variability by assigning different rainfall data to difference elements
Model Limitations
• As an event-based model, does not treat long periods of soil water redistribution, plant growth, and other
interstorm changes
• Does not contain subsurface component
• Simulates sediment only
Application History
USDA-ARS illustrated the application of KINEROS to simulate runoff from a 6.3 km2 semiarid experimental
watershed near Tombstone, Arizona.
250
-------
Appendix A: Model Fact Sheets
Ziegler applied KINEROS2 to modeling erosion on unpaved roads, the results of which are available at
http://www.age.uiuc.edu/s&w/hawaii/Ap30.htm.
Model Evaluation
The model application done by USDA-ARS-SWRC for an experimental watershed near Tombstone, Arizona,
showed that KINEROS2 could reasonably predict the overall trend of the outflow hydrograph.
The work by Ziegler showed that
"KINEROS2 can be parameterized to simulate reasonably total discharge, sediment transport, and sediment
concentration on small-scale road plots, for a range of slopes, during simulated rainfall events. KINEROS2,
however, did not accurately predict time-dependent changes in sediment output and concentration. In
particular, early flush peaks and the temporal decay in sediment output were not predicted, owing to the
inability of KINEROS2 to model removal of a surface sediment layer of finite depth." (Abstract is available
at http://www.age.uiuc.edu/s&w/hawaii/Ap30.htm.)
Model Inputs
• Overland flow element
o Plane geometry (length, width, and slope), Manning's n, Chezy conveyance factor
o Canopy cover, interception depth, average micro topographic relief, average micro topographic
spacing
o Infiltration-related parameters including saturated hydraulic conductivity (Ks), initial degree of
soil saturation, coefficient of variation of Ks, mean capillary drive, porosity, pore size distribution
index, volumetric rock fraction, thickness of soil layers (up to 2 layers)
o Rain splash coefficient, soil cohesion coefficient, and particle class fractions
• Channel element
o Type (simple or compound) and base flow discharge
o Channel geometry (length, width, bed slope, and bank slopes), Manning's n, Chezy conveyance
factor
o Infiltration-related parameters (same as specified in overland flow element)
o Cohesion coefficient of bed material.
• Pond element
o Initial storage volume
o Volume, surface area, and discharge rating table
o Seepage rate
• Rainfall file
• External flow file (optional)
Users' Guide
Available online: http://www.tucson.ars.ag.gov/kineros/
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS
Programming language:
• FORTRAN 77/90
251
-------
Appendix A: Model Fact Sheets
Runtime estimates:
• Seconds to minutes
Linkages Supported
None
Related Systems
BASINS
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Arc View extension to facilitate input data preparation
References
USDA-ARS-Southwest Watershed Research Center. KINEROS2 Documentation.
http://www.tucson.ars.ag.gov/kineros/Docs/Doc.htmlTechnicalManual
Smith, R.E., D.C. Goodrich, and J.N. Quinton. 1995. Dynamic, distributed simulation of watershed erosion: The
KINEROS2 and EUROSEM models. Journal of Soil and Water Conservation. 50(5):517-520.
Woolhiser, D.A., R.E. Smith and D.C. Goodrich. 1990. KINEROS, A kinematic runoff and erosion model:
documentation and user manual. ARS-77. USDA-Agricultural Research Service.
252
-------
Appendix A: Model Fact Sheets
LSPC: Loading Simulation Program in C++
Contact Information
Tim Wool
U.S. Environmental Protection Agency
Office of Research and Development
National Exposure Research Laboratory
Ecosystems Research Division
Watershed and Water Quality Modeling Technical Support Center
960 College Station Road
Athens, GA 30605
(706)355-8312
wool.timfg.epa. gov
http://www.epa.gov/athens/wwqtsc
Download Information
Availability: Nonproprietary
http ://www. epa. gov/athens/wwqtsc/html/lspc. html
Cost: N/A
Model Overview/Abstract
The Loading Simulation Program in C++ (LSPC) is a watershed modeling system that includes streamlined
Hydrologic Simulation Program FORTRAN (HSPF) algorithms for simulating hydrology, sediment, and general
water quality on land as well as a simplified stream transport model. It is a EPA-accepted TMDL modeling
application, developed by Tetra Tech, Inc., partially under contract with EPA. A key advantage of LSPC is that it
has no inherent limitations in terms of modeling size or model operations and has been applied to large, complex
watersheds. In addition, the Microsoft Visual C++ programming architecture allows for seamless integration with
modern-day, widely available software such as Microsoft Access and Excel.
Model Features
• Watershed modeling
• Hydrologic simulation and hydraulic transport
• Overland and in-stream sediment simulation
• Temperature simulation
Model Areas Supported
Watershed High
Receiving Water Medium
Ecological None
Air Medium
Groundwater Medium
253
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
In LSPC, a subwatershed is typically conceptualized as a group of various land uses all routed to a representative
stream segment. Several small subwatersheds and representative streams may be networked together to represent a
larger watershed drainage area. Various modules are available and may be readily activated to simulate various land
and stream processes.
Scientific Detail
Land processes for pervious and impervious areas are simulated through water budget, sediment generation and
transport, and water quality constituents' generation and transport. Hydrology is modeled as water balance of soil
(or storage) in different layers as described by the Stanford Watershed Model (SWM) methodology. Interception,
infiltration, evapotranspiration, interflow, groundwater loss, and overland flow processes are considered and are
generally represented by empirical equations. Sediment production is based on detachment and/or scour from a soil
matrix and transport by overland flow in pervious areas, whereas solids buildup and washoff is simulated for
impervious areas. For water quality, buildup and washoff of a quality constituent on a land surface, baseflow-
associated pollutant concentrations, sediment-associated pollutant load, and soil temperature and heat transfer to
water are available.
For in-stream simulation, the model simulates radiation and heat transfer, conservative substance routing, suspended
solids routing and settling, and general first-order pollutant loss, which is applicable for simulating a wide range of
conservative substances, such as fecal coliform bacteria, total nitrogen and phosphorus, and total metals. The
hydraulic compartment, which determines the advection terms for all of the other components, is based on a time-
interval budget of water volume between inflow from the above reach, user-specified outflows, and discharge to the
next downstream reach.
Model Framework
• Watershed hydrologic, sediment, and water quality
• Time series climate-driven, overland, subsurface, and in-stream simulation
Scale
Spatial Scale
• Lumped parameters at a land use subwatershed scale
• One-dimensional in-stream fate and transport
• Capable of simulating many subwatersheds (100+) over large drainage areas (8-digit HUCs)
Temporal Scale
• User-defined timestep, typically hourly
Assumptions
• Land simulation component is a distributed model by land use but ignores the spatial variation within a
land use in a subwatershed.
• For overland flow, model assumes one-directional kinematic-wave flow.
• Model also assumes subwatershed and streams as series of reservoirs while routing flows.
• Model assumes complete mixing along the width and depth of the receiving waterbody.
Model Strengths
LSPC is one the few watershed models that is capable of simulating land processes and receiving water processes
simultaneously. It is capable of simulating both peak flow and low flows and a variety of timesteps, including sub-
hourly to one minute, hourly or daily. The model simulates the hydraulics of complex natural and man-made
254
-------
Appendix A: Model Fact Sheets
drainage networks and includes capabilities to address variable water table. A key strength of LSPC model output is
that the model automatically aggregates results and manages the output at a subwatershed or reach-segment level.
The model can be setup as simple or complex, depending on the application requirement and data availability. The
design of the modeling system and supporting databases is particularly well suited for efficient application to large,
complex watersheds. Data management tools support the evaluation of loading and management within multiple
watersheds simultaneously.
Model Limitations
The model relies on many empirical relations to represent physical processes. For land simulation, processes are
lumped for each land use type at the subwatershed level; therefore, the model does not consider the spatial location
of one land parcel relative to another in the watershed. The model approaches a distributed model when smaller
subwatersheds are used; however, this may result in increased model size and simulation time. For in-stream
simulation, it is limited to well mixed rivers and reservoirs and one-directional flow. It requires extensive
calibration. It generally requires a high level of expertise for application.
Application History
LSPC is the key watershed modeling component of the TMDL Toolbox. TMDL's model applications have been
successfully developed using LSPC in Alabama, Mississippi, South Carolina, Georgia, California, Kentucky,
Tennessee, West Virginia, Virginia, Maryland, Arizona, Ohio, Montana, Puerto Rico, and U.S. Virgin Islands.
Several recent watershed TMDLs have been done using LSPC. Two are listed as follows:
• Published Tennessee TMDL documents may be accessed at the following web address:
http://www.state.tn.us/environment/wpc/tmdl/approved.php
• Published Alabama TMDL documents may be accessed at the following web address:
http ://www. epa. gov/region4/water/tmdl/alabama/index. htm
Model Evaluation
HSPF, on which LSPC is based, has been widely reviewed and applied throughout its history (Hicks, 1985; Ross, et
al., 1997; and Tsihrintzis, et al., 1996). Tsihrintzis, et al., (1994, 1995) applied HSPF in a CIS shell (using
ARC/INFO) to evaluate the impact of agricultural activities, specifically transport of sediments, nutrients, and
pesticides, on streams and groundwater in Southern Florida. The underlying HSPF algorithms used in the LSPC
model have been widely used and the applications have been documented over more than 20 years.
Model Inputs
• Continuous meteorological time series records including (at a minimum)
o Rainfall
o Potential evapotranspiration
For SNOW simulation, additional required meteorological time series include
o Temperature
o Wind speed
o Solar radiation
o Dewpoint temperature
• Soils data (auxiliary dataset to guide hydrologic calibration), pollutant buildup and washoff, stream
dimensions or rating curves, and point-source loading inputs
• A large number of parameters need to be specified (some default values are available)
Users' Guide
• An LSPC Users' Manual (Tetra Tech, 2002) is available through EPA's Watershed and Water Quality
Modeling Technical Support Center.
• For model documentation, underlying theory, and parameterization, the HSPF users' manual is a
recommended source (Bicknell, et al., 2001).
255
-------
Appendix A: Model Fact Sheets
• A browseable Windows help file version of the manual is available at http://hspf.com/pub/hspf/HSPF.chm.
Technical Hardware/Software Requirements
Computer hardware:
• PC (Pentium III or higher recommended, but not required)
Operating system:
• Windows 98 or later
Programming language:
• C++
Runtime estimates:
• Seconds to minutes or hours, depending on spatial/temporal resolution
Linkages Supported
EFDC, WASP, CE-QUAL-W2, EPD-Rivl
Related Systems
HSPF, WinHSPF
Sensitivity/Uncertainty/Calibration
LSPC includes a data analysis component that may be used to quickly compare model output against observed data
in time series form, as monthly summaries, or on a one-to-one graph. LSPC model output is especially tailored for
spreadsheet use; consequently, many users prefer to develop independent spreadsheet analysis, summarization,
calibration, and plotting applications, which are readily linked to LSPC model output.
HSPFParm is a free HSPF parameter database distributed with EPA's BASINS System. The software is installed
independent of the BASINS system. It provides regionalized model parameters for published applications across the
United States. It serves as a good starting point for parameter selection during model setup and calibration.
The Parameter Estimation software package (PEST) is a model calibration aid that can be run in conjunction with
HSPF (Doherty, 2003). Although PEST is not currently linked with LSPC, it is a generalized calibration system that
can potentially be linked to any model. Since parameter values and definitions in HSPF and LSPC are
interchangeable, results from a small-scaled HSPF PEST run can be used to calibrate LSPC. The objective
function's goal is to minimize the least squares of the difference between modeled and observed flow by varying
model parameters over a range that the user defines. PEST then iterates through a series of model runs until the
objective is satisfied.
Model Interface Capabilities
When launched from within the EPA Watershed Characterization System (WCS), LSPC has an extension for
automatically setting up the model using GIS information. The LSPC model interface has two components: (1) a
stand alone GIS component, and (2) a model parameter management component. The LSPC Map Objects GIS
interface, which is compatible with ArcView shapefiles, acts as the control center for managing and launching
watershed model scenarios. The stand alone interface communicates with shapefiles and the underlying Access
database. Therefore, once a watershed system is created, it is easily transferable to users who may not have ArcView
or MS Access. The LSPC model parameter management component can be used either independently on previously
saved model setup files or in conjunction with the GIS component to setup model scenarios on the fly.
256
-------
Appendix A: Model Fact Sheets
LSPC also includes a suite of additional tools, including data management tools for editing watershed data, data
inventory tools for reporting and summarizing model inputs and outputs, and data analysis tools for both visualizing
and summarizing model inputs or outputs and observed data. The model includes a TMDL calculation that allows
the user to specify source-specific allocations and generate corresponding model results for TMDL analysis.
References
Bicknell, B.R., J.C. Imhoff, J.L. Kittle, Jr., T.H. Jobes, and A. S. Donigian, Jr. 2001. HYDROLOGICAL
SIMULATION PROGRAM - FORTRAN, Version 12. (Computer program manual). AQUA TERRA Consultants.
Donigian, A.S. Jr., and A.S. Patwardhan. 1992. Modeling nutrient loadings from croplands in the Chesapeake Bay
Watershed. In Proceedings of water resources sessions at Water Forum '92, Baltimore, Maryland, August 2-6,
1992, pp. 817-822.
Donigian, A.S., Jr., B.R. Bicknell, L.C. Linker, J. Hannawald, C. Chang, and R. Reynolds. 1990. Chesapeake Bay
Program Watershed Model application to calculate bay nutrient loadings: preliminary Phase I findings and
recommendations. Prepared for the U. S. Chesapeake Bay Program, Annapolis, MD, by AQUA TERRA consultants.
Donigian, A.S., Patwardhan, A.S., and R.M. Jacobson. 1996. Watershed Modeling of Pollutant Contributions and
Water Quality in the Le Sueur Basin of Southern Minnesota. In Proceedings of Watershed 96, Baltimore, MD, June
8-12, 1996.
Hicks, C.N. 1985. Continuous Simulation of Surface and Subsurface Flows in Cypress Creek Basin, Florida, Using
Hydrological Simulation Program - FORTRAN (HSPF). Water Resources Research Center, University of Florida,
Gainesville, FL.
Lumb, A.M., McCammon, R.B., and Kittle, J.L., Jr. 1994. Users manual for an expert system (HSPEXP) for
calibration of the Hydrologic Simulation Program—FORTRA. U.S. Geological Survey Water-Resources
Investigations Report 94-4168. U.S. Geological Survey.
Moore, L.W., C. Y. Chew, R.H. Smith, and S. Sahoo. 1992. Modeling of Best Management Practices on North
Reelfoot Creek, Tennessee. Water Environment Research. 64(3):241-247.
Ross, M.A., P.O. Tara, J.S. Geurink, and M.T. Stewart. 1997. FIPR Hydrologic Model: Users Manual and
Technical Documentation. Prepared for Florida Institute of Phosphate Research, Bartow, FL, and Southwest Florida
Water Management District, Brooksville, FL, by University of South Florida, Tampa, FL
Scheckenberger, R.B., and A.S. Kennedy. 1994. The use of HSPF in subwatershed planning. In Current practices in
modelling the management of stormwater impacts, ed. W. James. Lewis Publishers, Boca Raton, FL. pp. 175-187.
Tsihrintzis, V.A., H.R. Fuentes and R. Gadipudi. 1996. Modeling Prevention Alternatives for Nonpoint Source
Pollution at a Wellfield in Florida. Water Resources Bulletin, Journal of the American Water Resources Association
(AWRA). 32(2):317-331.
Tsihrintzis, V., H. Fuentes, and R. Gadipudi. 1995. Modeling prevention alternatives for nonpoint source pollution
at a wellfield in Florida. Water Resources Bulletin. 32(2):317-331.
Tsihrintzis, V., H. Fuentes, and R. Gadipudi. 1994. Interfacing GIS and water quality models for agricultural areas.
Hydraulic Engineering '94, ed. G. Cotroneo and R. Rumer, ASCE,1, pp 252-256.
Tetra Tech, Inc., U.S. Environmental Protection Agency (USEPA). 2002. The Loading Simulation Program in C++
(LSPC) Watershed Modeling System—Users' Manual.
257
-------
Appendix A: Model Fact Sheets
Mercury Loading Model: Watershed Characterization System—Mercury
Loading Model
Contact Information
James Greenfield
U.S. Environmental Protection Agency, Region 4
6IForsyth Street, S.W.
Atlanta, GA 30303-8960
(404) 562-9238
greenfield.j im@epa. gov
http ://wcs .tetratech-ffx. com/Svstem/index. htm
Download Information
Availability: Nonproprietary
http ://wcs .tetratech-ffx. com/Svstem/index. htm
Cost: N/A
Model Overview/Abstract
Watershed-scale Mercury Loading Model is an Arc View 3.x grid-based modeling extension of the Watershed
Characterization System (WCS). The complexity of this loading function model falls between that of a detailed
simulation model, which attempts a mechanistic, time-dependent representation of pollutant load generation and
transport, and simple export coefficient models, which do not represent temporal variability. The WCS provides
simulation of precipitation-driven runoff and sediment delivery at a grid-based landscape. Solids load from runoff is
used to estimate pollutant delivery to the receiving waterbody from the watershed. This estimate is based on
mercury concentrations in wet and dry deposition and processed by soils in the watershed and ultimately delivered
to the receiving waterbody by runoff, erosion, and direct deposition. The main driving force for the WCS mercury
model is the input of the appropriate wet and dry deposition rates or maps for mercury as well as the climate
condition of a watershed.
Model Features
• Arc View GIS grid-based watershed mercury loading calculation
• Estimates total annual average mercury load from different sources of a watershed
• Model results are summarized in maps and tables that can be incorporated into a WCS automatic report
Model Areas Supported
Watershed Medium
Receiving Water None
Ecological None
Air Low
Groundwater None
258
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
The WCS Mercury Loading model is based on a soil-mercury mass balance model, IBM v2.05, developed by EPA's
Office of Health and Environmental Assessment and Environmental Research Lab, Athens, Georgia. The soil-
mercury mass balance model was expanded to account for the spatial heterogeneity of a watershed landscape using
Arc View GIS and grid-processing technology.
Scientific Detail
The WCS Mercury Loading model uses three major algorithms to calculate watershed mercury load:
1. Erosion and sediment transport algorithm for the calculation of mercury load from sediment. The
algorithm implemented in the model is the same as that used in the sediment budget model described in the
previous section. USLE is used to calculate the erosion potential, and an area-based sediment delivery ratio
equation is used as the default formula to calculate the sediment delivery.
2. Hydrology algorithm for the calculation of mercury load from surface runoff. Monthly rainfall is used to
calculate the average rainfall amount for rainfall events. The number of rain days is first compiled from selected
weather stations in EPA Region 4. Rainfall/event is then derived by dividing monthly rainfall by monthly
number of rain days. Rainfall/event is used to calculate the runoff using the curve number method developed by
USDA-NRCS. The hydrology algorithm also includes the computation of other components of the water
balance processes, including evapotranspiration and infiltration.
3. Chemistry algorithm for the calculation of mercury concentration in soil (particle phase and water
phase). U.S. national mercury deposition maps were used for atmospheric mercury input. Atmospheric input
deposits mercury directly on water surfaces, impervious land surfaces, and pervious land surfaces. For the
pervious land surface, mercury concentration in soils is calculated using the steps outlined in Mercury Study
Report to Congress (USEPA, 1997). The equation for the calculation of the soil mercury concentration can be
written as:
Cs = (C0 * e k"kStotal* T) + (D + W) * (1- e "kStotal* T) * 105 / ( zd * BD * ks total)
where Cs = soil mercury concentration at equilibrium (ng/g or ppb); C0 = initial (pre -industrial) soil mercury
concentration (ng/g); k = soil mercury mineralization rate (yr~ :); ks totai = soil mercury loss rate (yr"1) from leaching,
runoff, and erosion; zd = watershed soil incorporation depth or mixing depth (cm); BD = soil bulk density
(g/cm3); T = deposition period (yr"1); D = annual dry deposition rate (g m"2 yr"1); and W = annual wet deposition
rate (g m"2 yr"1). Based on atmospheric deposition rate and soil mercury concentration, mercury load from
sediment, runoff, and impervious surface, direct deposition on water can be calculated for the selected
subwatersheds in a WCS modeling project. The total watershed mercury load can be written as
total ~~ L, erosion ~"~ L, runoff ~"~ L, impervious runoff ~"~ L, water ~"~ L, point source
where L stands for mercury load.
The final results from the WCS mercury loading model are mercury loading tables and maps. All the loading tables
and maps, as well as results of loading comparison between the different subwatersheds, can be exported to a
Microsoft Word document using the automated reporting function.
Model Framework
• WCS Mercury Loading Model calculates sediment load, runoff, atmospheric deposition, and mercury
concentration in watershed soils using grid-based data (e.g., land use and elevation). The model
summarizes mercury load from various sources in each subwatershed of the study area.
259
-------
Appendix A: Model Fact Sheets
Scale
Spatial Scale
• One-dimensional, subwatershed, grid
Temporal Scale
• Annual and long-term average
Assumptions
• Soil concentrations within a deposition grid are uniform within the grid and can be estimated by the
following key parameters: dry and wet contaminant deposition rates, a set of soil transformation rates, a
soil bulk density, and a soil mixing depth.
• The partition of mercury components among soil water and soil particle phases can be described by
partition coefficient.
• The mercury gas phase in soil is insignificant in the total mercury mass balance, and the mercury reduction
loss is equal to the volatilization loss.
• The total runoff of an area is a simple sum of the runoff of each grid, and there is no mercury loss when it is
transported from a source cell to the watershed outlet through the runoff.
• The sediment transport can be described using an area-based delivery ratio formula.
Model Strengths
• Arc View 3.x based, calculates soil mercury concentrations and loading potential grid-by-grid
• Model is easy to setup and use
• Generates mercury load map showing contributions of different sources
• Generates maps and results table in an automatic report
Model Limitations
• Calculates only total mercury annual average load
• Lacks detailed algorithm for wetlands and forests
• Lacks mercury stream transport algorithm
• Lacks mercury input from soil parent material and mercury load calculation through shallow groundwater
Application History
The WCS Mercury Loading model was used to develop mercury TMDLs for the Middle and Lower Savannah River
basins and many other areas in EPA Region 4 (http://www.epa.gov/region4/water/tmdl/general.htm).
Model Evaluation
Model has been tested and evaluated against the original IEM-2M spreadsheet watershed model. The model was
also published in the Water Environment Federation Specialty Conference in 2002. The model has been applied to
many TMDLs in EPA Region 4 since 2000.
Model Inputs
• Mercury atmospheric deposition rate or map
• Monthly average precipitation and temperature
• Mercury point sources
• Land use/cover map
• Soil map
• Digital elevation model
• Stream reach files
260
-------
Appendix A: Model Fact Sheets
Users' Guide
Available online: http://wcs.tetratech-ffx.com/Documentation/index.htm
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• Microsoft Windows, requires Arc View 3.x and Arc View Spatial Analyst extension
Programming language:
• Arc View 3.x and Avenue script
Runtime estimates:
• Minutes to less than 1 hour
Linkages Supported
Output can be used with WASP (requires some manual processing)
Related Systems
Watershed Characterization System or WCS (http://wcs.tetratech-ffx.com')
BASINS (http://www.epa.gov/ostwater/BASrNS/')
Sensitivity/Uncertainty/Calibration
Not available. For a calibration example, see "Total Maximum Daily Load (TMDL) for Total Mercury in Fish
Tissue Residue in the Middle & Lower Savannah River Watershed" (USEPA Region 4, 2001).
Model Interface Capabilities
• Arc View 3.x as the user interface
• WCS system provides manual and auto delineation tools for watershed delineation
• WCS web site has core input datasets available for EPA Region 4, the data can be downloaded by USGS 8-
digit hydrologic units in Region 4
References
Greenfield J., T. Dai, and H. Manguerra. 2002. Watershed modeling extensions of the watershed characterization
system. In Proceedings of the Water Environment Federation Specialty Conference, Watershed 2002, Ft.
Lauderdale, Florida, February, 2002.
Dai, T., andH. Manguerra. 2000. User's Guide for WCS Mercury Tool. TetraTech, Inc., Fairfax, Virginia. Available at
http://wcs.tetratech-ffx.com/Documentation/index.htm
U.S. Environmental Protection Agency Region 4. 2001. Total Maximum Daily Load (TMDL) for Total Mercury in
Fish Tissue Residue in the Middle & Lower Savannah River Watershed. U.S. Environmental Protection Agency
Region 4. Available at http://www.epa.gov/owow/tmdl/examples/mercury.html
261
-------
Appendix A: Model Fact Sheets
U.S. Environmental Protection Agency (USEPA). 1997. Mercury study report to congress. EPA-452/R-97-003.
Office of Air Quality, Planning and Standards. Office of Research and Development. Washington, DC. Available at
http://www.epa.gov/oar/mercury.html
262
-------
Appendix A: Model Fact Sheets
MIKE 11
Contact Information
Peter De Golian
Danish Hyrdrologic Institute, Inc. (DHI)
311 S. Brevard Avenue
Tampa, FL 33606
(813)254-9427
pcg(@,dhigroup.com
www.dhigroup.com
Download Information
Availability: Proprietary
Demo versions available online only.
Cost (if applicable): $3000
Model Overview/Abstract
MIKE 11 is a general river modeling system developed by DHI. MIKE 11 is the most advanced and comprehensive
of its type today, and it is presently used by more than 1500 users around the world. MIKE 11 has become industry
standard in many countries, including Australia, New Zealand, and Bangladesh and in many European countries.
MIKE 11 contains modules for run-off simulations, hydrodynamics, flood forecasting, transport and dilution of
dissolved substances, sediment transport, and river morphology as well as various water quality processes. MIKE 11
has an interface to GIS allowing for preparation of model input and presentation of model output in a GIS
environment.
Model Features
A modular engineering tool for modeling conditions in rivers, lakes and reservoirs, irrigation canals, and other
inland water systems. It is designed for
• Flood risk analysis and mapping
• Design of flood alleviation systems
• Real-time flood forecasting
• Real-time water quality forecasting and pollutant tracking
• Hydraulic analysis and design of structures, including bridges
• Drainage and irrigation studies
• Optimization of river and reservoir operations
• Dam break analysis
• Water quality issues
• Integrated groundwater and surface water analysis
Model Areas Supported
Watershed Low
Receiving Water High
Ecological None
Air None
Groundwater Low
263
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
The hydrodynamic module (HD), which is the core of MIKE 11, contains an implicit, finite difference computation
of unsteady flows in rivers and estuaries. The formulations can be applied to branched and looped networks and
quasi two-dimensional flow simulation on flood plains. The computational scheme is applicable to vertically
homogeneous flow conditions ranging from steep river flows to tidally influenced estuaries. Both subcritical and
supercritical flow can be described by means of a numerical scheme, which adapts according to the local flow
conditions.
The complete nonlinear equations of open channel flow (Saint-Venant) can be solved numerically between all grid
points at specified time intervals for given boundary conditions. In addition to this fully dynamic description, a
choice of other flow descriptions is available:
• High-order, fully dynamic
• Diffusive wave
• Kinematic wave
• Quasi-steady state
• Kinematic routing (Muskingum, Muskingum-Cunge)
Within the standard HD module, advanced computational formulations enable flow over a variety of structures to be
simulated:
• Broad-crested weirs
• Culverts
• Bridges
• Pumps
• Regulating structures
• Control structures
• Dam-break structures
• User-defined structures
• Tabulated structures
The HD module is available in a number of different sizes, which can be upgraded at any time.
Furthermore, a number of add-on modules are available, which ensures that the system can be tailored to suit the
exact requirements.
Scientific Detail
The MIKE 11 hydrodynamic module (HD) uses an implicit, finite difference scheme for the computation of
unsteady flows in rivers and estuaries. The module can describe subcritical as well as supercritical flow conditions
through a numerical scheme that adapts according to the local flow conditions (in time and space). Advanced
computational modules are included for description of flow over hydraulic structures, including possibilities to
describe structure operation. The formulations can be applied to looped networks and quasi two-dimensional flow
simulation on flood plains. The computational scheme is applicable for vertically homogeneous flow conditions
extending from steep river flows to tidal influenced estuaries. The system has been used in numerous engineering
studies around the world.
Model Framework
MIKE 11 consists of a hydrodynamic core module and a number of add-on modules, each simulating certain
phenomena in a river system.
264
-------
Appendix A: Model Fact Sheets
The modular structure offers great flexibility, because
• Each module can be operated separately
• Data transfer between modules is automatic
• Complex physical processes can be coupled (e.g., river morphology, sediment re-suspension and water
quality)
• Updating or expansion of existing installations or models with new modules is simple
Scale
Spatial Scale
• One-dimensional model
Temporal Scale
• User-defined, variable timestep
Assumptions
The hydrodynamic module solves the complete nonlinear St. Venant equations for open-channel flow. It also
includes a quasi-steady state solver for rapid calculation of long-term simulations. The model automatically adapts
to subcritical and supercritical flow and can simulate a wide variety of flow structures, including weirs, culverts,
bridges and user-defined structures. It also allows the flexible operation of flood control and reservoir structures
(e.g., gates and pumps).
Model Strengths
Some of the strengths for the MIKE 11 model include
• MIKE 11 has an interface to GIS allowing for preparation of model input and presentation of model output
in a GIS environment
• Easily links up to other MIKE models
Model Limitations
Some of the limitations for the MIKE 11 model include
• Need to purchase multiple modules to take full advantage of the system
• Significant data needed to setup
Application History
Everglades Agricultural Area (CERP). The project was highly demanding on local knowledge and required
countless educated assumptions to be made with respect to agricultural and reservoir interactions.
Others: http://www.dhisoftware.com/MIKEl 1/News/MIKE 11 Applications.htm
Model Evaluation
http://www.dhisoftware.com/MIKEl 1/News/MIKE 11 Papers.htm
Model Inputs
• Graphical data input/editing
• Simultaneous input/editing of various data types
• Copy and paste facility for direct import (export) from, e.g., spreadsheet programs
• Fully integrated tabular and graphical windows
• Importing of river network and topography data from ASCII text files
265
-------
Appendix A: Model Fact Sheets
• User-defined layout of all graphical views (colors, font settings, lines, marker types, etc.)
Users' Guide
Available online: http://www.dhisoftware.com
Technical Hardware/Software Requirements
Computer hardware:
• 128 Mb DRAM
• 1 Gb hard drive space
• Pentium 200 Mhz minimum
Operating system:
• MS Windows 98/2000/NT/XP
Programming language:
• None
Runtime estimates:
• Depends on CPU
Linkages Supported
Other MIKE models by DHL
Related Systems
DHI series of MIKE models, built on the MIKE Zero interface.
Sensitivity/Uncertainty/Calibration
Built in Auto Calibration tool (http://www.dhisoftware.com/Generic tools/index.htni)
Model Interface Capabilities
MIKE 11 is operated through an efficient Windows-based interactive Graphical User Interface including both
graphical and tabular editing of data. The use of generic editors makes learning easy and efficient. Hence, the time
from learning to production is short.
The advanced graphical facilities enable visual data checking and presentation of the information stored in data files
and time series databases. The same graphical environment is used for data control, analysis, and presentation of
results. The graphical presentation includes river network plan plots, cross-sectional plots, pre-selection of
longitudinal profiles, time series plots, comparison of measured/simulated and simulated/simulated time series,
animation of flows and water levels on both plans and profiles, control of plotting parameters, etc.
References
DHI Software. Mike 11 References. http://www.dhisoftware.com/MIKEl 1/News/MIKE 11 Papers.htm
266
-------
Appendix A: Model Fact Sheets
MIKE 21
Contact Information
Peter De Golian
Danish Hyrdrologic Institute, Inc. (DHI)
311 S. Brevard Avenue
Tampa, FL 33606
(813)254-9427
pcg(@,dhigroup.com
www.dhigroup.com
Download Information
Availability: Proprietary
Demo versions available online only.
Cost (if applicable): $3000
Model Overview/Abstract
MIKE 21 is a professional engineering software package containing a comprehensive modeling system for two-
dimensional free-surface flows. MIKE 21 is applicable to the simulation of hydraulic and related phenomena in
lakes, estuaries, bays, coastal areas, and seas where stratification can be neglected.
MIKE 21 provides the design engineer with a unique and flexible modeling environment, using techniques that have
set the standard in 2D modeling.
It is provided with a modern user-friendly interface facilitating the application of the system. A wide range of
support software for use in data preparation, analysis of simulation results, and graphical presentation is included.
MIKE 21 is the result of more than 20 years of continuous development and is tuned through the experience gained
from thousands of applications worldwide. DHI continues to use MIKE 21 in its own studies, thus giving a valuable
symbiosis between development and application.
Model Features
MIKE 21 HD simulates water level and flow variations in river channels; on river floodplains (both urban and
rural); and in lakes, estuaries and coastal areas in response to various forcing functions. The water levels and flows
are resolved using a rectangular nested grid or tin or finite volume grid covering the area of interest, using the river
channel and floodplain topography, bed resistance coefficients, and hydrographic boundary conditions.
MIKE 21 HD solves the vertically integrated, fully dynamic equations of continuity and conservation of momentum
in two horizontal directions using implicit finite difference methods. The following effects are included in the
equations:
• Convective and cross momentum
• Momentum dispersion
• Floodplain flooding and drying
• Evaporation
• Wind shear stress
• Supercritical flow
267
-------
Appendix A: Model Fact Sheets
• Hydraulic structures
MIKE 21 HD forms the basis for calculations in additional MIKE 21 modules, describing advection dispersion,
water quality, and sediment transport.
Model Areas Supported
Watershed Low
Receiving Water High
Ecological None
Air None
Groundwater Low
Model Capabilities
MIKE 21 HD simulates water level and flow variations in river channels, on river floodplains (both urban and rural),
and in lakes, estuaries, and coastal areas in response to various forcing functions. The water levels and flows are
resolved using a rectangular nested grid or tin or finite volume grid covering the area of interest, using the river
channel and floodplain topography, bed resistance coefficients, and hydrographic boundary conditions.
MIKE 21 HD solves the vertically integrated, fully dynamic equations of continuity and conservation of momentum
in two horizontal directions, using implicit finite difference methods. The following effects are included in the
equations:
• Convective and cross momentum
• Momentum dispersion
• Floodplain flooding and drying
• Evaporation
• Wind shear stress
• Supercritical flow
• Hydraulic structures
MIKE 21 HD forms the basis for calculations in additional MIKE 21 modules describing advection dispersion,
water quality, and sediment transport.
Scale
Spatial Scale
In MIKE 21 HD, the floodplain and channel topography are described by a rectangular grid. The grid size is
specified by the user and should be based on the level of topographical detail to be represented. MIKE 21 HD can be
used to simulate two-dimensional flooding over a wide, flat, spatial range from small-scale urban situations, with
channels only meters wide, to broad-scale river floodplains of several kilometers' width.
Temporal Scale
The temporal scale of MIKE 21 HD is characterized by its flexibility and is based on the boundary time series
defined by the user. MIKE 21 HD has the flexibility to run short-scale, event-based time series or time series
covering months or years of flow simulation. Boundary time series are entered into a database by the user, who also
defines the length and timestep of the series.
Assumptions
MIKE 21 HD solves the vertically integrated, fully dynamic equations of continuity and conservation of momentum
in two horizontal directions. Model parameters for MIKE 21 are
268
-------
Appendix A: Model Fact Sheets
• Bed resistance
• Momentum dispersion (eddy) coefficients
• Wind friction factor (optional)
Model Strengths
Some of the strengths for the MIKE 21 model include
• MIKE 21 has an interface to GIS allowing for preparation of model input and presentation of model output
in a GIS environment.
• Easily links up to other MIKE models.
Model Limitations
Some of the limitations for the MIKE 21 model include
• Need to purchase multiple modules to take full advantage of the system.
• Significant data needed to setup.
Application History
MIKE 21 HD has been utilized in a wide range of flooding-related studies worldwide including:
• Flood studies
• Floodplain management studies
• Flood protection studies, especially where a detailed description of the impact on flood flow patterns is
required
• Dam break studies
• Urban flooding
Some of today's flood modeling systems offers a combination of a one-dimensional river hydraulics model and a
two-dimensional surface water model. Combining such models provides a highly efficient modeling framework as
benefits of both models can be applied where most appropriate. The modeler can apply detailed modeling and obtain
accuracy where needed without sacrificing computational or model construction time.
MIKE FLOOD is a dynamically linked one-dimensional and two-dimensional flood-modeling package. The new
tool is assembled from components taken from two of the most widely applied flood modeling packages - MIKE 11
and MIKE 21. MIKE FLOOD is FEMA-approved.
MIKE FLOOD provides a seamless coupling between river and floodplain, between the sea and inland
waterways^ays/lagoons. Long river reaches may be dynamically linked to the adjacent floodplain, enabling
accurate representation of complex lateral flood plain flooding and drainage. MIKE FLOOD offers extensive
capabilities for accurately representing dams, levees, bridges, road crossings, culverts, and operational structures.
GIS is used to process input data and for flood plain mapping.
Model Evaluation
A search of the Internet and recent publications did not yield any information for the MIKE 21 model evaluation.
Model Inputs
• Basic model parameters
o Model grid size and extent
o Timestep and length of simulation
o Type of output required and its frequency
269
-------
Appendix A: Model Fact Sheets
• River channel and floodplain topography
• Calibration factors
o Bed resistance coefficients
o Momentum dispersion coefficients
o Wind friction factor (optional)
• Boundary conditions
o Water levels or flow magnitudes
o Flow direction
MIKE 21 HD utilizes the powerful and flexible MIKE Zero graphical interface for the input and pre- and post-
processing of data.
Users' Guide
MIKE 21 HD is supported by thorough online-help system and user manual and technical reference documentation.
In addition, DHI offers a comprehensive system of technical support through its dedicated Software Support Center.
24-hour assistance from DHI's technical staff can be obtained through the Software Support Center via telephone
hotline, fax, or the Internet. As a part of License Service Agreements, DHI software users are updated regularly with
software developments via newsletters and Internet broadcasts.
Technical Hardware/Software Requirements
Computer hardware:
• 128 Mb DRAM
• 1 Gb hard drive space
• Pentium 200 Mhz minimum
Operating system:
• MS Windows 98/2000/NT/XP
Programming language:
• None
Runtime estimates:
• Depends on CPU
Linkages Supported
Other MIKE models by DHI.
Related Systems
DHI series of MIKE models, built on the MIKE Zero interface.
Sensitivity/Uncertainty/Calibration
Parameter Estimation/Model Calibration
MIKE 21 HD has three calibration parameters, namely bed resistance factor, momentum dispersion (eddy)
coefficient, and wind friction factor (optional). Calibration of the model can be achieved easily by adjustment of
these factors. In practice, the calibration of a model depends more on the accuracy of the available data (e.g.,
topography and boundary time series definition) than the model parameters. Model calibration parameters are
270
-------
Appendix A: Model Fact Sheets
chosen by the user. Instruction and guidelines for parameter selection are provided in the model documentation and
online help. Further information on parameter selection is available from a wide choice of published references and
case studies.
Model Testing and Verification
MIKE 21 HD has been extensively tested on a wide range of coastal and flood projects worldwide and has a proven
record of accomplishment within the wider consulting community. A list of applications and case studies is available
on the DHI web site or by contacting DHI directly.
Model Sensitivity
The sensitivity of MIKE 21 HD to calibration parameters is largely case-dependent (e.g., in areas where the
floodplain topography is uniform and the flood slope gentle, little sensitivity to parameters is observed). However,
on floodplains with rapidly varying topographies or steep floodwater slopes, model outputs may be more sensitive to
the parameter values chosen.
Model Reliability
MIKE 21 HD has a reliability proven over many years on numerous projects worldwide. When properly calibrated,
MIKE 21 HD can predict flood levels to within 0.05m and flood flows and velocities to within 10 percent of
observed data.
Model Interface Capabilities
MIKE 21 is operated through an efficient Windows-based interactive Graphical User Interface including both
graphical and tabular editing of data. The use of generic editors makes learning easy and efficient. Hence, the time
from learning to production is short.
The advanced graphical facilities enable visual data checking and presentation of the information stored in data files
and time series databases. The same graphical environment is used for data control, analysis, and presentation of
results. The graphical presentation includes river network plan plots, cross-sectional plots, preselection of
longitudinal profiles, time series plots, comparison of measured/simulated and simulated/simulated time series,
animation of flows and water levels on both plans and profiles, control of plotting parameters, etc.
References
DHI Software. 2004. Coastal and Inland Waters in 2D. http://www.dhisoftware.com/MIKE21/
271
-------
Appendix A: Model Fact Sheets
MIKE SHE
Contact Information
Peter De Golian
Danish Hyrdrologic Institute, Inc. (DHI)
311 S. Brevard Avenue
Tampa, FL 33606
(813)254-9427
pcg(@,dhigroup.com
www.dhigroup.com
Demo versions available online only.
Download Information
Availability: Proprietary
Cost (if applicable): $3000
Model Overview/Abstract
The MIKE SHE code is a powerful, physically based, distributed parameter, fully integrated code for three-
dimensional simulation of hydrologic systems. It has been successfully applied at multiple scales, using spatially
distributed and continuous climate data to simulate a broad range of integrated hydrologic, hydraulic, and transport
problems in humid as well as in more arid areas. (See http://tvphoon.mines.edu/software/igwmcsoft/mikeshe.htm.)
Model Features
MIKE SHE can be used for the analysis, planning, and management of a wide range of water resources and
environmental problems related to surface water and groundwater, such as
• Surface water impact from groundwater withdrawal
• Conjunctive use of groundwater and surface water
• Wetland management and restoration
• River basin management and planning
• Environmental impact assessments
• Aquifer vulnerability mapping with dynamic recharge and surface water boundaries
• Groundwater management
• Floodplain studies
• Impact studies for changes in land use and climate
• Impact studies of agricultural practices including irrigation, drainage, and nutrient and pesticide
management with DAISY
Model Areas Supported
Watershed Medium
Receiving Water High
Ecological Low
Air None
Groundwater High
272
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
• Can simulate all the major processes in the land phase of the hydrologic cycle
• Is applicable on spatial scales ranging from single soil profiles (for infiltration studies) to regional
watershed studies
• Includes both simple and advanced process descriptions to maximize computational efficiency
• Has a flexible modular structure that allows users to include only the necessary processes
• Easily links regional and local scale models
• Can be linked to ESRFs Arc View for advanced GIS applications
• Includes alternate process descriptions for different applications
• Links to original data rather than importing the data
• Allows you to update your original data, and your model is automatically updated
• Includes a dynamic data tree that gives you a precise overview of all your data
• Has automatic data and model verification routines
• Includes sophisticated output tools, including animations
• Manipulating time varying data
• Model calibration
• Water and mass balance analysis
Scientific Detail
The MIKE SHE code couples several partial differential equations that describe flow in the saturated and
unsaturated zones with overland and channel flow. Different numerical solution schemes are then used to solve the
different partial differential equations for each process. A solution to the system of equations associated with each
process is found iteratively by use of different numerical solvers.
Model Framework
Internally coupled groundwater and surface water model.
Scale
Spatial Scale
• Grid-based model.
Temporal Scale
• User-defined, variable timestep.
Assumptions
Several assumptions are associated with use of the specific partial differential equations. The significant
assumptions that have direct implications to the application of the MIKE SHE code to the RFETS SWWB model
include the following.
Unsaturated Zone
The main assumption is that flow is one-dimensional and vertical. In some cases, for example, beneath ephemeral
streams, or near buildings/paved areas, or below trenches, flow in the unsaturated zone may actually have local areas
where flow is horizontal, causing this vertical-flow assumption to be violated. However, it is currently believed that
these local areas will not significantly affect the interpretation of site-wide conditions.
Other Unsaturated Zone processes not simulated in MIKE SHE Zone include the following:
• Hysteresis
273
-------
Appendix A: Model Fact Sheets
• Air entrapment
• Vapor transport
• Freezing and thawing of soils
Although, locally these processes exert strong influences on unsaturated zone flow, their effects will likely be much
less pronounced on the site-wide model dynamics and mass balance than other factors (e.g., precipitation intensity
and distribution, saturated hydraulic conductivities, and hydrostratigraphic structure).
Saturated Zone
Properties are uniform within a single grid cell. In reality, porous media properties likely vary by orders of
magnitude within each grid cell. On average, however, these local scale variations are not expected to control the
site-wide flow dynamics or mass balance, and it is reasonable to assume that properties can be averaged.
Comparisons of model simulations with observed sitewide data will help to confirm this assumption.
Overland Flow
The kinematic wave approximation is used in MIKE SHE to simulate overland flow. This simplification to the full
Saint Venant flow equations does not permit detailed simulation of backwater effects; however, given the
anticipated grid resolution of the sitewide model, the assumption is reasonable. Specific hydrologic processes, such
as rill-flow, are not considered in this code, but at the scale considered for application these processes are not likely
to be strong controls of flow.
See http://www.dhisoftware.com/mikeshe/Reviews/External Evaluations/RFETS 2-20-01 .PDF.
Model Strengths
Some of the strengths for the MIKE SHE model include
• MIKE SHE has an interface to GIS allowing for preparation of model input and presentation of model
output in a GIS environment
• Easily links up to other MIKE models
Model Limitations
Some of the limitations for the MIKE SHE model include
• Need to purchase multiple modules to take full advantage of the system
• Significant data needed to setup
Application History
The model has limited verification. See http://www.dhisoftware.com/mikeshe/Reviews/ for more information.
Model Evaluation
See http://www.dhisoftware.com/mikeshe/Reviews/index.htm.
Model Inputs
MIKE SHE includes all of the processes in the land phase of the hydrologic cycle:
• Precipitation (rain or snow)
• Evapotranspiration, including canopy interception
• Overland sheet flow
• Channel flow
• Unsaturated sub-surface flow
• Saturated groundwater flow
274
-------
Appendix A: Model Fact Sheets
Within each of these, MIKE SHE offers several different approaches ranging from simple, lumped, and conceptual
approaches to advanced, distributed, and physically based approaches. Simple and advanced approaches may be
combined, which leads to an unparalleled flexibility that truly enables you to tailor the model to the hydrological
problem rather than the opposite. Changing from a simple to a more advanced approach, or vice versa, is seamless
and utilizes, to the extent possible, the existing data.
See http://www.sfwmd.gov/org/exo/cwmp/mikeshe/fmalreport.html# Toc455045031.
Users' Guide
Available online: http://www.dhisoftware.com
Technical Hardware/Software Requirements
Computer hardware:
• 128 Mb DRAM
• 1 Gb hard drive space
• Pentium 200 Mhz minimum
Operating system:
• MS Windows 98/2000/NT/XP
Programming language:
• None
Runtime estimates:
• Depends on CPU
Linkages Supported
Other MIKE models by DHL
Related Systems
DHI series of MIKE models, built on the MIKE Zero interface.
Sensitivity/Uncertainty/Calibration
Built in Auto Calibration tool (http://www.dhisoftware.com/Generic tools/index.htmX
Model Interface Capabilities
• Dynamic navigation tree - gives you a complete overview of your model while hiding irrelevant items
• Logical, dynamic dialogs - allows you to concentrate on the required data, because subdialogs contain
only relevant parameters
• Top-down/Left-right model design - leads you through the model development in a natural manner
• Automatic data checking - saves you time trying to decipher obscure run-time errors
• Online documentation - Comprehensive, context sensitive online documentation allows you to find
answers quickly
The MIKE SHE user interface has been designed to make it easy for you to move from your conceptual model to
your results and back again.
275
-------
Appendix A: Model Fact Sheets
• Stores the link to your original data - not the original data itself, which gives you unprecedented
flexibility, for example, you can
o Run multiple models from the same dataset
o Update your data and update all your models automatically
o Run rapid sensitivity analyses on your model domain, grid spacing, layer specifications, etc.
o Use your favorite editor for managing your data
• GIS integration - the seamless link to Arc View shape files for all distributed parameters and overlays
saves you time and effort
• Geo-objects for natural, object-oriented model design - input your natural geologic formations
(including lenses!) and let MIKE SHE take care of the conversion to the numerical grid
• Link models together - easily link local-scale models to regional-scale models or multiple regional
models together to include interactions between watersheds
See http://www.dhisoftware.com/mikeshe/Interface/index.htm.
References
DHI Software. Mike SHE External Reviews. http://www.dhisoftware.com/mikeshe/Reviews/index.htm
276
-------
Appendix A: Model Fact Sheets
MINTEQA2: Metal Speciation Equilibrium Model for Surface and
Groundwater
Contact Information
U.S. Environmental Protection Agency
Office of Research and Development
National Exposure Research Laboratory
Center for Environmental Assessment Modeling
(706) 355-8400
ceam(g),epamail.epa. gov
http://www.epa.gov/ceampubl/mmedia/minteq/index.htm
Download Information
Availability: Nonproprietary
Cost: N/A
Model Overview/Abstract
MINTEQA2, a geochemical equilibrium speciation model for dilute aqueous systems, is an update of MINTEQ,
which was developed by combining the fundamental mathematical structure of MINEQL with the thermodynamic
data base of WATEQ3. MINTEQA2 is an equilibrium speciation model that can be used to calculate the equilibrium
composition of dilute aqueous solutions in the laboratory or in natural aqueous systems. The model is useful for
calculating the equilibrium mass distribution among dissolved species, adsorbed species, and multiple solid phases
under a variety of conditions, including a gas phase with constant partial pressures. A comprehensive database is
included that is adequate for solving a broad range of problems without need for additional user-supplied
equilibrium constants. The model employs a predefined set of components that includes free ions, such as Na+, and
neutral and charged complexes (e.g., H4SiO4, Cr(OH)2+). The database of reactions is written in terms of these
components as reactants. An ancillary program, PRODEFA2, serves as an interactive preprocessor to help produce
the required MINTEQA2 input files.
Model Features
• The equilibrium mass distribution among dissolved species, adsorbed species, and multiple solid phases
under a variety of conditions, including a gas phase with constant partial pressures
Model Areas Supported
Watershed None
Receiving Water High for metals chemistry
Ecological None
Air None
Groundwater None
Model Capabilities
Conceptual Basis
MINTEQA2 calculates the equilibrium composition of dilute aqueous solutions in the laboratory or in natural
aqueous systems.
277
-------
Appendix A: Model Fact Sheets
Scientific Detail
To solve the chemical equilibrium problem, MINTEQA2 uses an initial guess for the activity of each component to
calculate the concentration of each species according to mass action expressions written in terms of component
activities. The total mass of each component is then calculated from the concentrations of every species containing
that component. The calculated total mass for each component is then compared with the known input total mass for
each component. If the calculated total mass and the known input total mass for any component differ by more than
a pre-set tolerance level, a new estimate of the component activity is made, and the entire procedure is repeated
(Newton-Raphson approximation method). After equilibrating the aqueous phase, MINTEQA2 computes the
saturation index (SI) for each possible solid with respect to the solution. The solid with the most positive SI is
allowed to precipitate by depleting the dissolved concentrations of those components comprising the solid in
accordance with the known stoichiometry of each component. The reverse process occurs if an existing solid is
found to be undersaturated with respect to the solution. In either case, it is necessary to re-equilibrate the solution
after mass has been added to or depleted from the aqueous phase. Thus, the aqueous solution is re-equilibrated just
as before except with one less degree of freedom, if precipitation has occurred, or one more, if dissolution has
occurred. The entire computational loop of iterating to equilibrium, checking for precipitation or dissolution, and
shifting mass from the aqueous to the solid phase or vice versa is repeated until equilibrium is achieved and there are
no oversaturated possible solids and no undersaturated existing solids. The model calculates simultaneous solutions
of nonlinear mass action expressions and linear mass balance relationships.
Model Framework
• Dilute aqueous solutions in the laboratory or in natural aqueous systems
Scale
Spatial Scale
None
Temporal Scale
None
Assumptions
• Chemical reactions are based on equilibrium concept
Model Strengths
• Capable of simulating detailed chemical speciation simulations, including dissolved species, adsorbed
species, and multiple solid phases under a variety of conditions
Model Limitations
• Does not include time series calculation capability
Application History
MINTEQA2 has been used for numerous applications to groundwater and surface water by USGS, EPA, and
academic institutions.
Model Evaluation
The model has been internally peer reviewed by EPA Science Advisory Board (SAB) and ERD-Athens internal
review panels. Model has been externally peer reviewed via publications in the technical literature.
278
-------
Appendix A: Model Fact Sheets
Model Inputs
• Total concentrations
Users' Guide
Available online: http://www.epa.gov/ceampubl/mmedia/minteq/index.htm
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS
Programming language:
• FORTRAN
Runtime estimates:
• Seconds
Linkages Supported
None
Related Systems
PHREEQ
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
Not available
References
Benjamin, M.M. and J.O. Leckie. 1981. Multiple-Site Adsorption of Cd, Cu, Zn, and Pb on Amorphous Iron
Oxyhydroxide. J. Coll. Inter. Sci. 79:209-221.
Davies, C.W. 1962. Ion Association. Butterworths Pub., Washington, DC.
Davis, J.A., R.O. James and J.O. Leckie. 1978. Surface lonization and Complexation at the Oxide/Water Interface: I.
Computation of Electrical Double Layer Properties in Simple Electrolytes. J. Coll. Inter. Sci. 63:480-499.
Davis, J.A. and J.O. Leckie. 1978. Surface lonization and Complexation at the Oxide/Water Interface: II. Surface
Properties of Amorphous Iron Oxyhydroxide and Adsorption of Metal Ions. J. Coll. Inter. Sci. 67:90-107.
Dzombak, D. A. 1986. Toward a Uniform Model for the Sorption of Inorganic Ions on Hydrous Oxides. Ph.D. diss.,
Massachusetts Institute of Technology, Cambridge Massachusetts.
279
-------
Appendix A: Model Fact Sheets
Felmy, A.R., D.C. Girvin, and E.A. Jenne. 1984. MINTEQ—A Computer Program for Calculating Aqueous
Geochemical Equilibria. EPA-600/3-84-O32. U.S. Environmental Protection Agency, Athens, GA..
280
-------
Appendix A: Model Fact Sheets
MUSIC: Model for Urban Stormwater Improvement Conceptualization
Contact Information
Cooperative Research Center for Catchment Hydrology
Centre Office
CRC for Catchment Hydrology
Department of Civil Engineering
Building 60
Monash University, VIC 3800
crcch(@,eng. monash.edu. au
(03) 9905 2704
http ://www. toolkit, net, au/music
Download Information
Availability: Proprietary
Cost: $330 for new user
Model Overview/Abstract
The Model for Urban Stormwater Improvement Conceptualization (MUSIC) was developed by the Cooperative
Research Center (CRC) for Catchment Hydrology in Australia. MUSIC is designed to simulate urban Stormwater
systems operating at a range of temporal and spatial scales: catchments from 0.01 km2 to 100 km2 and modeling
timesteps ranging from 6 minutes to 24 hours to match the catchment's scale. MUSIC provides a user-friendly
interface, to allow complex Stormwater management scenarios to be quickly and efficiently created, and the results
to be viewed using a range of graphical and tabular formats. The Stormwater control devices that can be simulated in
MUSIC include ponds, bioretention, infiltration buffer strips, sedimentation basins, pollutant traps, wetlands, and
swales. Major algorithms applied in BMP simulation are Continuously Stirred Tank Reactors (CSTRs) in series
model and first-order decay (k-C* model).
Model Features
• Urban Stormwater treatment conceptual design tool
Model Areas Supported
Watershed Medium (urban)
Receiving Water Low
Ecological None
Air None
Groundwater Low
Model Capabilities
Conceptual Basis
The algorithm adopted to generate urban runoff is based on a simplified mass balance model that was developed by
Chiew et al. (1997). For simulation of BMPs, the Plus Method for reservoir routing is used to simulate the
movement of water through the treatment system, and a first-order kinetic model combined with the Continuously
Stirred Tank Reactors (CSTRs) model to simulate the removal of pollutants within the treatment system.
281
-------
Appendix A: Model Fact Sheets
Scientific Detail
The simplified rainfall-runoff model (Chiew et al., 1997) is a mass balance model, which involves potential ET,
impervious storage, soil moisture storage, and a groundwater component.
For stormwater treatment process simulation, assume that the pollutant removal is a first-order decay process toward
an equilibrium value. This behavior is described by the first-order kinetic (k-C*) model, in which C* is the
background concentration, and k is the decay rate constant. In addition, the number of CSTRs is used to reflect the
shape factor on pollutant removal effectiveness. The concentration attenuation simulated by the k-C* model is
computed separately at each timestep for each CSTR.
Model Framework
A catchment (the entire catchment being simulated) is made up of a number of nodes, joined together by drainage
links. The catchment may contain a number of subcatchments, which are called source nodes. Three types of source
nodes represent three default land uses: urban, agricultural and forested, these source nodes differ only in their
default baseflow and stormflow pollutant concentrations. Users therefore can create source nodes to simulate any
type of land use (e.g. road runoff), using their own water quality data. A catchment contains only one receiving
node, which represents the receiving waterway (e.g. River, lake, bay). A catchment may also have a number of
junction nodes, which simply act as confluences. They have no effect on flow or water quality; they simply join
multiple upstream nodes into one. Treatment nodes are used to represent stormwater treatment measures.
Scale
Spatial Scale
• Catchment, 0.01 km2 to 100 km2
Temporal Scale
• 6 minutes to 24 hours
Assumptions
• Physical process (sedimentation) is the predominant pollutant removal mechanism during the event and is
described by the order kinetics (k-C*) model
• Constant seepage rate for infiltration processes
Model Strengths
• Includes an intuitive and user-friendly interface
• Capable of simulating various type of BMPs
• Suitable for evaluation of urban stormwater treatment conceptual design
Model Limitations
• The first-order kinetics (k-C*) modeling approach adopted in the USTM strictly applies only during event
operation. The parameter k lumps together the influence of a number of predominantly physical factors on
the removal of stormwater pollutants. While the assumption of a predominance of physical removal
processes during storm event operation is reasonable for paniculate (inorganic) contaminants, other factors
associated with chemical and biological processes can also be significant. These factors are currently not
accounted for in the determination of k. The background concentration C* is assumed to be a constant at
present and therefore does not reflect the influence of hydraulic loading, flow velocity, and other factors on
water quality of stormwater runoff.
• The treatment device infiltration process is simulated by applying a constant seepage rate.
282
-------
Appendix A: Model Fact Sheets
• Default empirical parameters values provided in the model were derived from monitoring data collected in
Australia, and it is strongly recommended that the model be calibrated before it is applied.
Application History
The model has been applied mainly in Australia.
Model Evaluation
Not available
Model Inputs
• Climate data (MUSIC comes with Bureau of Meteorology-formatted climate files for 30 reference cities
widely distributed throughout Australia.)
• Catchment characteristics (area, land use, impervious area, etc)
• Conceptual designs of stormwater treatment measures (type, size, etc)
Users' Guide
Available online: http://www.toolkit.net.au/music
Technical Hardware/Software Requirements
Computer hardware:
• Intel-based PC with CD-ROM drive
• Pentium III 750Mhz (preferably IGHz)
• 256Mb RAM (preferably 512 Mb)
• 5Gb of free hard drive space (preferably 10Gb)
• Monitor capable of 800 x 600 pixels (preferably 1024 x 768) @ 8-bit (256) colors (preferably 16-bit)
Operating system:
• Microsoft Windows 2000/XP
Programming language:
• Unknown
Runtime estimates:
• Minutes to hours
Linkages Supported
None
Related Systems
None
Sensitivity/Uncertainty/Calibration
Displays the observed value with modeled time series to facilitate calibration process.
283
-------
Appendix A: Model Fact Sheets
Model Interface Capabilities
• Graphical user interface for node/link network construction and data input, viewing, and editing
• Post-processor displays time series graphs, tables of statistics, cumulative frequency graphs, and file export
References
Chiew, F.H.S. & McMahon, T. A. 1997. Modelling Daily Runoff and Pollutant Load from Urban Catchments.
Water - Journal of the Australian Water Association. 24:16-17.
Wong, T. H. F., Duncan, H. P., Fletcher, T. D., Jenkins, G. A., & Coleman, J. R. 2001. A unified approach to
modelling urban stormwater treatment. Paper presented at the Second South Pacific Stormwater Conference, June
27-29, 2001, Auckland, New Zealand.
Wong, T.H.F. 2000. Improving Urban Stormwater Quality - From Theory to Implementation. Water - Journal of
the Australian Water Association. 27(6):28-31.
284
-------
Appendix A: Model Fact Sheets
P8-UCM: Program for Predicting Polluting Particle Passage through Pits,
Puddles, and Ponds—Urban Catchment Model
Contact Information
William Walker, Jr.,
Environmental Engineer
1127 Lowell Road
Concord, Massachusetts 01742-5522
(978) 369-8061
wwwalker(g),wwwalker. net
http ://www. wwwalker.net
Download Information
Availability: Nonproprietary
http ://wwwalker. net/p8/
Cost: N/A
Model Overview/Abstract
P8 is a model for predicting the generation and transport of stormwater pollutants in urban watersheds. Continuous
water balance and mass balance calculations are performed on a user-defined system consisting of watersheds,
devices (runoff storage/treatment areas, BMPs), particle classes, and water quality components. Simulations are
driven by continuous hourly rainfall and daily air temperature time series data. The model simulates pollutant
transport and removal in a variety of treatment devices (BMPs), including swales, buffer strips, detention ponds
(dry, wet, and extended), flow splitters, and infiltration basins (offline and online), pipes, and aquifers. Water quality
components include total suspended solids (TSS) (five size fractions), total phosphorus (TP), total Kjeldahl nitrogen
(TKN), copper, lead, zinc, and hydrocarbons.
Model Features
• Urban watershed hydrology
• Urban pollutants
• Stormwater BMPs
Model Areas Supported
Watershed Medium
Receiving Water Low
Ecological None
Air None
Groundwater Low
Model Capabilities
Conceptual Basis
In P8-UCM, continuous water balance and mass balance calculations are performed on a user-defined system
consisting of watersheds (pervious and impervious areas are separately considered), devices (runoff
storage/treatment areas, BMPs), particle classes, and water quality components.
285
-------
Appendix A: Model Fact Sheets
Scientific Detail
Runoff from pervious areas is computed using the Soil Conservation Service's (SCS) curve number technique.
Antecedent moisture conditions are adjusted, based on 5-day antecedent precipitation and season. Percolation from
pervious areas is estimated by water balance at the surface (percolation = precipitation - runoff -
evapotranspiration). Evapotranspiration is computed from air temperature and season using Hamon's method.
Runoff from impervious areas starts after the cumulative storm rainfall exceeds the specified depression storage.
Both rainfall and snowmelt are considered in runoff estimations. Particle concentrations in runoff from pervious
areas are computed using a method similar to the sediment rating curve included in EPA's Stormwater Management
Model (SWMM). Particle loads from impervious areas are computed using either or both of two techniques: (1)
particle accumulation and washoff and/or (2) fixed runoff concentration. The first method is used in default particle
datasets. An exponential washoff relationship similar to that employed in SWMM is used to simulate particle
buildup and washoff from impervious surfaces.
Receiving water processes are limited to devices, ponds, infiltration basins, and shallow channels. Storage area or
volume and outflow relations represent flow in ponds. Shallow channel flow is estimated by Manning equation.
Settling and transport of sediments are simulated in the model.
Model Framework
• Watershed model
• Shallow channels, ponds, infiltration basins, and storage devices
Scale
Spatial Scale
• Subwatersheds
Temporal Scale
• Hourly
Assumptions
• A watershed is divided into a lumped pervious area and a lumped impervious area.
• SCS Curve Number approach is appropriate for estimating surface runoff.
• All the pollutants entering the waterbodies are sediment-adsorbed.
Model Strengths
• A simple model that requires moderate effort to setup, calibrate, and validate
• Simulates urban stormwater BMPs and wetlands
Model Limitations
• SCS Curve Number approach at hourly timestep requires substantial calibration
• Limited capability in flow and pollutant routing
• Limited capability in groundwater process and groundwater and surface water interaction
Application History
P8 is widely applied in the Northeast and Midwest regions of the United States, especially to size stormwater BMPs
in urban watersheds.
286
-------
Appendix A: Model Fact Sheets
Model Evaluation
P8 documentation presents model evaluation in various applications.
Model Inputs
• Climate data: hourly precipitation and daily air temperature
• Device (hydraulic) parameters for pond, basin, buffer, pipe, splitter, and aquifer
• Watershed parameters: areas, impervious fraction and depression storage, street-sweeping frequency, SCS
runoff curve number for pervious portion
• Particle parameters: accumulation/washoff parameters, runoff concentrations, street-sweeper efficiencies,
settling velocities, decay rates, filtration efficiencies
• Water quality component parameters: pollutant concentrations
Users' Guide
Available online: http://www.wwwalker.net/p8/p8vldoc.pdf
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS
Programming language:
• FORTRAN
Runtime estimates:
• Minutes to less than an hour
Linkages Supported
None
Related Systems
None
Sensitivity/Uncertainty/Calibration
Built in capability for calibration and sensitivity analysis
Model Interface Capabilities
• DOS User Interface
• P8CONV - a preprocessor to develop precipitation input file
References
Palmstrom, N. & W. Walker. 1990. The P8 Urban Catchment Model for Evaluating Nonpoint Source Controls at
the Local Level, Enhancing States' Lake Management Programs. U.S. Environmental Protection Agency,
Washington, DC.
287
-------
Appendix A: Model Fact Sheets
U.S. Environmental Protection Agency (USEPA). 1992. Compendium of Watershed-Scale Models for TMDL
Development. EPA841-R-92-002. U.S. Environmental Protection Agency, Office of Water, Washington, DC.
HDR Inc. 1992. Evaluation of Storm Water Computer Models. Prepared for City of Minneapolis by HDR Inc.
U.S. Environmental Protection Agency (USEPA). 1997. Compendium of Tools for Watershed Assessment & TMDL
Development. EPA841-B-97-006. U.S. Environmental Protection Agency, Office of Water, Washington, DC.
Walker, W. 1990. P8 Urban Catchment Model Program Documentation Version 1.1. (Computer program). IEP,
Inc., Northboroug, MA and Narragansett Bay Project, Providence, RI.
288
-------
Appendix A: Model Fact Sheets
PCSWMM: Storm Water Management Model
Contact Information
Computational Hydraulics Int. (CHI)
3909WitmerRoad#960
Niagara Falls, NY 14305-1244
1-800-891-8447
infotgicomputationalhydraulics.com
http://www.computationalhydraulics.com/
Download Information
Availability: Proprietary
http://www.computationalhydraulics.com/Software/PCSWMM/index.html
Cost: $599.95
Model Overview/Abstract
PCSWMM 2003 has a menu-driven interface for running the EPA's Stormwater Management Model (SWMM) and
providing hydro/pollutographs and animated hydraulic gradelines. It is suitable for projects ranging from small BMP
installations to continuous hydrology, hydraulics and quality simulation of major and minor drainage systems. It also
provides GIS links to the EPA's SWMM core processes.
PCSWMM 2003 is flexible to be used with any of the following versions of the SWMM engine: WMM4.4h,
SWMM4.3 and later SWMM4.31, 4.40 and 4.4gu releases. PCSWMM fully supports all modules of SWMM,
including the Rain, Temperature, Runoff, Transport, Extran, Storage/Treatment, Combine, and Statistics modules.
Model Features
• Watershed hydrology and water quality
• Stream transport
• Urban stormwater systems and pipes
Model Areas Supported
Watershed High
Receiving Water Medium
Ecological None
Air None
Groundwater Low
Model Capabilities
Conceptual Basis
The basic spatial unit for SWMM is the subcatchment, into which the modeled watershed is subdivided. Several
small subwatersheds and representative streams may be networked together to represent a larger watershed drainage
area. The SWMM-RUNOFF and SWMM-TRANSPORT modules can be used to simulate various processes, both
on land and in-stream.
289
-------
Appendix A: Model Fact Sheets
Scientific Detail
Infiltration is calculated using the Horton or Green-Ampt methods, at the user's choice. A version of Manning's
equation is used to estimate flow rate from the subcatchment area based on a conceptual model of the subcatchment
as a "nonlinear reservoir." The lumped storage scheme is applied for soil and groundwater modeling. For
impervious areas, a linear formulation is used to compute daily and hourly increases in particle accumulation. For
pervious areas, a modified Universal Soil Loss Equation (USLE) determines sediment load. The concept of potency
factors is applied to simulate pollutants other than sediment.
Transport block has kinematic wave routing of flow and quality, base flow generation, and infiltration capabilities,
and it routes flow through user-defined system ranges from natural channel to concrete pipes. The EXTRAN block
carries out a numerical solution of the complete St. Venant equations for the urban drainageways and conduits, by
modeling the network as a link-node system (cf, DYNHYD). SWMM can directly be interfaced with EPA's WASP
receiving water quality model.
Model Framework
• Subwatersheds and watershed
• One-dimensional mass balance flow and pollutant routing
Scale
Spatial Scale
• Subwatershed. Flexible size
• One-dimensional channel/pipe system
Temporal Scale
• User-defined timestep, typically hourly
Assumptions
• The model performs best in urbanized areas with impervious drainage, although it has been widely used
elsewhere.
• Model parameters for quantity and quality simulations are developed such that the model will be calibrated
to enhance its capability.
• Water table elevation is assumed to be fixed value.
• All the pollutants entering the waterbodies are sediment-adsorbed.
Model Strengths
• Unlimited model size
• Flexible input file editor
• Easy to setup SWMM runs
• Hot-swapping SWMM engines
• User friendly
• Sensitivity, calibration, and error analysis tools
• Full support of all SWMM modules and procedures
Model Limitations
• Not a public domain product
• Lack of subsurface quality routing
• No interaction of quality processes
• Limited kinetics (a first-order decay rate can be specified for each pollutant in the Transport Block)
• Difficulty in simulation of wetlands quality processes
290
-------
Appendix A: Model Fact Sheets
• Rudimentary scour-deposition routine in the Transport Block
Application History
SWMM is applied to urban hydrologic quantity and quality problems in scores of U.S. cities as well as extensively
in Canada, Europe, and Australia. The model has been used for very complex hydraulic analysis of combined sewer
overflow mitigation, as well as for many stormwater management planning studies and pollution abatement projects,
and there are many instances of successful calibration and verification (Huber, 1992). Warwick and Tadepalli (1991)
describe calibration and verification of SWMM on a 10-square-mile urbanized watershed in Dallas, Texas.
Tsihrintzis et al. (1995) describe SWMM applications to four watersheds in South Florida representing high- and
low-density residential, commercial, and highway land uses.
Model Evaluation
It is widely applied in Florida, but the applications are limited primarily to urban areas, stormwater studies, and
event based applications.
Model Inputs
• Data requirements for hydrologic simulation include area, imperviousness, slope, roughness, width,
depression storage, and infiltration parameters. Land use data are used to determine ground cover type for
each model subarea.
• Depending on the options set for the loading calculations, additional parameters are necessary (e.g., buildup
coefficients would be needed for the dry weather buildup simulation).
• Additional data are necessary if the user intends to model subsurface drainage and interflow.
• Depending on the stormwater system, dimensions, slope, roughness coefficients, elevations, storage, etc.,
are required.
• Continuous records of evapotranspiration, temperature, and solar intensity
Users' Guide
Available online: http://www.computationalhvdraulics.com/Publications/Books/r219.html
Cost: $85
Technical Hardware/Software Requirements
Computer hardware:
• PC with 50MB available hard-drive space and preferably 256MB RAM
Operating system:
• Microsoft Windows
Programming language:
• FORTRAN (model) and Visual Basic (interface)
Runtime estimates:
• Minutes to less than an hour
Linkages Supported
EPA's Stormwater Management Model (SWMM)
291
-------
Appendix A: Model Fact Sheets
Related Systems
PCSWMM provides a SWMM environment for model development, display and analysis. GIS component of
PCSWMM is completely rewritten with innovative model creation and output visualization tools.
Sensitivity/Uncertainty/Calibration
The Genetic Algorithm Calibration tool is provided in PCSWMM for the complete calibration of SWMM Runoff,
Transport, Extran and/or Storage Treatment modules. It significantly reduces the effort required for calibration and
design optimization. This tool helps in model development and verification.
Model Interface Capabilities
• User-friendly model interface
• Graphical dialog boxes
• Data input is through wizard-style interface technology
• Data entry and model output in graphical form
• Animated hydraulic grade lines and multiple observed vs. computed plots provide visualizations of model
results
• Background layer support for Arc View, AutoCAD, Maplnfo, TIFF, JPEG, and BMP
References
Donigian, A.S., Jr., and W.C. Huber. 1991. Modeling ofnonpoint source water quality in urban and non-urban
areas. EPA/600/3-91/039. U.S. Environmental Protection Agency, Environmental Research Laboratory, Athens,
GA.
Huber, W. C. 1992. Experience with the EPA SWMM Model for analysis and solution of urban drainage problems.
In Proceedings of el Inundaciones Y Redes De Drenaje Urbano, ed. J. Dolz, M. Gomez, and J. P. Martin, Colegio de
Ingenieros de Caminos, Canales Y Puertos, Universitat Politecnica de Catalunya, Barcelona, Spain, 1992, pp. 199-
220.
Huber, W.C. 2001. New options for overland flow routing in SWMM. ASCE-EWRI. World Water and
Environmental Congress, Orlando, FL.
Huber, W.C., and R.E. Dickinson. 1988. Storm Water Management Model Version 4, User's manual. EPA 600/ 3-
88/ 00la (NTIS PB88-236641/ AS). U.S. Environmental Protection Agency, Athens, GA.
Irvine, K.N., E.G. Loganathan, E.J. Pratt and H.C. Sikka. 1993. Calibration of PCSWMM to estimate metals, PCBs
and HCB in CSOs from an industrial sewershed. In New Techniques for Modeling the Management ofStormwater
Quality Impacts, ed. W. James, Lewis Publishers, Boca Raton, FL. pp. 215-242.
James, W., Huber, W. C., Pitt, R. E., Dickinson, R. E., and James, R. C. 2002. Water Systems Models [1]:
Hydrology, User's guide to SWMM4 RUNOFF and supporting modules and to PCSWMM Version 2.4.
Computational Hydraulics International, Guelph, Ontario, Canada.
James, W., Huber, W. C., Pitt, R. E., Dickinson, R. E., Roesner, L. A., Aldrich, J. A., and James, R. C. 2002. Water
Systems Models [2]: Hydraulics, User's guide to SWMM4 TRANSPORT, EXTRAN and STORAGE modules and to
PCSWM. Version 2.4. Computational Hydraulics International. Guelph, Ontario, Canada.
Tshihrintzis, V. A., R. Hamid, and H. R. Fuentes. 1995. Calibration and verification of watershed quality model
SWMM in subtropical urban areas. In Proceedings of the First International Conference—Water Resources
Engineering, American Society of Civil Engineers, San Antonio, TX, August 14-16, 1995, pp. 373-377.
Tsihrintzis, V. and R. Hamid. 1998. Runoff quality prediction from small urban catchments using SWMM.
Hydrological Processes. 12(2):311-329.
292
-------
Appendix A: Model Fact Sheets
Warwick, J. I, and P. Tadepalli. 1991. Efficacy of SWMM application. Journal of Water Resources Planning and
Management. 117(3):352-366.
293
-------
Appendix A: Model Fact Sheets
PGC - BMP Module: Prince George's County Best Management Practice
Module
Contact Information
Mow-Soung Cheng
Section Head, Technical Support Section
Prince George's County Department of Environmental Resources
Programs and Planning Division
9400 Peppercorn Place, Sixth Floor
Largo, MD 20774
(301) 883-5836
mscheng(g),co.pg.md.us
Download Information
Availability: Nonproprietary
Cost: N/A
Model Overview/Abstract
Prince George's County, working with Tetra Tech, Inc., developed a best management practice (BMP) evaluation
module to assist in assessing the effectiveness of low-impact development (LID) technology. This module uses
simplified process-based algorithms to simulate BMP control of modeled flow and water quality time series
generated from runoff models such as the HSPF. These unit-process algorithms include weir and orifice control
structures, storm swale characteristics, flow and pollutant transport, flow routing and networking, infiltration and
saturation, evapotranspiration, and a general loss/decay representation for a pollutant. The module offers the user the
flexibility to design retention style or open-channel BMPs, define flow routing through a BMP or BMP network,
simulate IMPs such as reduced or discontinued imperviousness through flow networking, and compare BMP
controls against some defined benchmark such as a simulated pre-development condition. Because the underlying
algorithms are based on physical processes, BMP effectiveness can be evaluated and estimated over a wide range of
storm conditions, BMP designs, and flow routing configurations. Such a tool provides a quantitative medium for
assessing and designing TMDL allocation scenarios and evaluating the effectiveness of a proposed management
approach.
Model Features
• Retention-style BMP simulation
• Open-channel BMP simulation
• Soil media storage, infiltration, and filtration
• Time series input and output simulation
• Linked water quality, with generalized first-order pollutant loss and filtration loss
Model Areas Supported
Watershed Medium
Receiving Water Medium
Ecological Low
Air None
Groundwater Low
294
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
The BMP module requires unit-area time series of runoff from different land uses, typically generated by a related
watershed model of the area. It offers the flexibility to schematically represent the site's land parcel distribution and
flow routing (from land segment to BMP or BMP to BMP). Each BMP represents an evaluation point, where the
effect of the BMP or BMP network is assessed. An analysis spreadsheet allows the user to compare the developed
condition with and without BMPs to the pre-developed condition (typically represented as an equal-area forest site).
Analysis includes long-term volume and loads, storm-by-storm hydrograph and pollutograph comparison, and multi-
storm trend analysis for peak flow, volume control, and pollutant load reduction.
Scientific Detail
An external model that is capable of generating hourly time series of flow and pollutant load is used to generate the
required runoff inputs for the BMP Module. Examples of such models include HSPF, LSPC, and SWMM. The
primary function of the BMP Module is to provide detailed simulation of processes within BMPs. The BMP Module
presents two main categories for BMPs: Class A, retention/detention structures, and Class B, open-channel
structures. Between these two categories, the actual BMP unit processes may be combined interchangeably to
represent a wide range of BMPs. These BMP unit processes include (or are influenced by) evapotranspiration,
infiltration, orifice outflow, under-drain outflow, weir-controlled overflow or spillway, BMP bottom slope, bottom
roughness, soil media filtration of pollutant using under-drain outflow, and general first-order loss or decay of
pollutant. Computational methods include standard rectangular or triangular weir equations, orifice equations,
Manning's Equation coupled with the continuity equation, and the Holtan-Lopez infiltration model. This infiltration
model is built on the premise that infiltration is proportional to the capacity of the soil to store water, giving it an
advantage over other methods (e.g. Morton's equation, Green-Amp equation) in that it is physically based and
describes the infiltration and recovery capacity during low-flow or dry periods (Haan et al., 1994). This equation
was developed on the premise that soil moisture storage, surface-connected porosity, and the effects of root paths are
the dominant factors influencing infiltration capacity (Maidment, 1993).
Model Framework
• Modeled runoff inputs are land use-specific unit-area time series
• One-dimensional mass balance flow and pollutant routing (land-to-BMP or BMP-to-BMP)
• Combinations of BMP unit process used to simulate activity within a management structure
• Assumes complete mixing within BMP compartments
Scale
Spatial Scale
• Site-level or small watershed-scale analysis
Temporal Scale
• Hourly input and output time series
Assumptions
Simulated time series components can be spatially distributed and scaled by area; however, unit characteristics
remain constant within each unique land use type. Flow routing and simulation order are performed on a top-down
basis. Complete mixing is assumed within the surface storage volume of the BMP. Infiltrating or outflow water exits
at the current concentration of the completely mixed surface storage volume. The module assumes one-directional
flow from land to BMP or BMP to BMP.
295
-------
Appendix A: Model Fact Sheets
Model Strengths
The BMP Module interface allows the user to design the schematic site layout and represent flow diversions and
alterations by the placement of BMP structures. BMP scenarios such as reduced imperviousness can be represented
as a change in land use parcels. The BMP Module offers a great deal of detail in how a BMP can be physically
represented and simulates both hydrologic and generalized water quality activity within the BMP. It uses process-
based algorithms to represent BMP unit processes. The result is a time history of both influent and effluent flow and
pollutant levels on an hourly basis. A Visual Basic Applications (VBA)-enhanced Excel analysis spreadsheet
automatically queries and summarizes model results and produces a set of BMP performance measures for analysis
between multiple scenarios.
Model Limitations
Because the model currently supports hourly time series, the model assumptions are most valid with BMPs that have
residence times of at least an hour. Runoff characteristics must be predetermined within an external watershed
model; therefore, factors like land slope and its influence on influent runoff to BMPs cannot be changed from within
the BMP Module.
Application History
Conceptualization and design of the BMP Module began in late 1999 to early 2000. The Module has been under
internal testing, development, and enhancement during its relatively short recent history. Applications include both
proposed retrofit and development sites in Prince Georges County, Maryland. Examples include a commercial site
example from a Maryland state design manual, some proposed LID integrated site plans, and individual BMP types
in the LID Design Manual (Prince George's County, 1999). BMP Module application has been limited to screening-
level and planning purposes. The module has never been used for BMP design purposes.
Model Evaluation
A validation/confirmation exercise was performed using the BMP Module and laboratory data that documented
measured performance of a set of bioretention basins (Davis, 2001). The BMP Module was configured and loaded
consistent with the reported data to test the module's predictive capability. The results showed good agreement
between modeled and observed BMP performance.
Model Inputs
• Continuous hourly runoff unit-area time series model output by land parcel type
• Relevant BMP design and dimensional information
Users' Guide
A guide is available from Prince George's County (Tetra Tech, 2003).
Technical Hardware/Software Requirements
Computer hardware:
• IBM-compatible PC
Operating system:
• Windows 98 or later
Programming language:
• Module interface: C++, Analysis Tool: Visual Basic Applications in Microsoft Excel
296
-------
Appendix A: Model Fact Sheets
Runtime estimates:
• Seconds to minutes, depending on spatial/temporal resolution and computer performance
Linkages Supported
• Input format is closely compatible with HSPF, LSPC, or SWMM time series outputs
• Output format links to Excel Analysis Tool
Related Systems
The related systems listed below contain BMP simulation capability:
• Public domain models: SWMM, P8, VFSMOD
• Proprietary models: MUSIC, LIFE
Sensitivity/Uncertainty/Calibration
• Runoff inputs must be calibrated externally within the context of the relevant runoff generation model.
• BMP Module specifications include BMP size and/or geometry.
• Calibration parameters are associated with the infiltration method.
• The model is primarily designed to represent the relative change or impact of a BMP for comparison
between scenarios. Actual validation/confirmation of the model using data requires detailed BMP influent
and effluent monitoring.
Model Interface Capabilities
• User-friendly model interface.
• Data input allows user to build library of predefined BMPs.
• Drag-and-drop interface allows users to build complex networks of land, and single or multiple BMPs in
series.
• Postprocessing tools provide a graphic summary of multiple evaluation criteria.
References
Bowie G., W. Mills, D. Porcella, C. Campbell, J. Pagenkopf, G. Rupp, K. Johnson, P. Chan, S. Gherini, C.
Chamberlin. 1985. Rates, Constants, and Kinetics Formulations in Surface Water Quality Modeling. Ed. 2. EPA
600/3-85/040. U.S. Environmental Protection Agency, Washington, DC.
Davis, A.P., M. Shokouhian, H. Sharma, and C. Minami. 2001. Laboratory Study for Biological Retention for Urban
Stormwater Management. Water Environment Research. 73(1).
Haan, C.T., B.J. Barfield, and J.C. Hayes. 1994. Design Hydrology and Sedimentology for Small Catchments.
Academic Press, San Diego, CA.
Linsley, R., J. Franzini, D. Freyberb, G. Tchobanoglous. 1992. Water-Resource Engineering. 4th Ed.. McGraw-Hill,
New York.
Maidment, D. 1993. Handbook of Hydrology. McGraw-Hill, New York.
Prince George's County. 1999. Low-Impact Development Design Strategies: An Integrated Design Approach.
Department of Environmental Resources Programs and Planning Division, Largo, MD.
Tetra Tech, Inc. 2003. Low-Impact Development Management Practices Evaluation Computer Module - User's
Guide. Prepared for Prince George's County by Tetra Tech, Inc., Fairfax, VA.
297
-------
Appendix A: Model Fact Sheets
QUAL2E: Enhanced Stream Water Quality Model
Contact Information
Paul Cocca
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, DC 20460
(202) 566-0406
cocca.paul@epa. gov
Download Information
Availability: Nonproprietary
http://www.epa.gov/docs/OUAL2E WINDOWS/
http://www.epa.gov/ceampubl/swater/index.htm
Cost: N/A
Model Overview/Abstract
The QUAL2E model simulates nutrient dynamics, algal production, and dissolved oxygen with the impact of
benthic and carbonaceous demand in streams. Fifteen water quality variables are modeled in QUAL2E. The model
solves the time-variable water quality variable under steady, nonuniform flow. It can be applied to steady state and
diurnal time-variable situations.
Model Features
• Dissolved oxygen and oxygen demand
• Nutrients
• Tracer
• Algal dynamics
• Bacteria
Model Areas Supported
Watershed None
Receiving Water Medium
Ecological Medium
Air None
Groundwater None
Model Capabilities
Conceptual Basis
Rivers are conceptualized as a collection of reaches. Within each reach, the hydrogeometric properties are assumed
to be same. The reaches are further divided into a series of control volumes of the same length.
298
-------
Appendix A: Model Fact Sheets
Scientific Detail
The governing equations of QUAL2E are the advection-dispersion-reaction equations with external sources and
sinks. A flow balance is assumed, and steady, nonuniform flow is used to solve the advection. QUAL2E assumes
that the channel has a trapezoidal cross section. An empirical function is applied to obtain the dispersion coefficient.
The governing equations are solved with an implicit scheme in time and a backward difference in space.
Model Framework
• One-dimensional model
• Rivers, streams, channel network
Scale
Spatial Scale
• One-dimensional horizontal, nonuniform flow
Temporal Scale
• Steady state
Assumptions
• Steady, nonuniform flow
• Trapezoidal cross section channel
• Flow balance
Model Strengths
• A simple model with comprehensive nutrient, algal, and dissolved oxygen dynamics
• Easy to use, easy to understand
• Complete documentation
• Sensitivity analysis with QUAL2E-UNCAS
Model Limitations
• One-dimensional channel that cannot handle tidal impact
• Equal-length elements
• Steady flow (not able to model variable flow condition)
• Specified sediment oxygen demand (SOD); no sediment diagenesis
Application History
QUAL2E has been widely applied in the United States and around the world, including Chile, Italy, Spain, Slovenia,
India, and South Africa.
Model Evaluation
QUAL2E has been widely tested. Materials related to QUAL2E can be found in textbooks, journal papers, and
technical reports. An example is the evaluation of QUAL2E written by Francois Birgand, available at
http://www3.bae.ncsu.edu/Regional-Bulletins/Modeling-Bulletin/qual2e.html.
Model Inputs
• Reach identification and river mile/kilometer data
• Computational elements flag field data
299
-------
Appendix A: Model Fact Sheets
• Hydraulic data
• Biochemical oxygen demand and dissolved oxygen rate constants
• Initial conditions
• Incremental inflow
• Headwater sources
• Point source or withdrawal
Users' Guide
Available online: http://smig.usgs.gov/cgi-bin/SMIC/model homejages/modelhome?selection=qual2e.
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS, Windows
Programming language:
• FORTRAN
Runtime estimates:
• Minutes
Linkages Supported
BASINS
Related Systems
QUAL2K, DYNHYD5/WASP5, CE-QUAL-RIV1
Sensitivity/Uncertainty/Calibration
QUAL2E-UNCAS conducts uncertainty analysis
Model Interface Capabilities
• DOS version: QUAL2E reads text-based input file
• Latest QUAL2E Windows version provides Windows interface
References
Barnwell, T. O. and Brown, L. C. 1987. The Enhanced Stream Water Quality Models QUAL2E and QUAL2E-
UNCAS: Documentation and User Manual. EPA/600/3-87/007. U.S. Environmental Protection Agency,
Washington, DC.
Manual for Windows Interface available separately as: U.S. Environmental Protection Agency (USEPA). 1995.
QUAL2E Windows Interface Users Guide. EPA/823/B/95/003. U.S. Environmental Protection Agency, Washington,
DC.
Chapra, S.C. 1997. Surface Water Quality Modeling. McGRAW-HILL, Inc. New York
300
-------
Appendix A: Model Fact Sheets
QUAL2K
Contact Information
Steven C. Chapra
Professor, Berger Chair
Tufts University
Department of Civil & Environmental Engineering
Anderson Hall, Medford, MA 02155
(617)627-3654
Steven. chapra(@,tufts. edu
http://www.epa.gov/ATHENS/wwqtsc/html/qual2k.html
Download Information
Availability: Nonproprietary
http://www.epa.gov/ATHENS/wwqtsc/html/qual2k.html
Cost: N/A
Model Overview/Abstract
The QUAL2K model simulates nutrient dynamics, algal production, and dissolved oxygen with the impact of
benthic and carbonaceous demand in streams. QUAL2K is similar to QUAL2E but QUAL2K is enhanced with two
species of CBOD and internal sediment processes. In addition, QUAL2K models pH and alkalinity. The pathogen
die-off is modeled as a function of temperature, light intensity, and settling. The model solves the time-variable
water quality variable under steady, nonuniform flow. QUAL2K can be applied to model steady state and dinurnal
time-variable situations.
Model Features
• Dissolved oxygen and oxygen demand
• Nutrients
• Tracer
• Algal dynamics
• Bacteria
• pH, alkalinity
Model Areas Supported
Watershed None
Receiving Water Medium
Ecological Medium
Air None
Groundwater None
Model Capabilities
Conceptual Basis
Rivers are conceptualized as a collection of reaches. Within each reach, the hydrogeometric properties are assumed
to be same. The reaches are further divided into a series of control volumes of the same length.
301
-------
Appendix A: Model Fact Sheets
Scientific Detail
The governing equations of QUAL2K model are the advection-dispersion-reaction equations with external sources
and sinks. A flow balance is assumed, and steady, nonuniform flow is used to solve the advection. QUAL2K
assumes that the channel has a trapezoidal cross section. An empirical function is applied to obtain the dispersion
coefficient. The governing equations are solved with an implicit scheme in time and a backward difference in space.
Model Framework
Structural description of the model, internal linkages, such as one-dimensional model, internally coupled watershed
and one-dimensional stream models...
Narrow rivers and streams
Scale
Spatial Scale
• One-dimensional
Temporal Scale
• Steady state
Assumptions
• Steady, nonuniform flow
• Trapezoidal cross section channel
• Flow balance
Model Strengths
• A simple model with comprehensive nutrient, algal, and dissolved oxygen dynamics
• Easy to use, easy to understand
• Internal sediment processes
• Complete documentation
Model Limitations
• One-dimensional channel cannot handle tidal impact
• Equal length elements
• Steady flow; not able to model variable flow condition
Application History
Application examples include the Wind River temperature TMDL in Washington State and biochemical oxygen
demand simulation in the Hanjiang River, China.
Model Evaluation
Not available
Model Inputs
• Reach identification and river mile/kilometer data
• Computational elements flag field data
302
-------
Appendix A: Model Fact Sheets
• Hydraulic data
• Biochemical oxygen demand and dissolved oxygen rate constants
• Initial conditions
• Incremental inflow
• Headwater sources
• Point source or withdrawal
Users' Guide
Available online: http://www.epa.gov/ATHENS/wwqtsc/html/qual2k.html
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• Windows ME/2000/XP with MS Office 2000 or Higher
Programming language:
• Excel VBA
Runtime estimates:
• Minutes to hours
Linkages Supported
None
Related Systems
QUAL2E, CE-QUAL-RIV1, DYNHYD5/WASP5
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Microsoft Excel interface
References
Chapra, S.C. and Pelletier, G.J. 2003. QUAL2K: A Modeling Framework for Simulating River and Stream Water
Quality: Documentation and Users Manual. Civil and Environmental Engineering Department, Tufts University,
Medford, MA.
303
-------
Appendix A: Model Fact Sheets
REMM: Riparian Ecosystem Management Model
Contact Information
Richard Lowrance
U.S. Department of Agriculture
Agricultural Research Station
Southeast Watershed Research Laboratory
Coastal Plain Experiment Station
P.O. Box 946, Tifton, GA
(912) 386 3894
lorenz(@,tifton. cpes.peachnet.edu
http://sacs.cpes.peachnet.edu/remmwww/remm/remmoldwww/default.htm
Download Information
Availability: Nonproprietary
Cost: N/A
As of July 2004, REMM 1.0 is under testing. The testing version can be downloaded at
http://sacs.cpes.peachnet.edu/remmwww/remm/users.htm
Model Overview/Abstract
REMM has been developed as a tool that can help quantify the water quality benefits of riparian buffers. REMM
simulates the movement of surface and subsurface water; sediment transport and deposition; nutrient transport,
sequestration, and cycling; and vegetative growth in riparian forest systems on a daily timestep. In REMM, the
riparian system is considered to consist of three zones between the field and the waterbody. Each zone includes litter
and three soil layers, as well as a plant community that can have six plant types in two canopy levels. REMM can be
used to quantify nitrogen and phosphorus trapping in the riparian buffer zone, determine buffer effectiveness,
investigate the long-term fate of nutrients in buffer zones, evaluate the influence of vegetation type on buffer
effectiveness, and determine the impacts of harvesting on buffer effectiveness. As of July 2004, REMM is still under
development and has been continuously updated. A user interface is being built to assist input and output data
management. The strength of REMM is its capability to simulate subsurface compartments and comprehensive
nutrient cycling. Because of the model's complexity, application requires extensive data.
Model Features
• Surface and subsurface water movement
• Sediment and nutrients transport
• Nutrients cycling
• Vegetative growth
Model Areas Supported
Watershed High
Receiving Water None
Ecological Medium (Vegetative: High)
Air None
Groundwater Medium
304
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
In REMM, the riparian system consists of three zones between the field (drainage area) and the receiving waterbody.
Each zone includes litter and three soil layers that terminate at the bottom of the plant root system and a plant
community that can include six plant types in two canopy levels. Surface hydrology, erosion, vertical and horizontal
subsurface flows, carbon and nutrient dynamics, and plant growth that occur in each zone are modeled on a daily
timestep.
Scientific Detail
Movement and storage of water within riparian buffer systems is simulated by a process-based, two-dimensional
water balance on a daily basis. Infiltration is estimated using the explicit form of modified Green-Ampt equation.
Surface runoff is assumed to be generated when the sum of rainfall and upslope runoff depth exceeds the infiltration
capacity, also when the top soil layer is saturated and cannot accept any additional water. REMM also includes
subsurface drainage in both the vertical and lateral directions. Vertical subsurface drainage is simulated as
gravitational drainage between horizons and as deep groundwater seepage from the lower layer. The upward
evapotransporation flux is simulated in REMM using Darcy Buckingham equation given by Skaggs in the presence
of a shallow water table. In the absence of a shallow water table, evapotransporation loss is limited to the soil layer
wilting point moisture content. A simple routing scheme is used to distribute the incoming upland runoff down the
riparian slope based on its depth and flow velocity.
Overland soil erosion is simulated using USLE. Sediment routing is performed using equations applied in the
AGNPS model.
Carbon dynamics are simulated using the Century model, which divides carbon in the soil and residue (litter) layers
into different pools. Soil nitrogen is modeled in organic forms associated with soil carbon, residue carbon, and
dissolved carbon and in inorganic forms as ammonium and nitrate. Nitrification, denitrification, and immobilization
of nitrogen from plant residues are simulated. REMM simulates soil phosphorous in both organic and inorganic
forms. Partitioning of phosphorous into dissolved and adsorbed fractions is computed using the Langmuir isotherm.
Soil temperature is simulated in REMM using an empirical approach for the soil surface and a heat flux approach for
the subsurface soil. Vegetation can be represented in REMM using up to 12 plant types in two canopies. A thorough
model simulating the photosynthesis, respiration, growth, and development of the vegetation is applied in REMM.
Consumption of water and nutrients are related to the increase in biomass of various vegetation types.
Model Framework
• Vegetative buffer system
• Three buffer zones
• Litter and three soil layers
• Plant community (up to six plant types in two canopy levels)
Scale
Spatial Scale
• Field/hill slope
Temporal Scale
• Daily timestep
Assumptions
The model assumes that the dynamics of each physical, chemical and biological component can be described by the
principal of conservation of mass.
305
-------
Appendix A: Model Fact Sheets
Model Strengths
• Capability for simulating subsurface compartment
• Comprehensive nutrients cycling.
Model Limitations
• Extensive data requirement
• Employs a simplified method to distribute the incoming upland runoff down the riparian slope based on its
depth and flow velocity, which limits the accuracy of flow routing and, hence, infiltration calculation
Application History
REMM is still in the development stage, with a testing version released. Testing is being performed using data from
riparian buffer sites located on Gibbs Farm in Tifton, Georgia.
Model Evaluation
Testing is being performed using data from riparian buffer sites located on Gibbs Farm in Tifton, Georgia.
Model Inputs
• Daily weather data
o daily precipitation amount (mm)
o duration of precipitation (hr)
o ratio of time to rainfall peak/rainfall duration
o ratio of max. rainfall intensity/average rainfall intensity
o max. daily temperature (deg C)
o min. daily temperature (deg C)
o daily solar radiation (langleys/day)
o wind velocity (m/s)
o wind direction (deg from north)
o dew point temperature (deg C)
• Daily field input data
o surface runoff depth (mm/ha)
o subsurface depth (mm/ha)
o sediment loading (kg/ha)
o sediment-clay fraction
o sediment-silt fraction
o sediment-small aggregate fraction
o sediment-large aggregate fraction
o sediment-sand fraction
o carbon-humus-active-surface runoff (kg/ha)
o CN-ratio-surface runoff
o CP-ratio-surface runoff
o carbon-humus-active-suburface flow (kg/ha)
o CN-ratio subsurface flow
o CP-ratio subsurface flow
o carbon-humus-active-sediment (kg/ha)
o CN-ratio sediment
o CP-ratio sediment
o ammonium-surface runoff (kg/ha)
o ammonium-subsurface flow (kg/ha)
o ammonium-sediment (kg/ha)
o nitrate-surface runoff (kg/ha)
o nitrate-subsurface flow (kg/ha)
o phosphorus-surface runoff (kg/ha)
306
-------
Appendix A: Model Fact Sheets
o phosphorus-subsurface flow (kg/ha)
o phosphorus-sediment (kg/ha)
o rainfall-carbon-humus-active (kg/mm-ha)
o rainfall-CN ratio
o rainfall-CP ratio
o rainfall-nitrate (kg/mm-ha)
o rainfall-ammonium (kg/mm-ha)
o rainfall-phosphorus (kg/mm-ha)
• Zone parameters:
o field surface drainage area (ha)
o field subsurface drainage area (ha)
o field length (m)
o stream depth (m)
o latitude of location
o zone length (m)
o zone width (m)
o zone slope (%)
o seepage from aquiclude (mm/day)
o number of surface channels
• Litter layer parameters:
o layer depth (cm)
o evaporation factor
o evaporation constant
o litter transmission factor
o litter moisture (mm)
o litter humus moisture holding capacity by weight (%)
o litter residue moisture holding capacity weight (%)
o litter bulk density (g/cm3)
o litter CaCo3 (g/kg)
o litter P group
o litter base saturation (%)
o ammonium adsorption coefficients
o ammonium absorption coefficients
o litter pH
o litter C structural pool (kg/ha)
o litter C metabolic pool (kg/ha)
o litter C active pool (kg/ha)
o litter C slow pool (kg/ha)
o litter C passive pool (kg/ha)
o litter C lignin (kg/ha)
o litter ammonium pool (kg/ha)
o litter nitrate pool (kg/ha)
o litter N structural pool (kg/ha)
o litter N metabolic pool (kg/ha)
o litter N active pool (kg/ha)
o litter N slow pool (kg/ha)
o litter N passive pool (kg/ha)
o litter P structural pool (kg/ha)
o litter P metabolic pool (kg/ha)
o litter P active pool (kg/ha)
o litter P slow pool (kg/ha)
o litter P passive pool (kg/ha)
o litter P labile inorganic pool (kg/ha)
o litter P active inorganic pool (kg/ha)
307
-------
Appendix A: Model Fact Sheets
o litter P stable inorganic pool (kg/ha)
• Soil layer parameters (repeat for 3 soil layers):
o rock density (g/cm3)
o rock fraction (g/g)
o pore size distribution index
o bubbling pressure head (cm)
o soil layer depth (cm)
o wilting point (cm/cm)
o field capacity (cm/cm)
o porosity (cm/cm)
o initial moisture content (cm/cm)
o saturated conductivity (cm/hr)
o sand content (%)
o silt content (%)
o clay content (%)
o bulk density (g/(cm)3)
o CaCO3 content
o base saturation
o initial carbon structural pool (kg/ha)
o initial carbon metabolic pool (kg/ha)
o initial carbon active pool (kg/ha)
o initial carbon slow pool (kg/ha)
o initial carbon passive pool (kg/ha)
o initial carbon lignin pool (kg/ha)
o initial nitrogen ammonium pool (kg/ha)
o initial nitrogen nitrate pool (kg/ha)
o initial nitrogen structural pool (kg/ha)
o initial nitrogen metabolic pool (kg/ha)
o initial nitrogen active pool (kg/ha)
o initial nitrogen slow pool (kg/ha)
o initial nitrogen passive pool (kg/ha)
o initial phosphorus structural pool (kg/ha)
o initial phosphorus metabolic pool (kg/ha)
o initial phosphorus active pool (kg/ha)
o initial phosphorus slow pool (kg/ha)
o initial phosphorus passive pool (kg/ha)
o initial inorganic phosphorus labile pool (kg/ha)
o initial inorganic phosphorus active pool (kg/ha)
o initial inorganic phosphorus stable pool (kg/ha)
Users' Guide
Available online: http://sacs.cpes.peachnet.edu/remmwww/remm/documents/Userguide.pdf
Hard copy of technical document is available upon request.
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• Windows
308
-------
Appendix A: Model Fact Sheets
Programming language:
• C++
Runtime estimates:
• Seconds to minutes
Linkages Supported
None
Related Systems
None
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• A user interface is under development
References
U.S. Department of Agriculture, Agricultural Research Service (USDA-ARS). Riparian Ecosystem Management
Model: User's manual. USDA-ARS Southeast Watershed Research Laboratory. Tifton, GA.
.
Altier, L.S., R. Lowrance, R.G. Williams, S. P. Inamdar, D. D. Bosch, J. M. Sheridan, R. K. Hubbard, and D. L.
Thomas. 2002. Riparian Ecosystem Management Model - Simulator for Ecological Processes in Riparian Zones.
Report 46. U.S. Department of Agriculture, Agricultural Research Service, Conservation Research.
Lowrance, R., L.S. Altier, R.G. Williams, S.P. Inamdar, D.D. Bosch, J.M. Sheridian, D.L. Thomas and R.K.
Hubbard. 1998. The Riparian Ecosystem Management Model: Simulator for ecological processes in riparian zones.
In Proceedings of the First Federal Interagency Hydrologic Modeling Conference, Las Vegas, NV, April 1998, pp.
1.81-1.88.
Inamdar, S.P., J.M. Sheridan, R.G. Williams, D.D. Bosch, R. Lowrance, L.S. Altier, D.L. Thomas. 1998. The
Riparian Ecosystem Management Model: Evaluation of the hydrology component. In Proceedings of the First
Federal Interagency Hydrologic Modeling Conference, Las Vegas, NV, April 1998, pp. 7.17-7.24.
Bosch, D.D., R.G. Williams, S.P. Inamdar, J.M. Sheridan, and R. Lowrance. 1998. Erosion and sediment transport
through riparian forest buffers. In Proceedings of the First Federal Interagency Hydrologic Modeling Conference,
Las Vegas, NV, April 1998, pp. 3.31-3.38.
Inamdar, S.P., L.S. Altier, R. Lowrance, R.G. Williams, R. Hubbard. 1998. The Riparian Ecosystem Management
Model: Nutrient Dynamics. In Proceedings of the First Federal Interagency Hydrologic Modeling Conference, Las
Vegas, NV, April 1998, pp. 1.73-1.80.
L.S. Altier, R.G. Williams, R. Lowrance, and S.P. Inamdar. 1998. The Riparian Ecosystem Management Model:
Plant growth component. In Proceedings of the First Federal Interagency Hydrologic Modeling Conference, Las
Vegas, NV, April 1998, pp. 1.33-1.40.
Williams, R.G., R. Lowrance, L.S. Altier, and S.P. Inamdar. 1998. The Riparian Ecosystem Management Model: A
demonstration. In Proceedings of the First Federal Interagency Hydrologic Modeling Conference, Las Vegas, NV,
April 1998, pp. 8.133-8.138.
309
-------
Appendix A: Model Fact Sheets
RMA-11
Contact Information
Ian P. King
Resource Modelling Associates
9 Dumaresq Street
Gordon
NSW 2072, Australia
#61-2-9498-1043
ikingrma(@,bigpond. net, au
Download Information
Availability: Proprietary
Cost:
RMA-11 (One-dimensional/Two-dimensional) $1250
RMA-11 (One-dimensional/Two-dimensional/Three-dimensional) $2500
RMA-2 (One-dimensional/Two-dimensional hydrodynamics) $1500
RMAGEN (Pre-processor graphics) $750
RMAPLT (Post-process graphics) $750
Model Overview/Abstract
RMA-11 is a finite element water quality model for simulation of one-, two-, or three-dimensional estuaries, bays,
lakes, and rivers. It is also capable of simulating one- and two-dimensional approximations to systems. It is designed
to accept velocity and depth input from the results of the two-dimensional hydrodynamic model or the three-
dimensional stratified flow model. The input hydrodynamic data are used in the solution of the advection-diffusion
constituent transport equations. The model operates independently of the timesteps in the hydrodynamic model, and
the input data are automatically interpolated.
Model Features
• Finite element water quality model.
• Constituents that may be included in the simulation are
o Temperature with a full atmospheric heat budget at the water surface
o Biological oxygen demand/dissolved oxygen
o The nitrogen cycle
o The phosphorous cycle
o Algae growth and decay
o Cohesive suspended sediment
o Non-cohesive suspended sediment such as sand
o Salinity
o Coliform
Model Areas Supported
Watershed None
Receiving Water High
Ecological None
310
-------
Appendix A: Model Fact Sheets
Air None
Groundwater None
Model Capabilities
Conceptual Basis
RMA-11 represents the latest in the RMA model development series. The capabilities of all the earlier models, in
particular RMA4Q, have been systematically incorporated into the latest version. In addition, the advection-
diffusion-settling capabilities have been expanded to permit fully three-dimensional simulation. This model is
capable of one-, two-, three-dimensional approximations in any combination. It also has a bed model capable of
tracking the evolution of cohesive or noncohesive sediments.
Scientific Detail
The two- and three-dimensional advection diffusion equations are simulated for conservative and decaying
constituents. The equations are solved by the finite element method. The prototype system is represented by a
network of triangles and quadrilaterals/cubes and prisms that may have curved sides if desired. Within each element,
the model uses quadratic approximations for water quality constituents. A fully implicit solution scheme is used for
solution of time dependent problems. The primary features of RMA-11 are as follows:
• RMA11 shares many of the same capabilities of the RMA-2/RMA-10 hydrodynamics models, including
irregular boundary configurations, variable element size, one-dimensional elements, and the wetting and
drying of shallow portions of the modeled region.
• RMA 11 may be executed in steady state or dynamic mode. The velocities supplied may be constant or
interpolated from an input file (this may be RMA-2 or RMA-10 output).
• Source pollutant loads may be input to the system either at discrete points, over elements, or as fixed
boundary values.
• In formulating the element equations, the element coordinate system is realigned with the local flow
direction. This permits the longitudinal and transverse diffusion terms to be separated, with the net effect
being to limit excessive constituent dispersion in the direction transverse to flow.
• For increased computational efficiency, up to fifteen constituents may be modeled at one time, each with
separately defined loading, decay, and initial conditions.
• The model may be used to simulate temperature with a full heat exchange with the atmosphere, nitrogen
and phosphorous nutrient cycles, biochemical oxygen demand/dissolved oxygen, algae, cohesive or
noncohesive suspended sediments, and other nonconservative constituents.
• A multilayer bed model for the cohesive sediment transport constituent keeps track of thickness and
consolidation of each layer.
Model Framework
RMA-11 is a finite element water quality model for simulation of one-, two-, or three-dimensional estuaries, bays,
lakes, and rivers.
Scale
Spatial Scale
• Operation unit one-, two-, or three-dimensional
Temporal Scale
• User-defined timestep
Assumptions
• Represents the waterbody in a finite element model
311
-------
Appendix A: Model Fact Sheets
Model Strengths
• Simulates temperature with a full heat exchange with the atmosphere, nitrogen and phosphorous nutrient
cycles, biochemical oxygen demand/dissolved oxygen, algae, cohesive or noncohesive suspended
sediments, and other nonconservative constituents
• Has a pre-and post-processor (RMAGEN and RMAPLT)
Model Limitations
• Uses finite element solution.
Application History
It has recently been used by Australian Water and Coastal Studies for studies of sand transport and power station
effluents in Lake Macquarie and bacteria and nutrient loadings and dispersion in Salt Pan Creek, Sydney. It has also
been used for the waters surrounding Hong Kong and is currently being used for a study of Moreton Bay near
Brisbane.
The models have been applied in numerous coastal systems, estuaries, and rivers around the world. Resource
Management Associates (RMA) in California has made numerous outfall location studies in San Francisco Bay,
including studies for East Bay Municipal Utility District, City of San Francisco, and Tri Valley Waste Management
District. The U.S. Army Corps of Engineers has used the models extensively in its sedimentation studies for coastal
estuary systems. Examples include the Columbia River in Oregon and the Atchafalaya Estuary in Louisiana. RMA-
4, along with earlier version of RMA-2, was selected by Brigham Young University for inclusion in its
commercially marketed FASTTABS (now known as SMS) system.
Model Evaluation
See literature.
Model Inputs
• Initial conditions
• Time sequences of boundary conditions (inputs from watershed sources and discharges)
• Geometry of waterbody
• Physical coefficients
• Biological and chemical reaction rates
Users' Guide
Available with purchase of model
Technical Hardware/Software Requirements
Computer hardware:
• PC-DOS or UNIX
Operating system:
• PC-DOS or UNIX
Programming language:
• FORTRAN
312
-------
Appendix A: Model Fact Sheets
• Used on IBM PCs, UNIX workstations, Dec VAX systems, and Cray super computers. Source and
executable versions of the models are available for the finite element models.
Runtime estimates:
• Minutes to hours
Linkages Supported
RMA-2 and RMA-10
Related Systems
RMA-l, RMAGEN, RMA11PR, RMA4QPR, RMAPLT, RMA-2, RMA-10, CONVRM4 and CONVRM4Q
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• RMA-l and RMAGEN are pre-processor programs used to aid construction and display of finite element
networks.
• RMA-2 simulation results generate current vectors and stage for input into RMA-l 1.
• RMAPLT is a graphical postprocessor program for development of (1) velocity vector plots, (2) contour
plots of constituent concentration, water surface elevation, or velocity magnitude, and (3) time histories of
these parameters for selected locations.
References
King, I. P. 1998. RMA-11 - A Three Dimensional Finite Element Model For Water Quality in Estuaries and
Streams. Department of Civil and Environmental Engineering, University of California, Davis, CA.
313
-------
Appendix A: Model Fact Sheets
SED2D
Contact Information
Joseph V. Letter, Jr.
U.S. Army Corps of Engineers
Engineer Research and Development Center
Waterways Experiment Station
Coastal and Hydraulics Laboratory
Hydrologic Systems Branch
3 909 Halls Ferry Road
Vicksburg,MS39180
(601)634-2730
tabs(@,hl.wes.army.mil
http://chl.wes.army.mil/software/tabs
Download Information
Availability: Proprietary
Cost: Contact distributor
Organizations other than the U.S. Corps of Engineers can obtain the model from the vendors listed below; however,
WES cannot provide support nor offer guarantees of suitability to users outside the Corps of Engineers.
• Resource Management Associates (RMA), 4171 Suisun Valley Rd, Suite C, Suisun, CA (USA) 94585. Phone
707-864-2950
• Brigham Young University (BYU), Engineering Computer Graphics Laboratory, 368B CB, Provo, UT (USA)
84602. Phone 801-378-2812. A company named BOSS Corporation handles their distribution (1-800-488-
4775).
Model Overview/Abstract
SED2D, formerly STUDH, is a two-dimensional numerical model for depth-averaged transport of cohesive or a
representative grain size of noncohesive sediments and their deposition, erosion, and formation of bed deposits.
Model Features
• Sediment transport
• Deposition
• Erosion
Model Areas Supported
Watershed None
Receiving Water High
Ecological None
Air None
Groundwater None
314
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
SED2D provides a high level of sophistication in its representation of cohesive sediment bed dynamics and water
column exchange. The SED2D model can execute simultaneously with or externally linked with the RMA-2 model.
Scientific Detail
The derivation of the basic finite element formulation is presented in Ariathurai (1974) and Ariathurai, Mac Arthur,
and Krone (1977) and is summarized below. The model does four major computations:
1. Convection-Diffusion Governing Equation
2. Bed Shear Stress Calculation
3. The Bed Source/Sink Term
4. The Bed Strata Discretization
The model simulates cohesive and noncohesive sediment transport. Horizontal water column transport includes
advection and shear dispersion. Cohesive sediment transport is represented by a single size class with concentration-
dependent settling to approximately account for flocculation or aggregation and disaggregation. Water column-bed
sediment and sorbed contaminant exchange includes bottom-stress dependent-deposition and bottom stress and bed-
shear-strength-dependent erosion or resuspension. Cohesive bed erosion by mass erosion and surface erosion
processes are represented. Cohesive sediment beds are represented by a time-varying number of layers that increase
and decrease in number during periods of deposition and resuspension, respectively. Although the model does not
include a mathematically formulated consolidation simulation, the thickness, void ratio, density, and shear strength
of the layers vary with time since deposition, through the use of experimentally determined relationships. Vertical
advection of sediment and sorbed material in the bed is implicitly represented by the dynamic bed layering process.
Model Framework
SED2D WES requires that hydrodynamic data be externally supplied, usually by a numerical hydrodynamic model.
The TABS-MD modeling system has been designed to satisfy this and other needs for a comprehensive modeling
package. TABS-MD consists of RMA-2 WES, a general-purpose program for hydrodynamic modeling, in addition
to SED2D WES. The graphical user interface, SMS, and a number of utility programs are used to develop input,
translate data, analyze output, and provide graphical output from the models.
Scale
Spatial Scale
• Two-dimensional operation unit
Temporal Scale
• User-defined timestep
Assumptions
• The model considers a single, effective grain size during each simulation.
• An implicit assumption of the SED2D WES model is that the changes in the bed elevation due to erosion
and/or deposition do not significantly affect the flow field.
• The sediment transport model formulation assumes that the input geometric mesh and the resulting
hydrodynamic solution from RMA-2 are of adequate resolution, accuracy, and quality to allow for an
accurate and reasonable solution to the governing sediment transport equation to be solved.
Model Strengths
• The model simulates cohesive and noncohesive sediment transport.
315
-------
Appendix A: Model Fact Sheets
• It is useful for both deposition and erosion studies and, to a limited extent, for stream width studies.
• It has the ability to compute sediment loadings and bed elevation changes.
Model Limitations
• Both clay and sand may be analyzed, but the model considers a single, effective grain size during each
simulation. Therefore, a separate model run is required for each effective grain size.
• Fall velocity must be prescribed along with the water surface elevations, x-velocity, y-velocity, diffusion
coefficients bed density, critical shear stresses for erosion, erosion rate constants, and critical shear stress
for deposition.
• The program does not compute water surface elevations or velocities; these data must be provided from an
external calculation of the flow field. For most problems, a numerical model for hydrodynamic
computations, RMA-2 WES, is used to generate the water surface elevations and velocities.
• In addition, the sediment transport model formulation assumes that the input geometric mesh and the
resulting hydrodynamic solution from RMA-2 are of adequate resolution, accuracy, and quality to allow for
an accurate and reasonable solution to the governing sediment transport equation to be solved. In such a
case, then the mathematical solution from SED2D will potentially have severe oscillations with negative
concentrations.
Application History
Refer to References
Model Evaluation
See publications.
Model Inputs
• Initial conditions
• Time sequences of boundary conditions
• Geometry mesh
• Physical coefficients
Users' Guide
Available online: http://chl.erdc.usace.army.mil/CHL.aspx?p=s&a=ARTICLES:483
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS
Programming language:
• FORTRAN
Runtime estimates:
• Minutes to hours
316
-------
Appendix A: Model Fact Sheets
Linkages Supported
RMA-2, FastTABS
Related Systems
SED3D
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• A computer program (developed at the U.S. Army Corps of Engineers' Waterways Experiment Station
[WES] and Brigham Young University [BYU]), called FastTABS, provides a graphical, point and click
means for performing pre- and post-processing for surface water numerical models and can be used to
process data for SED2D.
References
Hayter, E.J., M.A. Bergs, R. Gu, S.C. McCutcheon, and S. J. Smith. 1996. SED2D, A Finite Element Model for
Cohesive Sediment Transport. Prepared for U. S. Environmental Protection Agency by Clemson University,
Clemson, SC.
Ackers, P., and W. R. White. 1973. Sediment Transport: New approach and analysis. Journal of the Hydraulics
Division, American Society of Civil Engineers, no. HY11.
Ariathurai, R. 1974. A Finite Element Model for Sediment Transport in Estuaries. Ph. D. thesis, University of
California, Davis.
Ariathurai, R., R.C. MacArthur, and R.B. Krone. 1977. Mathematical Model of Estuarial Sediment Transport.
Technical Report D-77-12. U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
Krone, R.B. 1962. Flume Studies of the Transport of Sediment in Estuarial Shoaling Processes, Final Report.
Hydraulic Engineering Laboratory and Sanitary Engineering Research Laboratory, University of California,
Berkeley.
McAnally, W.H., and W.A. Thoma,. 1980. Shear Stress Computations in a Numerical Model for Estuarine Sediment
Transport. Memorandum for Record, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
Swart, D.H. 1976. Coastal Sediment Transport, Computation of Longshore Transport. Report 968, Part 1. Delft
Hydraulics Laboratory, The Netherlands.
White, W.R., H. Milli, and A.D. Crabbe. 1975. Sediment Transport Theories: An Appraisal of Available Methods.
Report Int. 119. Hydraulics Research Station, Wallingford, England, vols. 1-2.
317
-------
Appendix A: Model Fact Sheets
SED3D: Three-Dimensional Numerical Model of Hydrodynamics and
Sediment Transport in Lakes and Estuaries
Contact Information
Steven McCutcheon
U.S. Environmental Protection Agency
Office of Research and Development (ORD)
National Exposure Research Lab
Ecosystems Research Division
Center for Exposure Assessment Modeling (CEAM)
960 College Station Road
Athens, Georgia 30605-2700
(706) 355-8235 or (706)355-8400
mccutcheon.steven(g)epamail.epa.gov or ceanngjepamail. epa.gov
http ://www. epa. gov/ceampubl/swater/sed3 d/
Download Information
Availability: Nonproprietary
Cost: N/A
Model Overview/Abstract
SED3D simulates the flow and sediment transport in lakes, estuaries, harbors, and costal waters. SED3D is a
dynamic modeling system that can be used to simulate the flow and sediment transport in various waterbodies under
the forcing of winds, tides, freshwater inflows, and density gradients with the influence of the Coriolis acceleration,
complex bathymetry, and shoreline geometry.
Model Features
• Hydrodynamics
• Sediment
• Transport
• Lakes
• Estuaries
• Harbors
• Coastal waters
• Three-dimensional
Model Areas Supported
Watershed None
Receiving Water High
Ecological Low
Air None
Groundwater None
318
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
SED3D simulates the flow and sediment transport in lakes, estuaries, harbors, and costal waters. SED3D is a
dynamic modeling system that can be used to simulate the flow and sediment transport in various waterbodies under
the forcing of winds, tides, freshwater inflows, and density gradients with the influence of the Coriolis acceleration,
complex bathymetry, and shoreline geometry.
Scientific Detail
Given proper boundary and initial conditions, the model can compute the time-dependent, three-dimensional
velocity components (u,v,w), temperature (T), salinity (S), and suspended sediment concentration (c) in the
Cartesian and vertically stretched grid system (x,y,s). The model contains a free surface, as opposed to a rigid-lid,
with proper boundary conditions for velocity, temperature, salinity, and sediment. A simplified second-order closure
model of turbulent transport is used to compute the vertical eddy viscosity and diffusivity contained in the model
equations.
Model Framework
SED3D can be run in a three-dimensional mode, a two-dimensional vertically integrated 'x-y' mode, or a two-
dimensional 'x-z' mode.
Scale
Spatial Scale
• Three-dimensional operation unit
Temporal Scale
• User-defined timestep
Assumptions
To be determined
Model Strengths
• Can simulate the flow and sediment transport in lakes, estuaries, harbors, and coastal waters.
• Provides a three dimensional numerical grid and a quasi-second-order closure scheme and sediment
transport capabilities.
Model Limitations
• The SED3D model and its associated files are designed for Digital Equipment Corporation (DEC)
installation, compilation, link edit, and execution. This model has not been successfully compiled, linked,
or executed at the EPA CEAM using a DOS-based FORTRAN development tool
• Complicated
Application History
To be determined
319
-------
Appendix A: Model Fact Sheets
Model Evaluation
To be determined
Model Inputs
• Initial conditions
• Time sequences of boundary conditions
• Reservoir geometry
• Physical coefficients
• Time sequences of hydrometeorological conditions
Users' Guide
Document must be requested from contact:
Sheng, Y.P., D.E. Eliason, X.J. Chen, and J.-K. Choi. 1991 A Three-Dimensional Numerical Model of
Hydrodynamics and Sediment Transport in Lakes and Estuaries: Theory, Model Development and Documentation.
U.S. Environmental Protection Agency, Athens GA.
Technical Hardware/Software Requirements
Computer hardware:
• DEC VAX 6310
Operating system:
• DEC VAX VMS version 5.3-1
Programming language:
• DEC VAX FORTRAN version 5.5-98
• VAX VMS/DCL LINK version V05-05
Runtime estimates:
• Minutes to hours or longer depending on the complexity of the system
Linkages Supported
None
Related Systems
None
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
Not available
320
-------
Appendix A: Model Fact Sheets
References
Sheng, Y.P., D.E. Eliason, X.-J. Chen, and J.-K. Choi. 1991. A Three-Dimensional Numerical Model of
Hydrodynamics and Sediment Transport in Lakes and Estuaries: Theory, Model Development and Documentation.
U.S. Environmental Protection Agency, Athens GA.
321
-------
Appendix A: Model Fact Sheets
SHETRAN
Contact Information
Professor P.E. O'Connell
Professor of Water Resources Engineering
Department of Civil Engineering
University of Newcastle
Newcastle upon Tyne
NE1 7RU, UK
(44) (0)191222 6319
p.e.o'connell(@,ncl.ac.uk
http://www.ncl.ac.uk/wrgi/wrsrl/rbms/rb ms.html#SHETRAN
Download Information
This information could not be obtained from either an Internet or recent publication search.
Model Overview/Abstract
The SHETRAN system was developed by the Water Resources Systems Research Laboratory (WRSRL), is based
on the SHE (Systeme Hydrologique Europeen), and was developed by international collaboration between groups in
the United Kingdom, Denmark, and France. SHETRAN is a three-dimensional, coupled surface/subsurface,
physically-based, spatially-distributed, finite-difference model for coupled water flow, multifraction sediment
transport and multiple, reactive solute transport in river basins. It gives a detailed description in time and space of
the flow and transport in the basin, which can be visualized using animated graphical computer displays. This makes
it a powerful tool for use in studying the environmental impacts of land erosion, pollution, and the effects of changes
in land use and climate and in studying surface water and groundwater resources and management. SHETRAN is
currently being integrated in a decision-support system to maximize its usefulness in environmental impact
management.
Model Features
The main difference between SHETRAN and existing physically based, spatially distributed, river basin modeling
systems lies in its comprehensive nature and its capabilities for modeling subsurface flow and transport. The
subsurface is treated as a variably saturated heterogeneous porous medium and fully three-dimensional flow and
transport can be simulated for combinations of confined, unconfmed, and perched systems. The "unsaturated zone"
is modeled as an integral part of the subsurface, and subsurface flow and transport are coupled directly to surface
flow and transport. So, for example, it is possible to model flow and transport in "deep" groundwater, while at the
same time modeling flow and transport in complex near-surface regions, which respond rapidly to rainfall and
strongly affect recharge and surface runoff.
SHETRAN represents physical processes using physical laws applied on a three-dimensional finite-difference mesh.
The mesh follows the topography of the basin, and the parameters of the physical laws vary from point to point on
the mesh, thus allowing the representation of the spatial heterogeneity of the physical properties of the rocks, soils,
and vegetation cover, etc. SHETRAN can be used for basins of less than 1 km2 to 2500 km2 in area and typically
uses a mesh with 20,000 finite-difference cells, stacked 50 deep, to model hourly flow and transport for periods of
up to a few decades. Stream channels are simulated as a network of links, each link running along the edge of a
finite-difference cell. The results from SHETRAN simulations can be viewed and analyzed using the SHEGRAPH
dedicated graphics package.
322
-------
Appendix A: Model Fact Sheets
Model Areas Supported
Watershed Medium
Receiving Water High
Ecological Low
Air None
Groundwater High
Model Capabilities
The main difference between SHETRAN and existing physically based, spatially distributed, river basin modeling
systems lies in its comprehensive nature and its capabilities for modeling subsurface flow and transport. The
subsurface is treated as a variably-saturated heterogeneous porous medium and fully three-dimensional flow and
transport can be simulated for combinations of confined, unconfmed, and perched systems. The "unsaturated zone"
is modeled as an integral part of the subsurface, and subsurface flow and transport are coupled directly to surface
flow and transport. So, for example, it is possible to model flow and transport in "deep" groundwater, while at the
same time modeling flow and transport in complex near-surface regions, which respond rapidly to rainfall and
strongly affect recharge and surface runoff.
Scale
Spatial Scale
• Physically based, spatially distributed
Temporal Scale
• User-defined, variable timestep.
Assumptions
Represents physical processes using physical laws applied on a three-dimensional finite-difference mesh. The mesh
follows the topography of the basin, and the parameters of the physical laws vary from point to point on the mesh,
thus allowing the representation of the spatial heterogeneity of the physical properties of the rocks, soils, and
vegetation cover, etc. Can be used for basins 1 km2 to 2500 km2 in area, and typically uses a mesh with 20,000
finite-difference cells, stacked 50 deep, to model hourly flow and transport for periods of up to a few decades.
Simulates stream channels as a network of links, each link running along the edge of a finite-difference cell.
Model Strengths and Limitations
This information could not be obtained from either an Internet or recent publication search.
Model Limitations
This information could not be obtained from either an Internet or recent publication search.
Application History
Examples of SHETRAN validation and application studies include
• Validation for water flow and the effect of fire on erosion, Rimbaud basin, France.
• Validation for nitrate transport, Slapton Wood basin, United Kingdom.
• Validation for flow and conservative and nonconservative solute transport through complex Quaternary
deposits, Cumbria, United Kingdom.
• Validation for flow and solute transport in a perched system at Hazelrigg, Lancaster, United Kingdom.
• Validation for erosion at Draix badlands gully basins, France.
• Validation of snowmelt modeling, Upper Sheep Creek, Idaho, USA.
323
-------
Appendix A: Model Fact Sheets
• Study of climate change impacts for Mediterranean desertification: Cobres (Portugal), Mula (Spain), and
Agri (Italy) basins.
• Impact assessment for radionuclide transport in the near surface and surface regions following releases
from a deep underground repository for radioactive wastes, United Kingdom.
Model Evaluation
This information could not be obtained from either an Internet or recent publication search.
Model Inputs
This information could not be obtained from either an Internet or recent publication search.
Users' Guide
This information could not be obtained from either an Internet or recent publication search.
Technical Hardware/Software Requirements
Computer hardware:
This information could not be obtained from either an Internet or recent publication search.
Operating system:
This information could not be obtained from either an Internet or recent publication search.
Programming language:
This information could not be obtained from either an Internet or recent publication search.
Runtime estimates:
This information could not be obtained from either an Internet or recent publication search.
Linkages Supported
This information could not be obtained from either an Internet or recent publication search.
Related Systems
This information could not be obtained from either an Internet or recent publication search.
Sensitivity/Uncertainty/Calibration
This information could not be obtained from either an Internet or recent publication search.
Model Interface Capabilities
SHETRAN models results can be viewed and analyzed using the SHEGRAPH dedicated graphics package.
References
Water Resources Group, Dept. of Civil Engineering, University of Newcastle, Tyne, UK.
.
Water Resource Systems Research Laboratory (WRSRL), Dept. of Civil Engineering at the University of Newcastle,
Tyne, UK. .
324
-------
Appendix A: Model Fact Sheets
SLAMM: Source Loading and Management Model
Contact Information
Robert E. Pitt
Professor, Director of Environmental Engineering Programs
The University of Alabama
Department of Civil and Environmental Engineering
Box 870205
Tuscaloosa, AL 35487-0205
(205) 348-2684
rpitt(@,coe.eng.ua.edu
http://unix.eng.ua.edu/~rpitt/SLAMMDETPOND/WinSlammMainWINSLAMM book.html
Download Information
Availability: Proprietary, http://www.winslamm.com/
Cost: $200
Model Overview/Abstract
SLAMM was originally developed to better understand the relationships between sources of urban pollutants and
runoff quality (Pitt, 1993). SLAMM is strongly based on field observations, with minimal reliance on pure
theoretical processes that have not been adequately documented or confirmed in the field. The EPA's Nationwide
Urban Runoff Program (NURP) has contributed significantly to the development of empirical relationships used in
SLAMM. SLAMM now also includes a wide variety of source area and outfall control practices (e.g., infiltration
practices, wet detention ponds, porous pavement, street cleaning, catch basin cleaning, and grass swales). Beginning
with version 5, SLAMM is Windows-based and thus is called WinSLAMM.
The model performs continuous mass balances for paniculate and dissolved pollutants and for runoff volumes.
Runoff is calculated by a method developed by Pitt (1987) for small-storm hydrology. Runoff is based on rainfall
minus initial abstraction, and infiltration is calculated for both impervious and pervious areas. Triangular
hydrographs, parameterized by a statistical approach, are used to simulate flow. Exponential buildup and washoff, as
well as wind removal functions, are used in computing runoff pollutant loadings. Water and sediment from various
source areas are tracked as they are routed through treatment devices. SLAMM is mostly used as a planning tool to
better understand sources of urban runoff pollutants and the effectiveness of their control.
SLAMM is capable of considering many stormwater controls (affecting source areas, drainage systems, and outfalls)
for a long series of rainfall events. The program considers how particulates filter or settle out in control devices.
Paniculate removal is calculated based on the structural design characteristics. Storage and overflow of devices are
also considered. At the outfall locations, the characteristics of the source areas are used to determine pollutant loads
in solid and dissolved phases. Another of its abilities is to accurately describe a drainage area in sufficient detail for
water quality investigations but without requiring a great deal of superfluous information that field studies have
shown to be of little value in accurately predicting discharge results. SLAMM also applies stochastic analysis
procedures to more accurately represent actual uncertainty in model input parameters to better predict the actual
range of outfall conditions (especially pollutant concentrations). Like all stormwater models, SLAMM needs to be
accurately calibrated and then tested (verified) as part of any local stormwater management effort.
325
-------
Appendix A: Model Fact Sheets
Model Features
• Urban stormwater runoff and water quality empirical simulation model: initial abstraction, infiltration, and
pollutant buildup and washoff
• Stormwater treatment devices (BMPs) simulation
Model Areas Supported
Watershed Medium
Receiving Water None
Ecological None
Air None
Groundwater None
Model Capabilities
Conceptual Basis
SLAMM represents the urban catchment by unconnected areas, connected areas, and drainage system. Unconnected
areas drain to adjacent pervious areas before the runoff enters drainage system. Connected areas directly drain to
drainage system. The drainage system consists of curbs, gutters, swale ditches, treatment devices, manholes, and
sewerage. SLAMM routes stormwater runoff and pollutants from unconnected source areas to the drainage system
directly, or to adjacent connected or pervious areas, which drain to the drainage system.
Scientific Detail
The model performs continuous mass balances for paniculate and dissolved pollutants and for runoff volumes.
Runoff is calculated by a method developed by Pitt (1987) for small-storm hydrology, which empirically determines
the initial losses and infiltration loss based on experiment data. Runoff is based on rainfall minus initial abstraction,
and infiltration is calculated for both impervious and pervious areas. Triangular hydrographs, parameterized by a
statistical approach, are used to simulate flow. Exponential buildup and washoff and wind removal functions, are
used in computing runoff pollutant loadings. The characteristics of the source areas are used to determine pollutant
loads in solid and dissolved phases based on an empirical method derived using available field observations. The
pollutant removal effectiveness of treatment devices are estimated, also based on empirical equations derived from
field data. SLAMM also applies stochastic analysis procedures to more accurately represent uncertainty in model
input parameters to better predict the range of outfall conditions (especially pollutant concentrations). SLAMM
applies Monte Carlo sampling procedures to consider the uncertainties in model input values
Model Framework
• Urban areas (impervious, pervious)
• Stormwater treatment devices
Scale
Spatial Scale
• Site
• Catchment
Temporal Scale
• Variable timestep (hourly or sub-hourly)
Assumptions
• Triangular runoff hydrograph
• Exponential pollutant buildup and washoff
326
-------
Appendix A: Model Fact Sheets
Model Strengths
• Better representation of small storms
• Treatment devices (BMPs) simulation capabilities
• Suitable for site-scale evaluation of urban runoff pollutants and the effectiveness of their control without
requiring a great deal of information
• Uncertainty analysis capability
Model Limitations
• Does not have channel routing capability; limited use at the watershed scale
• Does not simulate base flow
• Is strongly based on a statistical approach that uses the current available field observations; is not a process-
based model
Application History
SLAMM is mostly used as a planning tool to better understand sources of urban runoff pollutants and the
effectiveness of their control. Early users of SLAMM include the Ontario Ministry of the Environment's Toronto
Area Watershed Management Strategy (TAWMS) study and the Wisconsin Department of Natural Resources'
Priority Watershed Program.
Model Evaluation
The SLAMM hydrology simulation component was verified by Pitt using field data from three sites in Wisconsin
and one site in Michigan.
Model Inputs
• Rainfall depth
• Site characteristics (area, land use type, surface condition, soil type, and infiltration rate)
• Drainage system characteristics (type of drainage system, density, underlying soil type, and infiltration
rate)
• Treatment devices (type, size, outlet structure, underlying soil type, and infiltration rate; if applicable, street
cleaning date and/or frequency, and wet pond natural seepage and evaporation rate)
Users' Guide
Available online: http://unix.eng.ua.edu/~rpitt/SLAMMDETPOND/WinSlamm/MainWrNSLAMM book.html
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• DOS and Windows
Programming language:
• VB
327
-------
Appendix A: Model Fact Sheets
Runtime estimates:
• Seconds to minutes
Linkages Supported
The WinSLAMM-SWMM Interface Program has been developed to allow WinSLAMM to provide the runoff and
pollutant loads from input to SWMM TRANSPORT or EXTRAN blocks.
Related Systems
WinSLAMM
Sensitivity/Uncertainty/Calibration
SLAMM applies Monte Carlo sampling procedures to consider the uncertainties in model input values and enable
the model output to be expressed in probabilistic terms.
Model Interface Capabilities
• A Windows-based interface to facilitate data input
• Output is summarized in a series of user selectable tables
References
Pitt, R., and J. Voorhees. 2000. The Source Loading and Management Model (SLAMM), a Water Quality
Management Planning Model for Urban Stormwater Runoff. University of Alabama, Department of Civil and
Environmental Engineering, Tuscaloosa, AL.
Pitt, R. 1987. Small Storm Urban Flow and Particulate Washoff Contributions to Outfall Discharges. Ph.D. diss.,
University of Wisconsin, Madison, WI.
Pitt, R. 1993. Source Loading and Management Model (SLAMM). Presented at the National Conference on Urban
Runoff Management, March 30-April 2. Chicago, IL.
328
-------
Appendix A: Model Fact Sheets
SPARROW: SPAtially Referenced Regression On Watershed Attributes
Contact Information
Richard Alexander
U.S. Geological Survey - NAWQA Nutrient Synthesis
413 National Center
12201 Sunrise Valley Drive
Reston, VA20192
(703) 648-6869
ralex@usgs.gov
http://water.usgs.gov/nawqa/sparrow/
Other Contacts:
NATIONAL SPARROW (NAWQA Nutrient Synthesis, Reston, VA)
Nolan, Jacquie 703-648-5709 (jnolan@usgs.gov)
Schwarz, Gregory 703-648-5718 (gschwarz@usgs.gov)
Smith, Richard 703-648-6870 (rsmithl@usgs.gov)
REGIONAL SPARROW
Chesapeake Bay (Maryland District)
Brakebill, John 410-238-4257 (jwbrakeb@usgs.gov)
Preston, Steve 410-267-9875 (spreston@usgs.gov)
New England (New Hampshire District)
Johnston, Craig 603-226-7843 (cmjohnst@usgs.gov)
Moore, Richard 603-226-7825 (rmoore@usgs.gov)
Robinson, Keith 603-226-7809 (kwrobins@usgs.gov)
North Carolina Coastal (North Carolina District)
McMahon, Gerard 919-571-4068 (gmcmahon@.usgs.gov)
Download Information
Availability: Nonproprietary
Cost: N/A
Model Overview/Abstract
SPARROW relates in-stream water quality measurements to spatially referenced characteristics of watersheds,
including contaminant sources and factors influencing terrestrial and stream transport. The model empirically
estimates the origin and fate of contaminants in streams and quantifies uncertainties in these estimates based on
model coefficient error and unexplained variability in the observed data.
Model Features
• Empirically based method
• Riverine pollutant loading rates prediction
• Contaminants modeled: sediment, nutrients, etc.
• Datasets used: Reach File (RF1 or version), USGS's National Land Cover Dataset (NLCD), STATSGO,
and other spatial datasets
329
-------
Appendix A: Model Fact Sheets
Model Areas Supported
Watershed Medium
Receiving Water Low
Ecological None
Air Low (Model supports inclusion of atmospheric deposition loads)
Groundwater None
Model Capabilities
Conceptual Basis
The SPARROW model uses spatially referenced regressions of contaminant transport on watershed attributes to
support regional water quality assessment goals, including descriptions of spatial and temporal patterns in water
quality and identification of the factors and processes that influence those conditions. The method is designed to
reduce the problems of data interpretation caused by sparse sampling, network bias, and basin heterogeneity.
Scientific Detail
SPARROW uses statistical methods to calibrate a simple, structural model of riverine water quality, one that
imposes mass balance in accounting for changes in contaminant flux. Regression equations relate measured
transport rates in streams to spatially referenced descriptors of pollution sources and land surface and stream channel
characteristics. Spatial referencing of land-based and water-based variables is accomplished via superposition of a
set of contiguous land surface polygons on a digitized network of stream reaches that define surface water flow
paths for the region of interest. The primary spatial reference frame for the model is the RF1 reach network: all point
sources and landscape features are referenced to a particular RF1 reach.
Water quality measurements are obtained from monitoring stations located in a subset of the stream reaches. Water
quality predictors in the model are developed as a function of both reach and land surface attributes and include
quantities describing contaminant sources (point and nonpoint) as well as factors associated with rates of material
transport through the watershed (such as soil permeability and stream velocity). Predictor formulae describe the
transport of contaminant mass from specific sources to the downstream end of a specific reach. Loss of contaminant
mass occurs during both overland and in-stream transport. The model can also take into account pollutant loads
contributed by atmospheric deposition.
SPARROW was first used to estimate the distribution of nutrients in streams and rivers of the U.S. and subsequently
shown to describe land and stream processes affecting the delivery of nutrients (Smith, et al., 1997; Alexander, et
al., 2000; Preston and Brakebill 1999).
Model Framework
• Empirical, regression-based
• Uses national datasets, wide applicability
Scale
Spatial Scale
• Large watersheds
Temporal Scale
• Annual
• User-defined modeling period
330
-------
Appendix A: Model Fact Sheets
Assumptions
The model is based on an empirical regression approach using mass balance calculations. Regression equations
relate measured transport rates in streams to spatially referenced descriptors of pollution sources and land surface
and stream channel characteristics.
Model Strengths
The model is capable of simulating a variety of pollutants at different spatial scales using national level datasets,
including RF1 (stream reach file), NLCD (USGS land use/land cover), and STATSGO (NRCS soils data). The
model can be used to model large- and small-scale systems with flexibility in the datasets and level of detail
incorporated. The model is readily available from USGS and has been applied in several case studies.
Model Limitations
The model is limited to broadly estimating pollutant loads and fate/transport characteristics. Stream processes and
model output are based on statistical relationships that were developed using national and regional water quality
datasets.
Application History
The model has been primarily used to estimate nutrient and sediment loads at various spatial scales. Refer to the
USGS SPARROW website for case studies.
Model Evaluation
Refer to USGS website - http://water.usgs.gov/nawqa/sparrow/
Model Inputs
• Initial conditions
• Time sequences of boundary conditions (inputs from watershed sources and discharges)
• Stream reach file reference (e.g., RF1)
• Physical coefficients
• Biological and chemical reaction rates
• Land use, soils, and other spatial datasets
Users' Guide
Not readily available on website. Website provides several journal articles and contacts:
http://water.usgs.gov/nawqa/sparrow/
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS
Programming language:
• SAS
331
-------
Appendix A: Model Fact Sheets
Runtime estimates:
• Minutes
Linkages Supported
None
Related Systems
None
Sensitivity/Uncertainty/Calibration
In calibrating the model, measured rates of contaminant transport are regressed on predicted transport rates at the
locations of the monitoring stations, giving rise to a set of estimated linear and nonlinear coefficients from the
predictor formulae.
Once calibrated, the model is used to estimate contaminant transport and concentration in all stream reaches. A
variety of regional characterizations of water quality conditions are then possible based on statistical summarization
of reach-level estimates. The application of bootstrap techniques allows estimation of the uncertainty of model
coefficients and predictions.
Model Interface Capabilities
None
References
Refer to USGS website for complete list - http://water.usgs.gov/nawqa/sparrow/
Smith, R.A., G.E. Schwarz, and R.B. Alexander. 1997. Regional Interpretation of Water-quality Monitoring Data.
Water Resources Research, 33(12): 2781-2798.
Alexander, R.B., R.A. Smith, M.J. Focazio, and M.A. Horn. 1999. Source-Area Characteristics of Large Public
Surface-Water Supplies in the Conterminous United States: An Information Resource for Source-Water Assessment.
Open-File Report 99-248. U.S. Geological Survey, Reston, VA.
Preston, S.D. and J.W. Brakebill. 1999. Application of Spatially Referenced Regression Modeling for the Evaluation
of Total Nitrogen Loading in the Chesapeake Bay Watershed. Report 99-4054. U.S. Geological Survey Water
Resources Investigations, Baltimore, MD.
332
-------
Appendix A: Model Fact Sheets
STORM: Storage, Treatment, Overflow, Runoff Model
Contact Information
Mainframe version:
U.S. Army Corps of Engineers
Hydrologic Engineering Center (HEC)
609 Second Street
Davis, CA 95616
PC Version (ProStorm):
Dodson & Associates, Inc.
5629 FM 1960 West, Suite 314
Houston, Texas 77069
1-800-235-8069 or 281-440-3787
Download Information
Availability: Nonproprietary
Cost: N/A
No downloads available.
Model Overview/Abstract
The STORM model was developed by Water Resources Engineers, Inc., in 1973 under a contract with the U.S.
Army Corps of Engineers' Hydrologic Engineering Center (HEC). STORM is designed to model urban watersheds
and is capable of calculating loads and concentrations of water quality parameters, such as suspended and settleable
solids, biochemical oxygen demand, total nitrogen, orthophosphate, and total coliform. STORM is also capable of
calculating land surface erosion. STORM is used to aid in sizing of storage and treatment facilities to control the
quantity and quality of stormwater runoff and land surface erosion. A continuous simulation model, STORM
requires hourly precipitation data to model the seven stormwater elements, such as rainfall/snowmelt, runoff, dry
weather flow, pollutant accumulation and washoff, land surface erosion, treatment rates, and detention reservoir
storage. Dust and dirt and the associated pollutants are washed off from the watershed by the rainfall. The runoff is
routed to the treatment-storage facilities and the effect of treatment is calculated. Runoff in excess of treatment plant
capacity is stored and treated later except for the quantity in excess of storage, which is waste untreated and
becomes overflow directly into the receiving waters.
Model Features
• Urban watershed model
Model Areas Supported
Watershed High
Receiving Water None
Ecological None
Air None
Groundwater None
333
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
STORM is a quasi-dynamic model that uses modified rational formula for hydrology simulation. Erosion is
simulated using USLE, and water quality is simulated by linear buildup and washoff coefficients.
Scientific Detail
The runoff of water is computed by one of three methods — coefficient method, the SCS Curve Number technique,
or a combination of the two.
The model computes a soil moisture balance at the beginning of each time increment by the following equation:
where A = 0.7((SM -St_i)/SM)v
B = ((SM-St_l)ISM}p
where, S is the soil moisture capacity for storage of water (in), IN is the maximum infiltration rate from initial
abstractions (in/hr), EVis the pan evaporation rate (in/hr), MP is the maximum soil percolation rate (in/hr), SIM is the
maximum soil moisture capacity for storage of water (in), t is the time, At is time increment (1 hr), v is the exponent
regulating evapotranspiration, and;? is the exponent regulating percolation.
Dry weather flow in the combined sewer systems is computed by specifying either the total waste water flow and
infiltration flow (mgd), domestic, commercial, industrial, and infiltration flow separately or the coefficients required
to compute the individual flows based on population and areas of commercial and industrial land.
The STORM model calculates the pollutant loadings based on either dust and dirt method or the daily pollutant
accumulation method. The dust and dirt method calculates pollutants as fractions of the dust and dirt for each land
use. The amount of the dust and dirt is calculated based on accumulation rate specified in terms of pounds per 100
feet of gutter length per day for each land use. The factors such as the intensity of rainfall, the rate of runoff, the
accumulation of dust and dirt on the watershed, and the frequency and efficiency of street sweeping operations
affects the amount of pollutants entering the storm drains and the treatment facilities or the receiving waters.
Dry weather pollutant loading in the combined sewer systems is computed similarly to the flow calculation during
the same period by specifying either the total waste water flow and infiltration flow (mgd), domestic, commercial,
industrial, and infiltration flow separately or the coefficients required to compute the individual flows based on
population and areas of commercial and industrial land.
Model Framework
The model simulates runoff, erosion, and pollutant loading from urban watersheds, and routes through the treatment
storage facilities and discharges the excess runoff into receiving waters.
Scale
Spatial Scale
• Watershed scale
Temporal Scale
• Hourly
334
-------
Appendix A: Model Fact Sheets
Assumptions
• Pollutants accumulated over the land between the consecutive rainfall events will be washed off during a
rainfall event
Model Strengths
• Capable of calculating loads and concentration of many pollutants
• Simulates land surface erosion
• Can aid in sizing of storage and treatment facilities
Model Limitations
• Little flexibility in parameters to calibrate to observed hydrographs
• Requires a large amount of data
Application History
The STORM model was extensively used in the late 1970s and early 1980s. The model was applied to the San
Francisco master drainage plan for abatement of combined sewer overflows.
Model Evaluation
To be determined
Model Inputs
• Runoff coefficients
• SCS parameters
• Hourly precipitation
Users' Guide
Available upon request from U.S. Army Corps of Engineers
Technical Hardware/Software Requirements
Computer hardware:
• PC and Mainframe
Operating system:
• Windows and Mainframe
Programming language:
• FORTRAN
Runtime estimates:
• Minutes
Linkages Supported
None
335
-------
Appendix A: Model Fact Sheets
Related Systems
None
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Pro Storm
References
Donigian, A.S., Jr., and W. C. Huber. 1991. Modeling ofNonpoint Source Water Quality in Urban and Non-urban
Area. EPA/600/3-91/039. U.S. Environmental Protection Agency, Environmental Research Laboratory, Athens, GA.
U.S. Army Corps of Engineers, Hydrologic Engineering Center (USACE-HEC). 1977. Storage, Treatment,
Overflow, Runoff Model, STORM, Generalised Computer Program 723-58-L77520. USACE-HEC. Davis,
California.
Shoemaker, L., M. Lahlou, M. Bryer, D. Kumar, and K. Kratt. 1997. Compendium of Tools for Watershed
Assessment and TMDL Development. EPA 841/B/97/006. U.S. Environmental Protection Agency, Office of Water,
Washington, D.C.
336
-------
Appendix A: Model Fact Sheets
SWAT: Soil and Water Assessment Tool
Contact Information
Jeff Arnold
Grassland, Soil & Water Research Laboratory
Agricultural Research Service
U.S. Department of Agriculture
808 East Blackland Road
Temple, Texas 76502
(254) 770-6502
http://www.brc.tamus.edu/swat/index.html
Download Information
Availability: Nonproprietary, http://www.brc.tamus.edu/swat/swat2000.html
Cost: N/A
Model Overview/Abstract
SWAT is a river basin, or watershed, scale model developed by Dr. Jeff Arnold for the USDA Agricultural Research
Service (ARS). SWAT was developed to predict the impact of land management practices on water, sediment, and
agricultural chemical yields in large complex watersheds with varying soils, land use, and management conditions
over long periods of time. The model is physically based and computationally efficient; it uses readily available
inputs and allows users to study long-term impacts. SWAT is a continuous time model (i.e., a long-term yield
model). The model is not designed to simulate detailed, single-event flood routing.
Model Features
• Watershed hydrology, sediment, and water quality model
• Pesticide fate and transport simulation
• Channel erosion simulation
• Rural and agricultural management practices including detailed agricultural land planting, tillage,
irrigation, fertilization, grazing, and harvesting procedures
Model Areas Supported
Watershed High
Receiving Water Medium
Ecological Low
Air None
Groundwater Low
Model Capabilities
Conceptual Basis
SWAT divides a watershed into subwatersheds. Each subwatershed is connected through a stream channel and
further divided in to Hydrologic Response Unit (HRU). HRU is a unique combination of a soil and a vegetation type
in a subwatershed, and SWAT simulates hydrology, vegetation growth, and management practices at the HRU level.
337
-------
Appendix A: Model Fact Sheets
Water, nutrients, sediment, and other pollutants from each HRU are summarized in each subwatershed and then
routed through the stream network to the watershed outlet.
Scientific Detail
The following is a list of key processes and algorithms used in SWAT:
• Climate: Weather generator or user's input
• Hydrology: Canopy interception, runoff (SCS CN) infiltration (Green-Ampt), and evapotranspiration
(Penman-Monteith, Priestley-Taylor, or Hargreaves) flow
• Land cover/plant growth: Water and nutrient uptake, crop and plant growth database
• Erosion: MUSLE using peak runoff rate
• Nutrients: Nitrogen and phosphorus cycles
• Agricultural practices: Planting, tillage, irrigation, fertilization, pesticide management, grazing, and
harvesting. SWAT also has auto-fertilization and auto-irrigation as management options that are useful for
many agricultural areas
• For urban areas, SWAT employs buildup and washoff approach similar to approaches used in SWMM
• SWAT uses two methods to simulate water routing in stream network: Variable storage routing method
(flow continuity equation) and Muskingum routing method (wedge and prism storage)
• Sediment transport: Stream sediment transport power is related to the stream flow rate. Model calculates
the maximum sediment that can be transported, compares the maximum sediment concentration to the
actual sediment concentration, and extra sediment is deposited. If the actual sediment is less than the
maximum, settled sediment will be re-suspended and enter the water
• SWAT simulates in-stream biological and nutrient processes, including algal growth, death, and settling
and oxygen in water, aeration and photosynthesis, and changes in water temperature
• Temporal and spatial scale: Daily; watershed or flexible size
• SWAT simulates sediment settling and mass balance, pesticide transport and fate (reservoirs only), nutrient
(N and P) settling and lake chlorophyll a production (empirical) processes in ponds, wetlands, reservoirs,
and potholes.
Model Framework
• Hydrologic response unit, subwatershed, and watershed
• Simple one-dimensional stream and well mixed reservoir/lake model
Scale
Spatial Scale
• ID, cell or subwatershed
Temporal Scale
• Daily
Assumptions
• SCS CN approach (Green-Ampt approach for infiltration was also implemented but requires hourly data)
and MUSLE are appropriate for the area being modeled
• Flows in streams and reservoirs are one-dimensional
Model Strengths
• Physically based
• Great documentation
• Uses readily available inputs facilitated by the GIS interface (BASINS)
338
-------
Appendix A: Model Fact Sheets
• Detailed crop growth model and databases
• Good land management modules and databases
• Suitable for study watersheds from small to very large sizes
Model Limitations
• Not for simulating sub-daily events such as a single storm event and diurnal changes of dissolved oxygen in
a waterbody
• Only for simulating conservative metal species from the point source input
• Only route one pesticide each time through the stream network
• Cannot specify actual areas to apply fertilizers
• Assumes one-dimensional well mixed streams and reservoirs
• A large watershed can be divided into hundreds of HRUs resulting in many hundreds of input files, which
are difficult to manage and modify without a solid interface
• The current version does not have a good model post-processor
Application History
This model has been applied widely to study hydrology, nonpoint source pollution, and TMDLs since early 1990s.
See the reference list for details.
Model Evaluation
Many peer-review research papers have evaluated the model and model applications (see references)
Model Inputs
• Land uses (MRLC and others)
• Soil (STATSGO and others)
• Topography (30 x 30 m2 DEM or other resolutions)
• Subwatersheds (derived from manual or auto delineation tools in BASINS 3.0)
• Point source (PCS or other database)
• Climate data (daily temperature, precipitation, solar radiation, and wind speed)
• Crop and management databases
• USGS flow data (for calibration)
• Long-term watershed quality data (for model calibration)
• BASINS provides substantial input data and pre-processing to develop and run SWAT model
Users' Guide
The SWAT2000 Theoretical Documentation reviews all processes simulated with the model.
The SWAT2000 User's Manual provides definitions for all input variables.
Available online: http://www.brc.tamus.edu/swat/swatdoc.html
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS, Microsoft Windows
339
-------
Appendix A: Model Fact Sheets
Programming language:
• FORTRAN (model) and Arc View Avenue (interface)
Runtime estimates:
• Minutes to less than an hour
Linkages Supported
BASINS
Related Systems
SWAT incorporates features of several ARS models and is a direct outgrowth of the SWRRB model (Simulator for
Water Resources in Rural Basins). Specific models that contributed significantly to the development of SWAT were
CREAMS (Chemicals, Runoff, and Erosion from Agricultural Management Systems), GLEAMS (Groundwater
Loading Effects on Agricultural Management Systems), and EPIC 4 (Erosion-Productivity Impact Calculator).
Sensitivity/Uncertainty/Calibration
The calibration tool in the BASINS 3.0 interface allows basic model calibration and sensitivity analysis (27 key
parameters). No tools were developed for uncertainty analysis.
Model Interface Capabilities
• Non-GIS DOS interface (Util)
• Arc View interface—AVSWAT
• BASINS 3.0 Arc View GIS interface (includes input editing and output viewing utilities and a simple
calibration tool)
References
Neitsch S.L., J.G. Arnold, J.R. Kiniry, and J.R. Williams. 2001. Soil and Water Assessment Tool User's Manual,
Version 2000. U.S. Department of Agriculture, Agricultural Research Service, Temple, TX.
.
Arnold, J.G., R. Srinivasan, R.S. Muttiah, P.M. Allen and C. Walker. 1999. Continental scale simulation of the
hydrologic balance. Journal of the American Water Resources Association. 35(5):1037-1052.
Arnold, J.G., R. Srinivasan, R.S. Muttiah, and J.R. Williams. 1998. Large area hydrologic modeling and assessment
Parti: Model development. Journal of American Water Resources Association. 34(l):73-89.
Arnold, J.G., J.R. Williams, and D.A. Maidment. 1995. A continuous time water and sediment routing model for
large basins. Journal of Hydraulic Engineering, American Society of Civil Engineers. 121 (2): 171 -183.
Bingner, R.L., J. Garbrecht, J.G. Arnold, and R. Srinivasan. 1997. Effect of watershed subdivision on simulated
runoff and fine sediment yield. Transactions of the American Society of Agricultural Engineers. 40(5): 1329-1335.
Deliman, Patrick N. et al., 1999. Review of Watershed Water Quality Models. Technical Report W-99-1. U.S. Army
Corps of Engineers.
DiLuzio, M., R. Srinivasan, and J.G. Arnold. 2002. Integration of watershed tools and SWAT model into BASINS.
Journal of American Water Resources Association. 38(4):1127-1141.
Engel, B.A., R. Srinivasan, J.G. Arnold, C. Rewerts, and S.J. Brown. 1993. Nonpoint source (NFS) pollution
modeling using models integrated with Geographic Information Systems (GIS). Water Science Technology. 28: 685-
690.
340
-------
Appendix A: Model Fact Sheets
Fontaine, T.A., T.S. Cruickshank, J.G. Arnold and R.H. Hotchkiss. 2002. Development of a snowfall-snowmelt
routine for mountainous terrain for the Soil Water Assessment Tool (SWAT). Journal of Hydrology. 262 (1-4):209-
223.
King, K.W., J.G. Arnold, and R.L. Bingner. 1999. Comparison of Green-Ampt and Curve Number methods on
Goodwin Creek Watershed using SWAT. Transactions of the American Society of Agricultural Engineers.
42(4):919-925.
Kirsh, K., A. Kirsh, and J.G. Arnold. 2002. Predicting sediment and phosphorus loads in the Rock River Basin using
SWAT. Transactions of the American Society of Agricultural Engineers. 45(6): 1757-1769.
Manguerra, H.B., and B.A. Engel. 1998. Hydrologic parameterization of watersheds for runoff prediction using
SWAT. Journal of the American Water Resources Association. 34(5): 1149-1162.
Peterson, J.R. and J.M. Hamlet. 1998. Hydrologic calibration of the SWAT model in a watershed containing
fragipan soils. Journal of American Water Resources Association. 34(3): 531-544.
Saleh, A., J.G. Arnold, P.W. Gassman, L.W. Hauck, W.D. Rosenthal, J.R. Williams, and A.M.S. McFarland. 2000.
Application of SWAT for the Upper North Bosque Watershed. Transactions of the American Society of Agricultural
Engineers. 43(5): 1077-1087.
Santhi, C., J.G. Arnold, J.R. Williams, W.A. Dugas, and L. Hauck. 2001. Validation of the SWAT model on a large
river basin with point and nonpoint sources. Journal of American Water Resources Association. 37(5): 1169-1188.
Santhi, C., J.G. Arnold, J.R. Williams, W.A. Dugas, and L. Hauck. 2002. Application of a watershed model to
evaluate management effects on point and nonpoint source pollution. Transactions of the American Society of
Agricultural Engineers. 44(6): 1559-1570.
Spruill, C. A., S.R. Workman, and J.L. Taraba. 2000. Simulation of daily and monthly stream discharge from small
watersheds using the SWAT model. Transactions of the American Society of Agricultural Engineers. 43 (6): 1431-
1439.
Srinivasan, R. and J.G. Arnold. 1994. Integration of a basin-scale water quality model with GIS. Water Resources
Bulletin. (30)3:453-462.
Srinivasan, R.S., J.G. Arnold, and C.A. Jones. 1998. Hydrologic modeling of the United States with the soil and
water assessment tool. Water Resources Development. 14(3):315-325.
Srinivasan, R., T.S. Ramanarayanan, J.G. Arnold, and S.T, Bednarz. 1998. Large area hydrologic modeling and
assessment Part II: Model Application. Journal of American Water Resources Association. 34(1):91-101.
Van L., M.W. and J. Garbrecht. 2003. Hydrologic simulation of the little Washita River Experimental Watershed
using SWAT. Journal of American Water Resources Association. 39(2):413-426.
Ward, G. H., Jr. and J. Benaman. 1999. Models for TMDL Application in Texas Watercourses: Screen and Model
Review. Report CRWR-99-7. Center for Research in Water Resources, University of Texas, Austin.
341
-------
Appendix A: Model Fact Sheets
SWMM: Storm Water Management Model
Contact Information
SWMM 4.4 and previous versions:
Wayne C. Huber
Oregon State University
Dept. of Civil, Construction, and Environmental Engineering
202 Apperson Hall
Corvallis, Oregon 97331-2302
(541)737-4934
wayne.huber(@,orst.edu
SWMM version 5:
Lewis Rossman
U.S Environmental Protection Agency
Office of Research and Development
Water Supply and Water Resources Division
26 West Martin Luther King Drive
Cincinnati, OH 45268
(513)569-7603
rossman. lewis(g),epa. gov
Download Information
Availability: Nonproprietary
SWMM 4.4: http://ccee.oregonstate.edu/swmm
SWMM 5 (available for beta testing): http://www.epa.gov/ednnrmrl/swmm/index.htm
Cost: N/A
Model Overview/Abstract
SWMM is a dynamic rainfall-runoff simulation model developed by EPA. It is applied primarily to urban areas and
for single-event or long-term (continuous) simulation using various timesteps (Huber and Dickinson, 1988). It was
developed for the analysis of surface runoff and flow routing through complex urban sewer systems. The latest
official version of SWMM is 4.4h, which is recommended for all users. EPA SWMM5 is a completely revised and
updated release of SWMM. The first beta test version SWMM5 was released in June 2003. However, SWMM5 is
still under development, with additional functions being incorporated and released over time. In SWMM, flow
routing is performed for surface and sub-surface conveyance and groundwater systems, including the options of
nonlinear reservoir channel routing and fully dynamic hydraulic flow routing. In the fully dynamic hydraulic flow
routing option, SWMM simulates backwater, surcharging, pressure flow, and looped connections. SWMM has a
variety of options for quality simulation, including traditional buildup and washoff formulation as well as rating
curves and regression techniques. Universal Soil Loss Equation (USLE) is included to simulate soil erosion. SWMM
incorporates first order decay and particle settling mechanism in pollutant transport simulations and includes an
option of simple scour-deposition routine. Storage, treatment, and other BMPs can also be simulated.
Model Features
• Watershed hydrology and water quality
• Stream/conduit transport
342
-------
Appendix A: Model Fact Sheets
• Urban stormwater and sewage systems
Model Areas Supported
Watershed High
Receiving Water Medium
Ecological None
Air None
Groundwater Low
Model Capabilities
Conceptual Basis
The basic spatial unit for SWMM is the subcatchment into which the modeled watershed is subdivided. Multiple
small subwatersheds and representative streams may be networked together to represent a larger watershed drainage
area.
Scientific Detail
Infiltration is calculated using the Horton or Green-Ampt methods, at the user's choice. A version of Manning's
equation is used to estimate flow rate from the subcatchment area based on a conceptual model of the subcatchment
as a "nonlinear reservoir." The lumped storage scheme is applied for soil/groundwater modeling. For impervious
areas, a linear formulation is used to compute daily/hourly increases in particle accumulation. For pervious areas, a
modified USLE determines sediment load. The concept of potency factors is applied to simulate pollutants other
than sediment.
The Transport block has kinematic wave routing of flow and quality, base flow generation, and infiltration
capabilities and it routes flow through user-defined system ranges from natural channel to concrete pipes. The
EXTRAN block carries out a numerical solution of the complete St. Venant equations for urban drainageways and
conduits, by modeling the network as a link-node system. SWMM can be directly interfaced with EPA's WASP
receiving water quality model.
Model Framework
• Subwatersheds and watershed
• Channel/pipe network
• One-dimensional flow and pollutant routing
Scale
Spatial Scale
• Subwatershed of flexible size
Temporal Scale
• User-defined timestep, typically minutes to hourly
Assumptions
• The model performs best in urbanized areas with impervious drainage, although it has been widely used
elsewhere.
• Model parameters for quantity and quality simulations are developed such that the model will be calibrated
to enhance its capability.
• Water table elevation is assumed to be fixed.
• All the pollutants entering the waterbodies are sediment adsorbed.
343
-------
Appendix A: Model Fact Sheets
Model Strengths
• Fully dynamic hydraulic routing
• Hydraulic structure (manhole, weir, orifice, etc.) simulation
• Overland flow routing between pervious and impervious areas within a subcatchment (latest version)
• Various options for quality simulation, including buildup and washoff, rating curves, and regression
techniques
Model Limitations
• Only considers settling and first-order decay in in-stream pollutant routing and transformation
• Weak groundwater simulation capability
Application History
SWMM has been applied to urban hydrologic quantity/quality problems in scores of U.S. cities as well as
extensively in Canada, Europe, and Australia (Donigian and Huber, 1991; Huber, 1992). The model has been used
for very complex hydraulic analysis for combined sewer overflow mitigation, as well as for many stormwater
management planning studies and pollution abatement projects (Huber, 1992). Warwick and Tadepalli (1991)
describe calibration and verification of SWMM on a 10-square-mile urbanized watershed in Dallas, Texas.
Tsihrintzis, et al., (1995) describe SWMM applications to four watersheds in South Florida representing high- and
low-density residential, commercial, and highway land uses.
Model Evaluation
The applications are primarily limited to urban areas.
Model Inputs
• Data requirements for hydrologic simulation include area, imperviousness, slope, roughness, width,
depression storage, and infiltration parameters. Land use data are used to determine ground cover type for
each model subarea.
• Depending on what options are set for the loading calculations, additional parameters are necessary (e.g.,
buildup coefficients would be needed for the dry weather buildup simulation).
• Additional data are necessary if the user intends to model subsurface drainage and interflow.
• Depending on the stormwater system, dimensions, slope, roughness coefficients, elevations, and storage are
required.
• Continuous records of evapotranspiration, temperature, and solar intensity are required.
Users' Guide
• Huber, W.C. and R.E. Dickinson. 1988 Storm Water Management Model User's Manual, Version 4.
EPA/600/3-88/001a (NTIS PB88-236641/AS). U.S. Environmental Protection Agency, Athens, GA,
pp.595
• Roesner, L.A., J.A. Aldrich and R.E. Dickinson. 1988. Storm Water Management Model User's Manual,
Version 4: Addendum I, EXTRAN. EPA/600/3-88/00Ib (NTIS PB88236658/AS). U.S. Environmental
Protection Agency, Athens, GA. pp.203
• A revised and more readable User's Guide from William James at CHI can be purchased at
http://www.computationalhvdraulics.com/Publications/Books/r219.html
Cost: $85
344
-------
Appendix A: Model Fact Sheets
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• DOS and Windows
Programming language:
• FORTRAN (v4.4 and previous versions)
• C (v5)
Runtime estimates:
• Minutes
Linkages Supported
• SWMM can directly be interfaced with EPA's WASP receiving water quality model.
Related Systems
PCSWMM, XP-SWMM, MIKE-SWMM
Sensitivity/Uncertainty/Calibration
SWMM 4.4 includes a STATISTICS module, which performs simple statistical analyses on both quantity and
quality parameters.
Model Interface Capabilities
• SWMM 5 includes a Graphical User Interface for input data preparation and output data display
References
Donigian, A.S., Jr., and W.C. Huber. 1991. Modeling ofNonpoint Source Water Quality in Urban and Non-urban
Areas. EPA/600/3-91/039. U.S. Environmental Protection Agency, Environmental Research Laboratory, Athens,
GA.
Huber, W. C. 1992. Experience with the U.S. EPA SWMM Model for Analysis and Solution of Urban Drainage
Problems. In Proceedings, Inundaciones YRedes De Drenaje Urbano, ed. J. Dolz, M. Gomez, and J. P. Martin, eds.,
Colegio de Ingenieros de Caminos, Canales Y Puertos. Universitat Politecnica de Catalunya. Barcelona, Spain, pp.
199-220.
Huber, W.C. 2001. New Options for Overland Flow Routing in SWMM. American Society of Civil Engineers-
Environmental and Water Resources Institute, World Water and Environmental Congress, Orlando, FL.
Huber, W.C., and R.E. Dickinson. 1988. Storm Water Management Model Version 4, User's Manual. EPA 600/ 3-
88/ 00la (NTIS PB88-236641/ AS). U.S. Environmental Protection Agency, Athens, GA.
Irvine, K.N., E.G. Loganathan, E.J. Pratt and H.C. Sikka. 1993. Calibration of PCSWMM to estimate metals, PCBs
and HCB in CSOs from an industrial sewershed. In W. James, ed. New Techniques for Modeling the Management of
Stormwater Quality Impacts. Lewis Publishers, Boca Raton, FL. pp. 215-242.
345
-------
Appendix A: Model Fact Sheets
James, W., W. C. Huber, R. E. Pitt, R. E. Dickinson, and R. C. James. 2002. Water Systems Models [1]: Hydrology,
User's guide to SWMM4 RUNOFF and Supporting Modules and to PCSWMM. Version 2.4. Computational
Hydraulics International, Guelph, Ontario, Canada, pp. 311.
James, W., W. C. Huber, R. E. Pitt, R. E. Dickinson, L. A. Roesner, J. A. Aldrich, and R. C. James. 2002 Water
Systems Models [2]: Hydraulics, User's guide to SWMM4 TRANSPORT, EXTRAN and STORAGE Modules and to
PCSWMM. Version 2.4. Computational Hydraulics International. Guelph, Ontario, Canada, pp. 359.
Tshihrintzis, V. A., R. Hamid, and H. R. Fuentes. 1995. Calibration and verification of watershed quality model
SWMM in sub-tropical urban areas. In Proceedings of the First International Conference - Water Resources
Engineering. American Society of Civil Engineers, San Antonio, TX. pp 373-377.
Tsihrintzis, V. and R. Hamid. 1998. Runoff Quality Prediction from Small Urban Catchments using SWMM.
Hydrological Processes, 12 (2):311-329.
Warwick, J. J., and P. Tadepalli. 1991. Efficacy of SWMM application. Journal of Water Resources Planning and
Management 117(3):352-366.
346
-------
Appendix A: Model Fact Sheets
TMDL Modeling Toolbox
Contact Information
Tim Wool
U.S. Environmental Protection Agency
Office of Research and Development
National Exposure Research Laboratory
Ecosystems Research Division
Watershed and Water Quality Modeling Technical Support Center
960 College Station Road
Athens, GA 30605-2700
(706)355-8312
wool.tim(@,epa. gov
http://www.epa.gov/athens/wwqtsc
Download Information
Availability: Nonproprietary, http://www.epa.gov/athens/wwqtsc/index.html
Cost: N/A
Model Overview/Abstract
The TMDL Modeling Toolbox is a collection of models, modeling tools, and databases that have been utilized over
the past decade in the development of Total Maximum Daily Loads (TMDLs). The Toolbox takes these proven
technologies and provides the capability to more readily apply the models, analyze the results, and integrate
watershed loading models with receiving water applications. The design of the Toolbox is such that each of the
models is a stand-alone application. The toolbox provides an exchange of information between the models through
common linkages. Because of the modular design of the Toolbox, additional models can be added easily to integrate
with the other tools. In addition, the Toolbox provides the capability to visualize model results, a linkage to GIS and
non-geographic databases (including monitoring data for calibration), and the functionality to perform data
assessments.
Model Features
The Toolbox allows for the stead state dynamic simulation of mass transport and water quality processes in all types
of surface water environments, including overland flow, small creeks, rivers, lakes, estuaries, coastal embayments,
and offshore areas.
Model Areas Supported
Watershed High
Receiving Water High
Ecological None
Air Medium
Groundwater Medium
347
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
The Toolbox is designed to address a broad range of waterbody types and pollutants. EPA actively supports the
components of the TMDL Modeling Toolbox. EPA is committed to enhancing and improving components of the
Toolbox to meet the technical demands of the TMDL program and watershed protection. This will ensure
defensibility when TMDLs are faced with legal challenges. Additionally, knowledge gained through the TMDL
development experience and modeling with the Toolbox in one region of the country can be distributed readily
throughout the others regions.
Scientific Detail
The Toolbox contains assessment tools, watershed models, and receiving water models including the following:
Assessment Tools: (http://wcs.tetratech-ffx.com')
• Water Resources Database (WRDB)
• Watershed Characterization System (WCS)
• WCS Sediment Tool
• WCS Mercury Tool
• WCS LSPC Tool
Watershed Models:
• Loading Simulation Program in C++ (LSPC)
• Watershed Assessment Model (WAMView)
• Storm Water Management Model (SWMM)
Receiving Water Models:
• A Dynamic One-Dimensional Model of Hydrodynamics and Water Quality (EPDRivl)
• Stream Water Quality Model (QUAL2K)
• CONservational Channel Evolution and Pollutant Transport System (CONCEPTS)
• Environmental Fluid Dynamics Code (EFDC)
• Water Quality Analysis Simulation Program (WASP)
Model Framework
• Watershed hydrologic, sediment, and water quality
• Time series overland, subsurface, and in-stream simulation
Scale
Spatial Scale
• Watershed or Flexible size
• Lumped parameters at a land use-subwatershed basis
Temporal Scale
• Variety of timesteps, including hourly or daily depending on the model.
Assumptions
Each model and modeling tool in the TMDL Modeling Toolbox has its own assumptions based on the
physical/chemical processes modeled.
348
-------
Appendix A: Model Fact Sheets
Model Strengths
• Has standardized tools and models used by EPA, states, and consultants.
• Includes models with proven track records.
• Provides an exchange of information between the models through common linkages.
• Includes a flexible design for adding new models to integrate with the existing tools.
• Is in the public domain.
Model Limitations
• Requires a high level of expertise for each application.
• Each model and modeling tool has own limitation.
Application History
The Toolbox models and databases have been used both independently and collectively to develop defensible
TMDLs for a wide array of issues including pathogens, sediment, nutrients, dissolved oxygen, metals, temperature,
and toxicants. The WCS Sediment Tool has been applied to sediment-impaired waters throughout the southeast.
Mercury TMDLs were developed in Georgia using a combination of the WCS Mercury Tool and WASP. LSPC has
been used in Alabama for pathogen TMDLs; Georgia, Tennessee, Kentucky, and Alabama for nutrient and/or
dissolved oxygen TMDLs; and Alabama, West Virginia, and Arizona for metals TMDLs. EFDC has been used
widely throughout the country to support TMDL development—Washington, California, Oklahoma, Florida,
Mississippi, Alabama, North Carolina, West Virginia, Delaware, Pennsylvania, and Massachusetts.
Model Evaluation
Toolbox model linkages have been successful in a number of situations, most notably for TMDL development using
EFDC and WASP in the Neuse Estuary, North Carolina; Cape Fear River, North Carolina; and Fenholloway River
Estuary, Florida; and TMDL development using LSPC, EFDC, and WASP for Mobile Bay, Alabama; Flint Creek,
Alabama; Coosa Lakes, Alabama; Lake Allatoona, Georgia; and Alabama River, Alabama.
Model Inputs
The Toolbox provides an exchange of information between the models through common linkages. Each model has
its own specific input data requirement. In general, model inputs include
• Continuous meteorological time series records
• Soils data (auxiliary dataset to guide hydrologic calibration)
• Pollutant buildup and washoff
• Stream dimensions or rating curves
• Point source loading inputs
• A large number of specified calibration parameters
Users' Guide
Documentation is available online at http://www.epa.gov/athens/wwqtsc
Technical Hardware/Software Requirements
Computer hardware:
• IBM Compatible PC (Pentium III or higher recommended, but not required)
Operating system:
• Microsoft Windows 98/NT/2000/XP
349
-------
Appendix A: Model Fact Sheets
Programming language:
• Different tools use different programming platforms
Runtime estimates:
• Minutes to less than an hour and daily depending on the model.
Linkages Supported
• Loading Simulation Program in C++ (LSPC)
• Watershed Assessment Model (WAMView)
• Storm Water Management Model (SWMM)
• A Dynamic One-Dimensional Model of Hydrodynamics and Water Quality (EPDRivl)
• Stream Water Quality Model (QUAL2K)
• CONservational Channel Evolution and Pollutant Transport System (CONCEPTS)
• Environmental Fluid Dynamics Code (EFDC)
• Water Quality Analysis Simulation Program (WASP)
Related Systems
The TMDL Modeling Toolbox supports several numerical models as listed in the Linkage Supported section.
Sensitivity/Uncertainty/Calibration
The TMDL Modeling Toolbox supports several numerical models such as LSPC, which includes a data analysis
component that may be used to quickly compare model output against observed data in time series form, as monthly
summaries, or on a one-to-one graph. LSPC model output is specially tailored for spreadsheet use; consequently,
many users prefer to develop independent spreadsheet analysis, summarization, calibration, and plotting
applications, which are readily linked to LSPC model output.
Model Interface Capabilities
The Model Visualization Enhancement Module (MOVEM), a graphical post processor, provides an efficient method
for reviewing model predictions and comparing them with field data for calibration. MOVEM has the ability to
display results from all of the WASP models as well as EFDC, DYNHYD, and EPD-RIV1. MOVEM allows the
modeler to displays the results in two graphical formats:
• Spatial Grid - a two dimensional rendition of the model network is displayed in a window where the model
network is color shaded based upon the predicted concentration.
• x/y Plots - generates an x/y line plot of predicted and/or observed model results in a window.
There is no limit on the number of x/y plots, spatial grids, or even model result files the user can utilize in a session.
Separate windows are created for each spatial grid or x/y plot created by the user.
References
Tetra Tech, Inc. and U.S. Environmental Protection Agency (USEPA). 2002. The Loading Simulation Program in
C++ (LSPC) Watershed Modeling System—Users' Manual.
WAMView User's Manual. Soil and Water Engineering Technology, Inc.
Wool, T.A, R.B. Ambrose, J.L. Martin, and E.A. Comer. Water Quality Analysis Simulation Program (WASP)
Version 6.0 Draft: User's Manual.
350
-------
Appendix A: Model Fact Sheets
TOPMODEL
Contact Information
Keith Beven
Department of Environmental Science
Institute of Environmental and Natural Sciences
Lancaster University
Lancaster LAI 4YQ
United Kingdom
+44 (0)1524 593892
K.Beven@lancaster.ac.uk
http://www.es.lancs.ac.uk/hfdg/TOPMODEL.html
Download Information
Availability: Nonproprietary, (http://www.es.lancs.ac.uk/hfdg/TOPMODEL.html')
Cost: None for non-commercial uses.
Contact the author for other uses.
Model Overview/Abstract
TOPMODEL is a physically based, distributed watershed model that simulates hydrologic fluxes of water
(infiltration-excess overland flow, saturation overland flow, infiltration, exfiltration, subsurface flow,
evapotranspiration, and channel routing) through a watershed. The model simulates explicit groundwater/surface-
water interactions by predicting the movement of the water table, which determines where saturated land-surface
areas develop and have the potential to produce saturation overland flow.
Model Features
• Rainfall-runoff modeling in single or multiple subwatersheds.
• The Windows version of the model allows Monte Carlo runs with parameter sets chosen from specified
parameter ranges.
• The Windows version of the model displays simulated hydrograph time series—the topographic index
derived from the elevation data—and map of saturated area in a watershed.
Model Areas Supported
Watershed Medium
Receiving Water Low
Ecological None
Air None
Groundwater Medium
Model Capabilities
Conceptual Basis
TOPMODEL is a rainfall-runoff model that bases its distributed predictions on an analysis of watershed topography.
The model predicts saturation excess and infiltration excess surface runoff and subsurface stormflow. Since the first
article was published about the model in 1979 (Beven and Kirkby, 1979) there have been many different versions.
351
-------
Appendix A: Model Fact Sheets
The idea has always been that the model should be simple enough to be modified by the user so that the predictions
conform as far as possible to the user's perceptions of how a watershed works.
Scientific Detail
TOPMODEL is defined as a variable contributing area conceptual model in which the dynamics of surface and
subsurface saturated areas is estimated on the basis of storage discharge relationships established from a simplified
steady state theory for downslope saturated zone flows. The theory assumes that the local hydraulic gradient is equal
to the local surface slope and implies that all points with the same value of the topographic index, a/tan B will
respond in a hydrologically similar way. This index is derived from the basin topography, where a is the drained
area per unit contour length and tan B is the slope of the ground surface at the location. Thus the model need make
calculations only for representative values of the index. The results may then be mapped back into space by
knowledge of the pattern of the index derived from a topographic analysis.
The soil profile is defined by a set of stores. The upper one is the root zone storage, where rainfall infiltrates until
the field capacity is reached. When forest canopies appear, an additional interception and surface storage may be
necessary. In this store, evapotranspiration is assumed to take place at the potential rate to decrease at a linear rate
when the root zone becomes depleted.
Once the field capacity is exceeded, a second store starts filling until the water content reaches saturation. The
gravity drainage store links the unsaturated and saturated zones, according to a linear function that includes a time
delay parameter for vertical routing through the unsaturated zone. An alternative approach based on the Darcian flux
at the base of the unsaturated zone may be considered.
When the deficit in the gravity drainage store or the water table depth equals 0 the saturation condition is reached
and the rainfall produces direct surface runoff. Hence the main goal of TOPMODEL is the computation of the
storage deficit or the water table depth at any location for every timestep. The theory relates mean watershed storage
deficit to local storage deficits using the local value of a function of the topographic index. In the original version of
TOPMODEL the soil hydraulic conductivity, or by extension the soil transmissivity, is assumed to decay following
a negative exponential law. In this case, the expression that estimates the value of the local storage deficit or the
water table depth is given in terms of the topographic index ln(a/tan B). Other forms of soil hydraulic conductivity
decay functions lead to different index functions. When distributed values of soil transmissivity (TO) are known a
soil-topographic index may be considered, ln(a/TO tan B).
The topographic index derivation was obtained by manual analysis of contour maps and hillslope streamtubes in the
early versions of the model. The current version of the model provides a program to derive its distribution from a
regular raster grid of elevations for any watershed or subwatershed using the multiple direction flow algorithm and
the channel initiation threshold for positioning river headwaters.
To compute runoff according to the infiltration excess mechanism TOPMODEL uses the exponential Green-Ampt
model. If infiltration excess does occur it does so over the whole area of the subwatershed (although alternatively a
statistical distribution of hydraulic conductivity values in the watershed can be assumed). A parameter for
controlling the fraction of watershed area that generates runoff by infiltration excess was considered recently by a
few studies to compute runoff using the Philip' two term-equation.
Subwatershed discharges are routed to the watershed outlet using a linear routing algorithm with constant velocity
both in the main channel and in the internal subwatershed.
Model Framework
• Watershed and subwatersheds
• Watershed surface are divided into surface zone, root zone, and saturated zone.
• Channel networks
352
-------
Appendix A: Model Fact Sheets
Scale
Spatial Scale
• Grid or subwatersheds
Temporal Scale
• Variable, from 1 to 24 hours
Assumptions
• The hydraulic gradient of subsurface flow is equal to the land-surface slope.
• The actual lateral discharge is proportional to the specific watershed area (drainage area per unit length of
contour line).
• The redistribution of water within the subsurface can be approximated by a series of consecutive steady
states.
• The soil profile at each point has a finite capacity to transport water laterally downslope.
Model Strengths
• It is a simple distributed watershed model and results can be visualized in a spatial context.
• It requires few watershed parameters and low level of expertise.
• It has been studied extensively.
• The model code is available for modification.
Model Limitations
• TOPMODEL only simulates watershed hydrology, although studies have been conducted to modify it to
simulate water quality dynamics.
• TOPMODEL can be applied most accurately to watersheds that do not suffer from excessively long dry
periods and have shallow homogeneous soils and moderate topography.
• Model results are sensitive to grid size, and grid size <=50 m is recommended.
Application History
TOPMODEL has a long history of application. See an extensive list of publications at
http://www.es.lancs.ac.uk/hfdg/TOPMODEL/new-bibliog.html.
Model Evaluation
The model has been validated with rainfall-discharge data (e.g. Beven et al. 1984, Hornberger et al. 1985, Robson et
al. 1993, Obled et al. 1994, Wolock 1995) and several studies have examined its applicability to water quality
problems (Wolock et al. 1990, Robson etal. 1992).
Model Inputs
• Project file: Text description of application and input file names and paths.
• Catchment (watershed) data file: Watershed and sub watershed topographic index—In(a/tan5) distributions
and the following parameters:
o The mean soil surface transmissivity
o A transmissivity profile decay coefficient
o A root zone storage capacity
o An unsaturated zone time delay
o A main channel routing velocity and internal subwatershed routing velocity
To use the infiltration excess mechanism, a hydraulic conductivity (or distribution), a wetting front suction and
the initial near surface water content should be added.
353
-------
Appendix A: Model Fact Sheets
The initialization of each run requires an initial stream discharge and the root zone deficit.
• Hydrological input data file: rainfall, potential evapotranspiration, and observed discharge time series in
m/h
• Topographic index map data file: the topographic index map may be prepared from a raster digital
elevation file using the DTM-ANALYSIS program. This file includes number of pixels in X direction,
number of pixels in Y direction, grid size, and topographic index values for each pair of X and Y.
Users' Guide
Available online: http://www.es.lanes.ac.uk/hfdg/TOPMODEL.html
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS, PC-WINDOWS
Programming language:
• FORTRAN, Visual Basic
Runtime estimates:
• Minutes
Linkages Supported
Links to GLUE (Generalized Likelihood Uncertainty Estimation) program for sensitivity/uncertainty/calibration
analyses.
Related Systems
TOPMODEL is integrated in GRASS GIS version 5. TOPSIMPL, another Windows version of the model written by
Georges-Marie Saulnier can be downloaded directly from the main TOPMODEL site
http://www.es.lancs.ac.uk/hfdg/TOPMODEL.html.
Sensitivity/Uncertainty/Calibration
The Windows version of TOPMODEL allows the sensitivity analysis of the objective functions to changes of one or
more of the parameters to be explored. An initial run of the model is made with the current values of the parameters.
Then each chosen parameter is varied across its range, keeping the values of the other parameters constant. The
results are displayed as graphs.
TOPMODEL's Monte-Carlo simulation output can be exported for further sensitivity and uncertainty analyses on
the model results using the GLUE (Generalized Likelihood Uncertainty Estimation) program.
TOPMODEL calibration procedures are relatively simple because it uses very few parameters in the model
formulas. The model is very sensitive to changes of the soil hydraulic conductivity decay parameter, the soil
transmissivity at saturation, the root zone storage capacity, and the channel routing velocity in larger watersheds.
The calibrated values of parameters are also related to the grid size used in the digital terrain analysis. The timestep
and the grid size also have been shown to influence TOPMODEL simulations.
354
-------
Appendix A: Model Fact Sheets
Model Interface Capabilities
There are three options available in the program interface:
• The Hydrograph Prediction Option: This option allows the model to be run and hydrographs displayed. If a
Topographic Index Map File is available, then a map button is displayed that allows the display of
predicted simulation, either as a summary over all timesteps or animated.
• The Sensitivity Analysis Option: This screen allows the sensitivity of the objective functions to changes of
one or more of the parameters to be explored.
• The Monte Carlo Analysis Option: In this option a large number of runs of the model can be made using
uniform random samples of the parameters chosen for inclusion in the analysis. Check boxes can be used to
choose the variables and objective functions to be saved for each run. The results file produced will be
compatible with the GLUE analysis software package.
References
Beven, K J and M J. Kirkby. 1979. A physically based variable contributing area model of basin hydrology.
Hydrologic Science Bulletin. 24(l):43-69.
Beven, K.J., M.J. Kirkby, N. Schofield, and A.F. Tagg. 1984. Testing a physically-based flood forecasting model
(TOPMODEL) for three U.K. Catchments. Journal of Hydrology. 69:119-143.
Hornberger, G.M., KJ. Beven, BJ. Cosby, andD.E. Sappington. 1985. Shenandoah watershed study: Calibration of
a topography-based, variable contributing area hydrological model to a small forested catchment. Water Resources
Research. 21:1841-1850.
Obled, Ch., J. Wendling, and KJ. Beven. 1994. The sensitivity of hydrological models to spatial rainfall patterns:
An evaluation using observed data. Journal of Hydrology. 159: 305-333.
Robson, A.J., KJ. Beven, and C. Neal. 1992. Towards identifying sources of subsurface flow: A comparison of
components identified by a physically based runoff model and those determined by mixing techniques. Hydrological
Processes. 6:199-214.
Robson, A.J., P.O. Whitehead, and R.C. Johnson. 1993. An application of a physically based semi-distributed model
to the Balquhidder Catchments. Journal of Hydrology. 145:357-370.
Wolock, D.M. 1995. Effects of subbasin size on topographic characteristics and simulated flow paths in Sleepers
River Watershed, Vermont. Water Resources Research. 31(8): 1989-1997.
Wolock, D.M., G.M. Hornberger, and T.M. Musgrove. 1990. Topographic effects on flow path length and surface
water chemistry of the Llyn Brianne Catchments in Wales. Journal of Hydrology. 115:243-259.
355
-------
Appendix A: Model Fact Sheets
WAMView: Watershed Assessment Model with an Arc View Interface
Contact Information
Barry Jacobson
Soil and Water Engineering Technology, Inc.
3448 NW 12th Avenue
Gainesville, FL 32605
(352) 378-7372
jacobson(@,swet.com
http ://www. swet. com
Download Information
Availability: Proprietary.
Cost: N/A
The source code is not available. The executable code can be downloaded from http://www.swet.com with a
registration.
Model Overview/Abstract
WAMView allows engineers and planners to assess the water quality of surface and groundwater based on land use,
soils and weather stations.
Model Features
• Overland attenuation
• Stream routing and various water control structures
Model Areas Supported
Watershed High
Receiving Water None
Ecological Low
Air None
Groundwater Low
Model Capabilities
Conceptual Basis
In WAMView, the watershed is conceptualized as a series of cells/grids with different land use, soil, and land slope.
The streams are conceptualized as a series of reaches with computed geometries based on upstream drainage areas,
which may be redefined by the user.
Scientific Detail
The watershed runoff model BUCSHELL generates grid-based runoff based on land use, soil, topography, and
rainfall with two built-in watershed model GLEAMS and EAAMOD. GLEAMS is applied to upland while
EAAMOD is applied to the area with a shallow groundwater table.
356
-------
Appendix A: Model Fact Sheets
The stream routing algorithm BLASROUTE is developed based on Manning's equation without a momentum
component. A looping technique is implemented in the stream routing algorithm, and various flow structure
configurations and operation schedules are implemented. It is noted that BLASROUTE employs first order
attenuation from cells to reaches, depression and wetlands.
Model Framework
• Grid-based watershed
• One-dimensional stream routing
Scale
Spatial Scale
• Grid-based watershed; typical grid size 100m x 100m
• Typical reach/stream length 1000m to 10000m
Temporal Scale
• User-defined timestep: typically, a day
Assumptions
• Transport of water and constituent is dependent on flow distances, gradients, and type of conveyance
system.
• All input data such as land use, soil, hydrology coverages, and land use management activities are accurate.
• Rainfall data from individual stations are representative of rainfall across the entire basin.
• A reservoir routing technique without a momentum component is representative of low gradient streams.
• Phosphorous and nitrogen process models within the submodels are representative of actual transport
processes.
Model Strengths
• Capable of simulating the water quality of surface and groundwater based on land use, soil and weather.
• Capable of simulating various BMP scenarios.
• Provides a higher resolution of results than models that rely on polygon coverages.
• Works well for wetlands.
• Capable of routing attenuated runoff into a complex reach network with flow structures in the latest
version.
Model Limitations
• Does not include a momentum component in the stream routing algorithm.
• May predict flow inaccurately when applied to streams with steep slopes.
• Considers limited chemistry constituents.
• Includes groundwater component empirically, not fully integrated into the system.
Application History
Past applications of WAMView include:
• St. Johns River Watershed Assessment Project
• Suwannee River Watershed Assessment Project
• Lower St. Johns River Mainstem Subbasins Hydrologic/Water Quality Modeling
• Hydrologic Water Quality Assessment for Myakka River Basin
• North Palm Beach County Basin Pollutant Loading and Abatement Analysis
357
-------
Appendix A: Model Fact Sheets
Model Evaluation
Model evaluation was done through various projects in which the model was calibrated and then validated to a
different time period. The two watershed models used in WAMView are GLEAMS (modified slightly by SWET)
and EAAMOD (developed by SWET for areas with a shallow groundwater table). They were tested and evaluated in
the past. (See EAAMOD ~ Everglades Agricultural Area Model at http://www. swet.com/.)
Model Inputs
• Time sequences of boundary conditions—outflow with chemistry constituents of interest (if tidal outflow
exits), point sources flow with chemistry constituent of interest
• Basin polygon coverage, topography coverage, land use coverage, soil coverage, reach coverage, rain
station coverage, and utility coverage if any
• Time sequences of rainfall data for each station and other weather data, including monthly maximum and
minimum temperature, monthly average dew point temperature, wind speed and solar radiation
• Water control structure configurations and operation schedules, if any
Users' Guide
Available with the download of the executable code: http://www.swet.com. The guide is too simple.
Technical Hardware/Software Requirements
Computer hardware:
• Minimum 100 Mb hard disk space
• Minimum 64 Mb RAM
• Minimum 200 MHZ co-processor
• Minimum 600 x 800 screen resolution
Operating system:
• Windows 95/98/ME/NT/2000/XP
• Arc View 3.2a with Spatial Analyst 1.1 (or 2.0)
Programming language:
• FORTRAN for BUCSHELL and BLASROUTE
• AVENUE for pre- and post-processor in a customized Arc View
Runtime estimates:
• Hours to a day
Linkages Supported
None
Related Systems
None
Sensitivity/Uncertainty/Calibration
Not available.
358
-------
Appendix A: Model Fact Sheets
Model Interface Capabilities
The model interface is a customized Arc View interface with good pre- and post-processors. The pre-processor can
process GIS coverage data and create some model input files, and the post-processor can display simulated results in
GIS view and layout windows.
References
WAMView User's Manual. Soil and Water Engineering Technology, Inc.
Jacobson, B.M., A.B. Bottcher, N.B. Pickering, and J. G. Hiscock. 1998. Unique routing algorithm for watershed
assessment model. American Society of Agricultural Engineers Paper No. 98-2237. American Society of
Agricultural Engineers, St. Joseph, MI.
Bottcher, A.B., J.G. Hiscock, N.B. Pickering, and B.M. Jacobson. 1998. WAM: Watershed Assessment Model for
Agricultural and Urban Landscapes. Presented at the 7th International Conference on Computers in Agriculture.
October 26-30, 1998, Orlando, FL.
359
-------
Appendix A: Model Fact Sheets
WARMF: Watershed Analysis Risk Management Framework
Contact Information
Carl W. Chen
Systech Engineering, Inc.
3180 Crow Canyon Place, Suite 260
San Ramon, CA 94583
(925) 355-1780
www. svstechengineering.com
Download Information
Availability: Proprietary
Cost: Depending on projects (Systech conducts an initial model setup)
Model Overview/Abstract
WARMF was developed by Systech Engineering under sponsorship from Electric Power Research Institute (EPRI)
as a decision support system to support the watershed approach. It was designed to take stakeholders through a
series of steps to develop and evaluate water quality management alternatives for a river basin. WARMF also
provides a procedure to calculate the total maximum daily load (TMDL) of pollutants. All necessary databases,
simulation models, and graphical software are integrated into a Windows Graphical User Interface (GUI).
Model Features
• Calculates TMDLs using a bottom-up approach. TMDLs are determined for multiple control points and
different allocations of point and nonpoint loads.
• Links catchments, river segments, and lakes to form a watershed model.
• Uses data commonly available from National Climatic Data Center, EPA, USGS, Natural Resources
Conservation Services, state and other local agencies, and private data purveyors.
• Uses meteorological data to dynamically simulate runoff and nonpoint source loads from land.
• Predicts daily hydrology and water quality of rivers and lakes.
• Simulates flow, pH, temperature, dissolved oxygen, ammonia, nitrate, phosphate, suspended sediment,
coliform bacteria, major cations and anions, three algal species, and periphyton.
• Simulate metals such as iron, zinc, manganese, and copper.
• Displays spatial distributions of point and nonpoint loading using GIS map format.
• Displays water quality status in terms of suitability for fish habitat, swimming, water supply, and other
uses.
• Accounts for the source controls of atmospheric deposition, nonpoint and point source loads.
• Evaluates cost sharing schemes for pollution trading between point and nonpoint dischargers.
Model Areas Supported
Watershed High
Receiving Water Medium
Ecological None
Air Low
Groundwater Medium
360
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
WARMF is organized into five linked modules—Engineering, Data, Knowledge, Consensus, and TMDL. The
Engineering module is the dynamic, simulation model that drives WARMF. The Data module provides time series
input data (meteorological, point source) and calibration data. The Knowledge module is a utility to store documents
for the watershed. The Consensus and TMDL modules provide road maps to engage stakeholders in building
consensus on a watershed management or TMDL plan.
Scientific Detail
The computing engine is taken from the Integrated Lake-Watershed Acidification Study (ILWAS) model. The
ILWAS model divides a watershed into catchments, stream segments, and lake layers. The hydrologic module
simulates the processes of canopy interception, snow pack accumulation and snow melt, infiltration through soil
layers, evapotranspiration from soil, ex-filtration of groundwater to stream segments, kinematic wave routing of
stream flows, and flow routing of the terminal reservoir. The chemistry module performs mass balance and chemical
equilibrium calculations to account for the processes of dry deposition to the canopy, nitrification of ammonia on the
canopy, ion leaching from sap to the canopy surface, washoff by throughfall, ion leaching by snowmelt, and the soil
processes (e.g. litter fall, litter breakdown, litter decay, nitrification, denitrification, cation exchange, anion
adsorption, weathering, and nutrient uptake). Algorithms for sediment erosion and pollutant transport from farm
lands and other land uses were adapted from ANSWERS and the Universal Soil Loss Equation (USLE). The
pollutant accumulation and washoff from urban areas was adapted from the Storm Water Management Model
(SWMM).
Model Framework
• Watershed (up to five soil layers)
• One-dimensional stream segments
• Lake layers (option to select CE-QUAL-W2)
Scale
Spatial Scale
• Watershed
• One-dimensional stream
• Lake layers
Temporal Scale
• Daily
Assumptions
• The catchment is idealized by a series of compartments (canopy, snow pack, and soil layers).
• Each compartment is considered as a continuously stirred tank reactor (CSTR) for flow routing and mass
balance calculation.
• Stream hydrology is based on conservation of mass.
• The model represents the stream segment as a CSTR.
Model Strengths
• For lakes and reservoirs, it has an option of one-dimensional horizontally mixed and vertically stratified or
two-dimensional vertically stratified.
361
-------
Appendix A: Model Fact Sheets
Model Limitations
• The user needs to contract Systech to setup and calibrate the model.
• In the TMDL module, percent reductions can be specified for either point or nonpoint source and WARMF
will calculate the other. However, as the TMDL module runs through its iterations, it will reduce all
upstream sources equally. For example, if multiple point sources exist within a subwatershed or upstream,
all nonpoint sources are reduced equally; the model can not account for individual source contributions.
Application History
WARMF has been applied to various watersheds in the United States and the Techi Watershed of Taiwan.
Model Evaluation
WARMF is being tested on the Catawba River (Duke Energy), Cheat River (AEP and Allegheny Power), Chartiers
Creek (Pennsylvania power companies), and Oostanaula Creek (TVA) basins. In addition, WARMF is being tested
in cooperation with government agencies and other stakeholders in the Truckee River (California/Nevada) and Blue
River (Colorado) watersheds.
Model Inputs
• Meteorological data
• Point source loading information
• Atmospheric deposition loads
• Fertilizer application
• Subbasin shape file
• Land use shape file
• Reach Network shape file
Users' Guide
Documentation reports (e.g., User's Guide to WARMF, Documentation of Graphical User Interface) are available to
WARMF users and all reports are available for purchase from Electric Power Research Institute (EPRI).
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• Graphical User Interface
Programming language:
• Computational code: FORTRAN
Runtime estimates:
• Minutes to hours
Linkages Supported
None
362
-------
Appendix A: Model Fact Sheets
Related Systems
None
Sensitivity/Uncertainty/Calibration
An uncertainty analysis can be performed to evaluate the chance of failure for a management plan.
Model Interface Capabilities
• Pre- and post-processors
• Data display tools
References
Chen, C. W., J. Herr and W. Tsai. 2003. Enhancement of WARMF to Track Mercury Species in a River Basin from
Atmospheric Depositions to Fish Tissues. Publication No. 1005470. Electric Power Research Institute, Palo Alto,
CA.
Chen, C.W., J. Herr, andL. Weintraub. 2001. Watershed Analysis Risk Management Framework (WARMF): Update
One - A Decision Support System for Watershed Analysis and Total Maximum Daily Load Calculation, Allocation
and Implementation. Publication No.1005181. Electric Power Research Institute, Palo Alto, CA.
Chen, C., J. Herr, and L. Weintraub. 2000. Watershed Analysis Risk Management Framework (WARMF) User's
Guide: Documentation of Graphical User Interface. Report TR1000729. Electric Power Research Institute, Palo
Alto, CA.
Chen, C. W., J. Herr, and L. Ziemelis. 1998. Watershed Analysis Risk Management Framework - A Decision
Support System for Watershed Approach andTMDL Calculation. Documentation Report TR110709. Electric Power
Research Institute, Palo Alto, CA.
Chen, C.W., L. Weintraub, L. Olmsted, and R.A. Goldstein. Decision framework for sediment control in muddy
creek watershed. Journal of the American Water Resources Association.
Chen, C.W., J. Herr, and L. Weintraub. 2004. Decision support system for stakeholder involvement. Journal of
Environmental Engineering, American Society of Civil Engineers. 130(6):714-721.
Chen, C.W., J. Doherty, and J.M. Johnston. 2003 Methodologies for calibration and predictive analysis of a
watershed model. Journal of the American Water Resources Association.
Herr, J., C.W. Chen, R.A. Goldstein, R. Herd, and J.M. Brown. 2003. Modeling acid mine drainage on a watershed
scale for TMDL calculations. Journal of the American Water Resources Association. 39(2).
Chen, C.W., L.H.Z. Weintraub, J. Herr, and R.A. Goldstein. 2000. Impacts of a thermal power plant on the
phosphorus TMDL of a reservoir. Environmental Science and Policy. 3:217-223.
Chen, C. W., J. Herr, L. Ziemelis, R. A. Goldstein, and L. Olmsted. 1999. Decision support system for total
maximum daily load. Journal of Environmental Engineering, American Society of Civil Engineers. 125(7):653-659.
Chen, C.W., W.T. Tsai, and A.A. Lucier. 1998. A model of air-tree-soil system for ozone impact analysis.
Ecological Modeling. 111:207-222.
Chen, C. W., J. Herr, L. Ziemelis, M. C. Griggs, L. L. Olmsted, and R. A. Goldstein. 1997. Consensus module to
guide watershed management decisions for catawba River Basin. The Environmental Professional. 19:75-79.
Chen, C. W., J. Herr, R. A. Goldstein, F. J. Sagona, K. E. Rylant, and G. E. Hauser. 1996. Watershed risk analysis
model for TVA's Holston River Basin. Water, Air and Soil Pollution. 90:65-70.
363
-------
Appendix A: Model Fact Sheets
Chen, C.W., D. Leva, and A. Olivieri. 1996. Modeling the fate of copper discharged to San Francisco Bay. Journal
of Environmental Engineering, American Society of Civil Engineers. 122(10).
Chen, C.W., W.T. Tsai, and L.E. Gomez. 1994. Modeling Responses of Ponderosa Pine to Interacting Stresses
Ozone and Drought Forest Science. 40(2):267-288.
364
-------
Appendix A: Model Fact Sheets
WASP: Water Quality Analysis Simulation Program
Contact Information
Tim Wool
Center for Exposure Assessment Modeling (CEAM)
U.S. Environmental Protection Agency
Office of Research and Development
Environmental Research Laboratory
960 College Station Road
Athens, Georgia 30605-2720
(706)355-8312
wool.tim(@,epa. gov
Download Information
Availability: Nonproprietary
Version 5.1: http://www.epa.gov/ceampubl/swater/wasp/index.htm
Version 6.2: http://www.epa.gov/athens/wwqtsc/html/wasp.html
Cost: N/A
Model Overview/Abstract
WASP is a generalized modeling framework based on the finite-volume concept for quantifying fate and transport
of water quality variables in surface waters. The three components of the model are WASP for mass transport;
EUTRO for dissolved oxygen, nutrients, and algal kinetic; and TOXI for toxic substances. WASP is capable of
analyzing time-variable or steady state, one-, two-, or three-dimensional water quality problems. WASP5 is a DOS
application and WASP6 is a Windows application. WASP model has been widely applied to investigate dissolved
oxygen, bacteria, eutrophication, suspended solids, and toxic substance problems.
Model Features
• Tracer transport
• Eutrophication
• Dissolved oxygen
• Nutrients
• Toxic
• Sediment transport
Model Areas Supported
Watershed None
Receiving Water High
Ecological Medium
Air None
Groundwater None
365
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
The waterbodies are conceptualized as well-mixed control volume, and the law of conservation of mass is applied to
each control volume.
Scientific Detail
The governing equations for the WASP model are the advection-dispersion-reaction equations for the water quality
variables. The WASP module provides the advection and dispersion solution and the EUTRO and TOXI modules
provide the reaction solutions of dissolved oxygen, nutrients, and algae. A one-step Euler solution technique is
applied for the time difference. The advection terms in the governing equations are solved with UPWIND difference
in space. The water quality variables can be turned on/off depending on modeling requirements.
Model Framework
• One-, two-, or three-dimensional
• Any type of waterbody
Scale
Spatial Scale
• One-, two-, or three-dimensional
Temporal Scale
• User-defined timestep
Assumptions
• Completely mixing control volume
Model Strengths
• WASP model is a very flexible modeling framework and can simulate water quality in one-, two-, or three-
dimensional space.
• The control volume structure promises the conservation of mass. WASP provides the transport computation
framework and can be incorporated with EUTRO to simulate eutrophication, nutrient, and dissolved
oxygen. It also can be incorporated with TOXI to model metals, toxics, and sediment transport.
Model Limitations
• Requires external hydrodynamic model to provide flow file for solving advection. The file size might be
very large in several gigabytes for long-term simulation.
• User specified dispersion coefficient and temperature.
• First-order UPWIND difference in space may cause significant numerical diffusion.
• Over-simplified sediment flux calculation.
• No periphyton or macroalgae.
• Sediment transport processes are not related to shear stress.
Application History
A significant amount of WASP applications can be found in technical reports, journal and conference papers.
Examples of application include modeling eutrophication of Tampa Bay, Neuse River, the Great Lakes, and
Potomac Estuary; and examining phosphorus loading to Lake Okeechobee, PCB pollution of the Great Lakes, and
kepone pollution of the James River Estuary.
366
-------
Appendix A: Model Fact Sheets
Model Evaluation
WASP model is widely cited in peer reviewed journal papers.
Model Inputs
• Initial conditions
• Point and nonpoint sources inputs
• Flow file
• Vertical mixing coefficients
• Open boundary conditions
• Biological and chemical reaction rates
Users' Guide
WASP5:
• The Water Quality Analysis Simulation Program, WASPS, Part A: Model Documentation.
• The Water Quality Analysis Simulation Program, WASPS, Part B: The WASPS Input Data Set
• Available in model download file: http://www.epa.gov/ceampubl/swater/wasp/index.htm
WASP6:
• Water Quality Analysis Simulation Program (WASP) Version 6.0, Draft: User's Manual
• Available online: http://www.epa.gov/athens/wwqtsc/html/wasp.html:
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• PC-DOS, Windows
Programming language:
• FORTRAN (WASPS),
Runtime estimates:
• Minutes to hours
Linkages Supported
DYNHYD5 provides the flow information to WASPS. Other models that provides flow file include RIVMOD,
EFDC, and SWMM.
Related Systems
QUAL2E, QUAL2K, CE-QUAL-RIV1, CE-QUAL-W2, CW-QUAL-ICM, CAEDYM
Sensitivity/Uncertainty/Calibration
Not available
367
-------
Appendix A: Model Fact Sheets
Model Interface Capabilities
• WASP6 provides a Windows interface including pre-processor and post-processor
References
Ambrose, R.B., T.A. Wool, and J.L. Martin. 1993a. The Water Quality Analysis Simulation Program, WASPS, Part
A: Model Documentation. U.S. Environmental Protection Agency, Center for Exposure Assessment Modeling,
Athens, GA.
Ambrose, R.B., T.A. Wool, and J.L. Martin. 1993b. The Water Quality Analysis Simulation Program, WASPS, Part
B: The WASPS Input Dataset. U.S. Environmental Protection Agency, Center for Exposure Assessment Modeling,
Athens, GA.
Wool, T.A, R.B. Ambrose, J.L. Martin, and E.A. Comer. Water Quality Analysis Simulation Program (WASP)
Version 6.0 Draft: User's Manual
368
-------
Appendix A: Model Fact Sheets
WEPP: Water Erosion Prediction Project
Contact Information
U.S. Department of Agriculture
Agricultural Research Service
National Soil Erosion Research Laboratory
1196 Building SOIL, Purdue University
West Lafayette, Indiana 47907-1196
(765) 494-8673
nserl(@,horizon. nserl.purdue.edu
Download Information
Availability: Nonproprietary
Download site: http://topsoil.nserl.purdue.edu/nserlweb/weppmain/wepp.html
Download instructions/limitations: Registration optional
Cost: N/A
Model Overview/Abstract
The WEPP model is a process-based, distributed parameter, continuous simulation, erosion prediction model for use
on personal computers running Windows 95/98/NT/2000/XP. The current model version (v2002.700) available for
download is applicable to hillslope erosion processes (sheet and rill erosion) as well as simulation of the hydrologic
and erosion processes on small watersheds. The WEPP model (version 2002.700), WEPP Windows interface
(March 2004), CLIGEN climate generators (versions 4.3 and 5.2), documentation, and example data are included in
the download package.
Model Features
• Process-oriented
• Continuous simulation
• Erosion prediction
• Applicable to small watersheds (field-sized); can simulate small profiles (USLE types) up to large fields.
• Mimics the natural processes that are important in soil erosion. Everyday it updates the soil and crop
conditions that affect soil erosion. When rainfall occurs, the plant and soil characteristics are used to
determine if surface runoff will occur. If predicted, then the program will compute estimated sheet and rill
detachment and deposition and channel detachment and deposition.
Model Areas Supported
Watershed Medium (sediment only)
Receiving Water Low (channels)
Ecological None
Air None
Groundwater None
369
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
Conceptually WEPP is similar to CREAMS, SWRRB, and EPIC, but formulations are more process-based, rather
than statistical relations. The model is based on stochastic weather generations, infiltration theory, hydrology, soil
physics, plant science, hydraulics, and erosion mechanics.
Scientific Detail
The WEPP model includes components for weather generation, frozen soils, snow accumulation and melt, irrigation,
infiltration, overland flow hydraulics, water balance, plant growth, residue decomposition, soil disturbance by
tillage, consolidation, and erosion and deposition.
• Weather generator - precipitation by two-state Markov chain model; daily maximum and minimum
temperatures and solar radiation from normal distributions; and wind speed and direction based on the
historical distributions
• Winter processes - soil frost by fundamental heat theory and snow melt by generalized snow melt equation
• Irrigation - stationary sprinkler systems with four irrigation-scheduling options—(1) no irrigation, (2)
depletion-level scheduling, (3) fixed-date scheduling, and (4) a combination of the second and third options
• Infiltration - Green-Ampt Mein-Larson infiltration equation
• Evapotransipration - Penman equation
• Hillslope erosion and deposition - RUSLE for hillslope version; deterministic equations based on
infiltration theory, soil physics, and erosion mechanics, for other versions; and detachment, transport, and
deposition based on steady state solution to the sediment continuity equation
• Flow depth and hydraulic shear stress - steady state, spatially varied flow equations
• Impoundment component - runoff and sediment routed through several types of impoundment structures,
including farm ponds, culverts, filter fences, and check dams.
Model Framework
• Single watershed composed of a network of hillslopes and channels.
Scale
Spatial Scale
• Hillslope or small watershed
Temporal Scale
• Daily, monthly or annual
Assumptions
• Sediment delivery rate is assumed to be proportional to the product of rainfall intensity and interrill runoff
rate.
• Broad sheet flow on an idealized surface is assumed for overland flow routing and hydrograph
development.
• Steady state conditions are assumed at the peak runoff rate for erosion calculations.
• Heat flow in a frozen or unfrozen soil or soil-snow system is unidirectional.
Model Strengths
• WEPP is capable of estimating the spatial and temporal distributions of soil loss.
370
-------
Appendix A: Model Fact Sheets
• Since the model contains processed based hydrology, water balance, plant growth, and residue
decomposition components, the model has wider range of applications compared to other erosion
prediction models.
Model Limitations
• The model is not suitable for large watersheds. It is appropriate only for hillslope profiles of tens of meters
and small watersheds up to hundreds of meters.
Application History
Application of WEPP other than by the developers is very limited. Kincaid (2002) used WEPP to predict runoff and
erosion under sprinkler irrigation and found that the model can be used as an irrigation and management tool to help
prevent runoff from sprinkler irrigation.
Model Evaluation
Though there are a lot of publications available evaluating the applicability of the model. Most of them are
publications by members of the development team and are not peer reviewed. There are also few peer-reviewed
publications available. Tiwari et al. (2000) conducted a study to evaluate WEPP and also to compare the results with
the predictions by USLE and RUSLE. Bjorneberg et al. (1999) evaluated the model under furrow irrigated
conditions and found that the model could not be used without modifications. Reyes et al. (2004) conducted a study
to compare the runoff volume predictions of GLEAMS, EPIC and WEPP and the soil loss predictions of GLEAMS,
RUSLE, EPIC and WEPP. They found that the WEPP consistently under-predicted the soil loss.
Model Inputs
The WEPP model includes a number of conceptual components, which are used to predict and calculate estimates of
soil detachment and deposition. The different input requirements for these components include:
• Climate-rainfall parameters, temperature, solar radiation, wind
• Winter-freeze-thaw, snow accumulation, snow melting
• Irrigation-stationary sprinkler, furrow
• Hydrology-infiltration, depressional storage, runoff
• Water balance-evapotranspiration, percolation, drainage
• Soils-types and properties
• Crop growth-cropland, rangeland, forestland
• Residue management and decomposition
• Tillage impacts on infiltration and credibility
• Erosion-interrill, rill, channel
• Deposition-rills, channels, and impoundments
• Sediment delivery, particle sorting and enrichment
Users' Guide
Available online: http://topsoil.nserl.purdue.edu/nserlweb/weppmain/wepp.html
Technical Hardware/Software Requirements
Computer hardware:
• PC
Operating system:
• Windows 95/98/NT/2000/XP
371
-------
Appendix A: Model Fact Sheets
Programming language:
• FORTRAN 77
Runtime estimates:
• Minutes
Linkages Supported
None
Related Systems
CREAMS, SWRRB and EPIC
Sensitivity/Uncertainty/Calibration
Not available
Model Interface Capabilities
• Windows-based GUI interface for input creation
• GIS interface for WEPP (GeoWEPP)
References
Blorneberg, D. L., T. J. Trout, R. E. Sojka, and J. K. Aase. 1999. Evaluating WEPP-predicted infiltration, runoff,
and soil erosion for furrow irrigation. Transactions of the American Society of Agricultural Engineers. 42(6): 1733-
1741.
Cochrane, T. A., and D. C. Flanagan. 2003. Representative hillslope methods for applying the WEPP model with
DEMs and GIS. Transactions of the American Society of Agricultural Engineers. 46(4): 1041-1049.
Flanagan, D. C., and M. A. Nearing, eds. 1995. USDA-water Erosion Prediction Project: Technical documentation.
NSERL Report No. 10. U.S. Department of Agriculture-Agricultural Research Service- National Soil Erosion
Research Lab, West Lafayette, ID.
Kincaid, D. C. 2002. The WEPP model for runoff and erosion prediction under sprinkler irrigation. Transactions of
the American Society of Agricultural Engineers. 45(1): 67-72.
Reyes, M. R., C. W. Raczkowski, G. A. Gayle, and G. B. Reddy. 2004. Comparing the soil loss predictions of
GLEAMS, RUSLE, EPIC, and WEPP. Transactions of the American Society of Agricultural Engineers. 47(2): 489-
493.
Tiwari, A. K., L. M. Risse, and M. A. Nearing. 2000. Evaluation of WEPP and its comparison with USLE and
RUSLE. Transactions of the American Society of Agricultural Engineers. 43(5): 1129-1135.
Ward, George H., Jr. and Jennifer Benaman. 1999. Models for TMDL Application in Texas Watercourses: Screening
and Model Review. Online Report CRWR-99-7. Center for Research in Water Resources, University of Texas,
Austin.
372
-------
Appendix A: Model Fact Sheets
WinHSPF: An Interactive Windows Interface to HSPF
Contact Information
U.S. Environmental Protection Agency
Office of Water
Office of Science and Technology (430IT)
1200 Pennsylvania Avenue, N.W.
Washington, D.C. 20460
basins(@,epa.gov
AQUA TERRA
http://www.aquaterra.com/
Download Information
Availability: Nonproprietary
http://www.epa.gov/ost/ftp/basins/svstem/BASINS3/gww.htm
Cost: N/A
Model Overview/Abstract
WinHSPF was designed as an interactive Windows interface to Hydrologic Simulation Program FORTRAN
(HSPF). It is a fully-integrated component of the BASINS system but can be run stand-alone. WinHSPF can be used
to build User Control Input (UCI) files from GIS data, especially data from the EPA's BASINS system. After the
UCI file is built, WinHSPF can be used to view and modify the model representation of a watershed. The
FORTRAN program HSPF can be run from within WinHSPF. Modifying a given UCI file and saving as another
name creates multiple simulation scenarios. Within the BASINS system, WinHSPF is intended to be used in
conjunction with the interactive program known as "GENeration and analysis of model simulation SCeNarios," or
GenScn. GenScn allows the user to analyze results of model simulation scenarios and compare scenarios. This
system architecture allows the user to choose the system environment that best suits his or her needs.
Model Features
• Detailed watershed simulation model
• Watershed hydrology
• Runoff/Sediment/Pollutant generation and transport
• One-dimensional stream hydrology and transport
• Pesticide fate and transport simulation
Model Areas Supported
Watershed High
Receiving Water Medium
Ecological None
Air None
Groundwater Low
373
-------
Appendix A: Model Fact Sheets
Model Capabilities
Conceptual Basis
HSPF was the primary watershed model adopted into the original BASINS system. HSPF simulates the complete
watershed hydrology precipitation, interception, evapotranspiration, runoff, interflow, groundwater flow, and
groundwater loss to deep aquifers. A continuous simulation model driven by meteorological time series data, HSPF
also simulates the fate and transport of a wide variety of pollutants, such as nutrients, sediments, tracers, dissolved
oxygen/biochemical oxygen demand, temperature, bacteria, and user-defined constituents. It simulates pollutant
accumulation and reactions on pervious and impervious land segments, runoff and pollutant discharge loads to
stream segments, in-stream reactions, flow, and pollutant routing through river reach networks.
The BASINS WinHSPF Windows interface replaces the former Nonpoint Source Model (NPSM) interface included
in the original versions of BASINS. WinHSPF provides complete access to HSPF's functionality and user input data
files. The model estimates land use (rural and urban mixtures) specific nonpoint source loadings for selected
pollutants at an 8-digit HUC or subwatershed scale. The model uses GIS landscape data such as land use distribution
and elevation data together with the watershed and drainage stream network characteristics to automatically prepare
many of the input data it requires. WinHSPF is a continuous simulation model driven by meteorological events and
used to analyze water quality impacts from multiple point and nonpoint pollutant sources. It can be used for single
or multiple hydrological connected watersheds and is designed for evaluating alternative pollution control scenarios.
A user also can simulate point source contributions in WinHSPF. Data from EPA's Permit Compliance System
(PCS) are provided. This is a simplified implementation in that only a single value of flow from the facility and
pollutant concentration (based on NPDES permit limits) can be specified.
Scientific Detail
Land processes for pervious and impervious areas are simulated through water budget, sediment generation and
transport, and water quality constituents' generation and transport. Hydrology is modeled as water balance of soil
(or storage) in different layers as described by the Stanford Watershed Model (SWM) methodology. Interception,
infiltration, evapotranspiration, interflow, groundwater loss and overland flow processes are considered and are
generally represented by empirical equations. Sediment production is based on detachment or scour from a soil
matrix and transport by overland flow in pervious areas, while solids buildup and washoff are simulated for
impervious areas. It includes agricultural components for land-based nutrient and pesticide processes and a special
actions block for simulating management activities. HSPF also simulates the in-stream fate and transport of a wide
variety of pollutants such as nutrients, sediments, tracers, dissolved oxygen/biochemical oxygen demand,
temperature, bacteria, and user-defined constituents.
Model Framework
• Hydrologic response unit, subwatershed, and watershed
• Simple one-dimensional stream and well-mixed reservoir/lake model
• WinHSPF interface uses an object-oriented data structure to link to the HSPF model
Scale
Spatial Scale
• Lumped parameters at a land use-subwatershed basis
• Subwatershed—variable size up to internal operations limit
Temporal Scale
• User-defined timestep, typically hourly
Assumptions
• It is a distributed model by land use but ignores the spatial variation within a land use in a subwatershed.
374
-------
Appendix A: Model Fact Sheets
• The receiving waterbody is well mixed along the width and depth. Assumes one-directional flow.
Model Strengths
• One of the few watershed models capable of simulating land processes and receiving water processes
simultaneously.
• Capable of simulating both peak flow and low flow.
• Simulates at a variety of timesteps, including sub-hourly to one minute, hourly or daily.
• Simulates the hydraulics of complex natural and man-made drainage networks.
• Includes capabilities to address a variable water table.
• Simulate results for many locations along a reservoir or tributary.
• Includes user-defined model output options by defining the external targets block.
• Can be setup as simple or complex, depending on application, requirement, and data availability.
Model Limitations
• Relies on many empirical relations to represent physical process.
• Lumps simulation processes for each land use type at the subwatershed level (i.e., does not consider the
spatial location of one land parcel relative to another in the watershed). The model approaches a distributed
model when smaller subwatersheds are used; however, this may result in increased model size and
simulation time.
• Requires extensive calibration.
• Requires a high level of expertise for application.
• Is limited to well-mixed rivers and reservoirs and one-directional flow.
Application History
The underlying HSPF model is a proven, tested continuous simulation watershed model. It is one of the models
recommended by the EPA for complex TMDL studies. The HSPF model has been used widely and the applications
have been documented for more than 20 years.
Model Evaluation
The WinHSPF interface has an application history dating to the recent release of BASINS 3.0. However, the
underlying HSPF model has been widely reviewed and applied throughout its history (Hicks, 1985, Ross et al.,
1997, and Tsihrintzis et al., 1996). One of the largest applications of the model was to the Chesapeake Bay
Watershed, as part of the Chesapeake Bay Program's management initiative (Donigian, 1990, 1992). Tsihrintzis et
al. (1994, 1995) applied HSPF in a GIS shell (using ARC/INFO) to evaluate the impact of agricultural activities,
specifically the transport of sediments, nutrients, and pesticides, on streams and groundwater in southern Florida. An
extensive HSPF bibliography has been compiled to document model development and application, and is Available
online: http://hspf.com/hspfbib.html.
Model Inputs
• Continuous meteorological time series records including (at a minimum):
o Rainfall
o Potential evapotranspiration
For snow simulation, additional required meteorological time series include:
o Temperature
o Wind speed
o Solar Radiation
o Dewpoint Temperature
For additional simulation options, other required meteorological time series may include:
o Pan Evaporation
o Cloud Cover
375
-------
Appendix A: Model Fact Sheets
• Soils data (auxiliary dataset to guide hydrologic calibration), pollutant buildup and washoff, stream
dimensions or rating curves, point-source loading inputs
• A large number of parameters need to be specified (some default values are available).
Users' Guide
• WinHSPF Users' Manual is available online: http://www.epa.gov/waterscience/basins/bsnsdocs.htmMwin
• For model documentation, underlying theory, and parameterization the HSPF users' manual is a
recommended source (Bicknell et al, 2001).
• A browsable Windows help file version of the manual is available at http://hspf.com/pub/hspf/HSPF.chm.
Technical Hardware/Software Requirements
Computer hardware:
• PC with Windows Operation System
Operating system:
• Microsoft Windows
Programming language:
• FORTRAN (model) and Visual Basic (interface)
Runtime estimates:
• Minutes to less than an hour
Linkages Supported
None
Related Systems
WinHSPF, an interface to the Hydrological Simulation Program-FORTRAN (HSPF), is a key component of Better
Assessment Science Integrating Point and Nonpoint Sources (BASINS) Version 3.0. BASINS 3.0 is being
developed for the EPA's Office of Water to respond to the continued needs of various agencies to perform
watershed and water quality assessments integrating point and nonpoint sources.
Sensitivity/Uncertainty/Calibration
WinHSPF assists the user in building the necessary datasets and making the necessary modifications to the input
sequence for hydrologic calibration using the U.S. Geological Survey's Expert System for the Calibration of HSPF
(HSPEXP).
Model Interface Capabilities
WinHSPF provides an interactive interface to HSPF in a Windows environment. WinHSPF may be used for creating
a new HSPF input sequence or for modifying an existing HSPF input sequence. The program HSPF may be run
from within WinHSPF. Input sequences may be modified and saved under another name to create simulation
scenarios.
References
Bicknell, B.R., J.C. Imhoff, J.L. Kittle, Jr., T.H. Jobes, and A. S. Donigian, Jr. 2001. Hydrological Simulation
Program - Fortran, Version 12, User's Manual. AQUA TERRA Consultants.
376
-------
Appendix A: Model Fact Sheets
Hicks, C.N. 1985. Continuous Simulation of Surface and Subsurface Flows in Cypress Creek Basin, Florida, Using
Hydrological Simulation Program - FORTRAN (HSPF). Water Resources Research Center, University of Florida,
Gainesville, FL.
Lumb, A.M., McCammon, R.B., and Kittle, J.L., Jr. 1994. Users Manual for an Expert System (HSPEXP) for
Calibration of the Hydrologic Simulation Program—FORTRAN. Report 94-4168 U.S. Geological Survey Water-
Resources Investigations.
Ross, M.A., P.O. Tara, J.S. Geurink, and M.T. Stewart. 1997. FIPR Hydrologic Model: Users Manual and
Technical Documentation. Florida Institute of Phosphate Research and Southwest Florida Water Management
District, University of South Florida, Tampa, FL.
Scheckenberger, R.B., and A.S. Kennedy. 1994. The Use of HSPF in Subwatershed Planning. In Current Practices
in Modeling the Management of Stormwater Impacts. W. James. Lewis Publishers, Boca Raton, FL. pp. 175-187.
Tsihrintzis, V., H. Fuentes, and R. Gadipudi, 1994. Interfacing GIS and water quality models for agricultural areas.
Hydraulic Engineering American Society of Civil Engineers. 1:252-256.
Tsihrintzis, V.A., H.R. Fuentes and R. Gadipudi. 1996. Modeling prevention alternatives for nonpoint source
pollution at a wellfield in Florida. Water Resources Bulletin, Journal of the American Water Resources Association
32(2):317-331.
377
-------
Appendix A: Model Fact Sheets
WMS: Watershed Modeling System
Contact Information
Environmental Modeling Systems, Inc.
1204 West South Jordan Parkway, Suite B
South Jordan, UT 84095-4612
(801) 302-1400
info(@,ems-i.com
U.S. Army Corps of Engineers
Engineer Research and Development Center
Coastal and Hydraulics Lab
Hydrological Systems Branch
Vicksburg,MS39180
Download Information
Availability: Proprietary
http://www.ems-i.com/WMS/wms.html
A limited public domain version also is available at this link:
http://www.ems-i.com/WMS/WMS Freeware/wms freeware.html
Cost: Complete Package ($4,600)
Model Overview/Abstract
WMS is a comprehensive graphical modeling environment for all phases of watershed hydrology and hydraulics. It
was developed by the Environmental Modeling Research Laboratory of Brigham Young University in cooperation
with the U.S. Army Corps of Engineers Waterways Experiment Station. WMS consists of tools that automate
modeling processes such as automated basin delineation, geometric parameter calculations, and GIS overlay
computations (CN, rainfall depth, roughness coefficients, etc.). WMS 7 supports hydrologic modeling with HEC-1
(HEC-HMS), TR-20, TR-55, Rational Method, NFF, MODRAT, and HSPF. It also supports hydraulic models such
as HEC-RAS and CE-QUAL-W2. WMS is designed to be modular, enabling the user to select only those modules
and hydrologic modeling capabilities that are required. Additional WMS modules can be added at any time. To
facilitate data transfer between Arc View GIS and WMS, an extension called WMSHydro has been developed. This
extension creates a WMS/Arc View "super file" which is a collection of Arc View shapefiles and ASCII grid files.
The super file can be exported either from WMS or this extension. This super file also can be imported into WMS or
Arc View as themes (coverages) and grids.
Model Features
WMS is structured into eight modules. Only one module is active at any given time.
• Terrain Data: Used for basin delineation with Triangulated Irregular Networks (TINs).
• Drainage: Used for basin delineation with gridded Digital Elevation Models (OEMs).
• Map: Used to create data layers from GIS objects (drainage, soil, land use, etc.)
• Hydrologic Modeling: Contains interfaces to hydrologic models.
• River: Contains tools for creating one-dimensional hydraulic models.
• GIS: Used to open shape file data and convert them to feature objects.
• 2D Grid: Used for finite difference models (currently research models only).
378
-------
Appendix A: Model Fact Sheets
• 2D Scatter Point: Contains two-dimensional scatter point interpolation tools.
Model Areas Supported
Watershed High
Receiving Water High
Ecological None
Air None
Groundwater Medium
Model Capabilities
Conceptual Basis
In WMS, a subwatershed is delineated based on DEM (grid) or TIN. Several small subwatersheds and representative
streams may be networked together to represent a larger watershed drainage area. Various hydrological and
hydraulic models are supported by WMS and can be used through WMS to simulate the various land and stream
processes.
Scientific Detail
WMS supports several numerical models to compute peak flow, hydrographs, and water quality. Each model is
supported through the Hydrologic Modeling Module with a completely integrated interface for parameter input,
model run, and output display. The models available for use with WMS are:
• HEC-1 (HMS): HEC-1 is the most commonly used lumped parameter model and was developed by the
U.S. Army Corp of Engineers' (USAGE) Hydrologic Engineering Center. HEC-1 simulates the surface
runoff from a single precipitation event.
• TR-20: TR-20 is designed to compute surface runoff from natural or synthetic rainstorm events and was
developed by the U.S. Department of Agriculture's Natural Resource Conservation Service (NRCS).
• TR-55: TR-55 was developed by the NRCS as a simplified method to compute storm runoff in small,
urbanized watersheds.
• MODRAT: MODRAT is the specialized Modified Rational Method program used by the Los Angeles
County, California, to compute surface runoff.
• StormDrain: The StormDrain model does complete storm sewer network analysis for steady state or
transient flow conditions. It is based on the HYDRAIN analysis code from the Federal Highway
Administration (FHWA).
• CE-QUAL-W2: CE-QUAL-W2 is a two-dimensional (profile) hydraulic model used for water quality
analysis in rivers and reservoirs where vertical variation analysis is required.
• National Flood Frequency (NFF): The NFF program evaluates regional regression equations for estimating
flood peak discharges. It is developed by the U.S. Geological Survey (USGS) in cooperation with FHWA
and Federal Emergency Management Agency (FEMA).
• Rational Method: The Rational Method is one of the simplest and best-known methods of hydrology. It
computes peak discharge from an area based on rainfall intensity and a runoff coefficient.
• Hydrological Simulation Program-FORTRAN (HSPF): HSPF simulates hydrologic and water-quality
processes on land surfaces, streams, and impoundments. It is commonly used in the development of
TMDLs.
• HEC-RAS: HEC-RAS is a one-dimensional hydraulic model for computing water surface profiles for
steady state or gradually varied flow.
• GSSHA: GSSHA is a distributed (two-dimensional) hydrologic model developed for analysis of surface
runoff, channel hydraulics, and groundwater interaction. Water quality and sediment transport also are
supported. It was developed by the USAGE Engineer Research and Development Center (ERDC).
Model Framework
• Watershed hydrologic, sediment, and water quality
379
-------
Appendix A: Model Fact Sheets
• Time series overland, subsurface, and in-stream simulation
Scale
Spatial Scale
• Watershed or subwatershed—flexible size
• Lumped parameters at a land use-subwatershed basis
Temporal Scale
• Variety of timesteps, including hourly or daily
Assumptions
• It is a distributed model by land use but ignores the spatial variation within a land use in a subwatershed.
Model Strengths
• It is more robust in TIN triangulation and processing.
• It provides tools to extract the significant elevation points from DEMs.
• TINs or DEMs contours can be exported as feature lines and then to shape files.
• It supports several hydrological and hydraulic models through WMS interfaces.
Model Limitations
• The model relies on many empirical relations to represent physical process.
• It requires a high level of expertise for application.
• There is a limited public domain version and the full package is expensive.
Application History
WMS provides the interface linkage to various popular and tested models such as HSPF, which is a continuous
simulation watershed model. HSPF is one of the models recommended by the EPA for complex TMDL studies. The
HSPF model has been used widely and the applications have been documented for more than 20 years.
Model Evaluation
HSPF has been widely reviewed and applied throughout its long history (Hicks, 1985, Ross et al., 1997, and
Tsihrintzis et al., 1996). One of the largest applications of the model was to the Chesapeake Bay Watershed, as part
of the EPA's Chesapeake Bay Program's management initiative (Donigian, 1990, 1992). An extensive HSPF
bibliography has been compiled to document model development and application, and is available online at
http://hspf.com/hspfbib.html.
Model Inputs
Some of the most common coverage types used in WMS include:
• Drainage
• Storm Drain
• ID-Hyd Centerline
• ID-Hyd Cross Section
• Area Property
• NFF Region
• Soil Type
• Land Use
380
-------
Appendix A: Model Fact Sheets
• Time Computation
A large number of parameters need to be specified for the specific model through the graphical user interface.
Users' Guide
WMS documentation is available online:
http://www.ems-i.com/WMS/WMS Downloads/wms downloads.html
WMS online help is available at these links:
http://www.ems-i.com/wms70help/WMSHELP.htm
http://www.bossintl.com/online help/wms/source/introduction/introduction.htm
Technical Hardware/Software Requirements
Computer hardware:
• 500MhZ processor
• 128 MB RAM
• 100 MB disk space
Operating system:
• Microsoft Windows NT/Me/2000/XP
Programming language:
• Visual Basic with ArcObjects and OpenGL Technology
Runtime estimates:
• Minutes to less than an hour
Linkages Supported
• HEC-l (HEC-HMS)
• TR-20
• TR-55
• MODRAT
• StormDrain
• CE-QUAL-W2
• National Flood Frequency Program (NFF)
• Rational Method
• Hydrologic Simulation Program - FORTRAN (HSPF)
• HEC-RAS
• GSSHA (Gridded Surface Subsurface Hydrologic Analysis)
Related Systems
WMS supports several numerical models as listed in Linkages Supported section.
Sensitivity/Uncertainty/Calibration
The Hydrologic Modeling Module of WMS contains the tools for computing complex hydrologic parameters by
using digital terrain or GIS data needed for input to a model. It serves as a good starting point for model setup and
calibration.
381
-------
Appendix A: Model Fact Sheets
Model Interface Capabilities
WMS provides GIS style tools and functionality that make it easy to build models and view results. All modeling
parameters are entered through interactive graphics and easy-to-use dialog boxes. The system reads and writes
native model input/output files through graphical user interface.
References
Dellman, P.N., C.E. Ruiz, C.T Manwaring, and E.J. Nelson. 2002. Watershed Modeling System Hydrological
Simulation Program; Watershed Model User Documentation and Tutorial. Engineer Research and Development
Center, Environmental Lab, Vicksburg, MS.
Donigian, A.S. Jr., and A.S. Patwardhan. 1992. Modeling nutrient loadings from croplands in the Chesapeake Bay
Watershed. In Proceedings of water resources sessions at Water Forum '92, Baltimore, MD August 2-6, pp. 817-
822.
Donigian, A.S., Jr., B.R. Bicknell, L.C. Linker, J. Hannawald, C. Chang, and R. Reynolds. 1990. Chesapeake Bay
Program Watershed Model Application to Calculate Bay Nutrient loadings: Preliminary Phase I Findings and
Recommendations. Prepared for the U. S. Chesapeake Bay Program, Annapolis, MD, by AQUA TERRA
consultants.
Downer, Charles W.; Nelson, E. J.; Byrd, Aaron. 2003. Primer: Using Watershed Modeling System (WMS) for
Gridded Surface Subsurface Hydrologic Analysis (GSSHA) Data Development - WMS 6.1 and GSSHA 1. 43C.
Engineer Research and Development Center, Coastal and Hydraulics Lab, Vicksburg, MS.
Environmental Modeling Research Laboratory (EMRL). 1998. Watershed modeling system (WMS) reference
manual and tutorial. Brigham Young University, Provo, Utah.
Hicks, C.N. 1985. Continuous Simulation of Surface and Subsurface Flows in Cypress Creek Basin, Florida, Using
Hydrological Simulation Program - FORTRAN (HSPF). Water Resources Research Center, University of Florida,
Gainesville, FL.
Ross, M.A., P.O. Tara, J.S. Geurink, and M.T. Stewart. 1997. FIPR Hydrologic Model: Users Manual and
Technical Documentation. Florida Institute of Phosphate Research, and Southwest Florida Water Management
District, University of South Florida, Tampa, FL.
Tsihrintzis, V.A., H.R. Fuentes and R. Gadipudi. 1996. Modeling prevention alternatives for nonpoint source
pollution at a wellfield in Florida. Water Resources Bulletin, Journal of the American Water Resources Association,
32(2):317-331.
382
-------
Appendix A: Model Fact Sheets
XP-SWMM: Stormwater and Wastewater Management Model
Contact Information
XP Software, Inc
2000 NE 42nd Avenue, Suite 214,
Portland, OR, 97213
(888) 554-5022
info(g),xpsoftware.com
Download Information
Availability: Proprietary
http ://www. xpsoftware .com/products/xpswmm. htm
Cost: XP-SWMM2000 is sold as a link/node model; the cost of the software is proportionate to the size of the
project, for which it will be used.
Model Overview/Abstract
XP-SWMM is an enhanced version of SWMM coupled with the XP interface. The graphical EXPERT environment
(XP) is a friendly, graphics-based environment that encompasses data entry, run-time graphics, and post-processing
of results in graphical form. Drainage networks are drawn either on the screen over real-world topographical
backgrounds or imported from a database. It has the ability to handle systems comprising pipes and open channels,
rivers, loops, bifurcations, pumps, weirs, and ponds.
Model Features
• Watershed hydrology and water quality
• Stream transport
• Urban stormwater systems and pipes.
Model Areas Supported
Watershed High
Receiving Water Medium
Ecological None
Air None
Groundwater Low
Model Capabilities
Conceptual Basis
The SWMM model simulates overland water quantity and quality produced by storms in urban watersheds. Several
modules or blocks are included to model a wide range of quality and quantity watershed processes. A distributed
parameter sub-model (RUNOFF) describes runoff based on the concept of surface storage balance. The
rainfall/runoff simulation is accomplished by the nonlinear reservoir approach. The lumped storage scheme is
applied for soil/groundwater modeling. For impervious areas, a linear formulation is used to compute daily/hourly
increases in particle accumulation. For pervious areas, a modified Universal Soil Loss Equation (USLE) determines
sediment load. The concept of potency factors is applied to simulate pollutants other than sediment.
383
-------
Appendix A: Model Fact Sheets
Scientific Detail
In addition to EPA SWMM's Non-linear Runoff Routing, XP-SWMM has 10 additional ways to estimate surface
runoff:
1. SCS Unit Hydrographs using a Curve Number with curvilinear unit hydrographs
2. SCS Unit Hydrographs using a Curve Number with triangular unit hydrographs
3. Kinematic Wave
4. Snyder Unit Hydrograph
5. Snyder (Alameda) Unit Hydrograph
6. Nash Unit Hydrograph
7. Santa Barbara Urban Hydrograph
8. Laurenson's Non-linear Runoff Routing (RAFTS)
9. Rational Method
10. Colorado Urban Hydrograph Procedure (CUHP)
The model has more capabilities in infiltration, sub-surface flow, and groundwater flow than that of EPA SWMM.
In addition to SWMM capabilities, pollutant routing is available for all modules, including the Hydraulics layer.
More than 30 types of conduits can be input into the model.
Model Framework
• One-dimensional mass balance flow and pollutant routing
Scale
Spatial Scale
• Watershed or subwatershed—flexible size
Temporal Scale
• Variety of timesteps, including hourly or daily
Assumptions
• The model performs best in urbanized areas with impervious drainage, although it has been widely used
elsewhere.
• Model parameters for quantity and quality simulations are developed such that the model will be calibrated
to enhance its capability.
• All the pollutants entering the waterbodies are sediment adsorbed.
Model Strengths
• Completely interactive analysis engine, allowing the user to change parameters mid-run, pause or terminate
a run, and to graph results on the fly.
• Dynamic memory allocation.
Model Limitations
• Not a public domain product.
• Lack of subsurface quality routing
• No interaction of quality processes
• Limited kinetics (A first order decay rate can be specified for each pollutant in the Transport Block.)
• Difficulty in simulation of wetlands quality processes
• Rudimentary scour-deposition routine in the Transport Block
384
-------
Appendix A: Model Fact Sheets
Application History
SWMM has been applied to urban hydrologic quantity/quality problems in scores of U.S. cities as well as
extensively in Canada, Europe, and Australia. The model has been used for very complex hydraulic analysis for
combined sewer overflow mitigation, as well as for many stormwater management planning studies and pollution
abatement projects (Huber, 1992). Warwick and Tadepalli (1991) describe calibration and verification of SWMM on
a 10-square-mile urbanized watershed in Dallas, Texas. Tsihrintzis et al. (1995) describe SWMM applications to
four watersheds in South Florida representing high- and low-density residential, commercial, and highway land uses.
Ovbiebo and She (1995) describe an application of SWMM in a subbasin of the Duwamish River, Washington.
Model Evaluation
[SWMM or XP-SWMM] is widely applied in Florida, but the applications are primarily limited to urban areas and
event-based applications.
Model Inputs
• Data requirements for hydrology depend on the user's choice. Land use data are used to determine ground
cover type for each model subarea.
• Depending on what options are set for the loading calculations additional parameters are necessary (e.g.,
buildup coefficients would be needed for the dry weather buildup simulation).
• Additional data are necessary if the user intends to model snowmelt, subsurface drainage, and interflow.
• Depending on the stormwater system, dimensions, slopes, roughness coefficients, elevations, and storage
are required.
Users' Guide
Available online: http://www.xpsoftware.com/products/xpswmm.htm
Technical Hardware/Software Requirements
Computer hardware:
• 200 Mb of available disk space
• 128 Mb of RAM
Operating system:
• Microsoft Windows
Programming language:
• FORTRAN (model) and Visual Basic (interface)
Runtime estimates:
• Minutes to less than an hour
Linkages Supported
EPA SWMM model
Related Systems
XP-SWMM can optionally run EPA SWMM model.
385
-------
Appendix A: Model Fact Sheets
Sensitivity/Uncertainty/Calibration
S WMM has a high level of accuracy with careful calibration and sufficient data.
Model Interface Capabilities
• XP-SWMM has intuitive graphical dialog boxes.
• Data input is subject to expert system filtering and checking.
• Storage and retrieval of typical infrastructure attributes such as pipe depths and locations is made easy
through the user-friendly model interface.
• Data entry and model output is in graphical form.
• XP-SWMM uses an object oriented Graphical Expert Environment in which the user can create the
drainage network interactively on the screen using a mouse and toolbar icons.
References
Donigian, A.S., Jr., and W.C. Huber. 1991. Modeling of Nonpoint Source Water Quality in Urban and Non-urban
Areas. EPA/600/3-91/039. U.S. Environmental Protection Agency, Environmental Research Laboratory, Athens,
GA.
Huber, W. C. 1992. Experience with the US. EPA SWMM Model for analysis and solution of urban drainage
problems. In Proceedings, Inundaciones Y Redes De Drenaje Urbano, ed. J. Dolz, M. Gomez, andJ. P. Martin, eds.,
Colegio de Ingenieros de Caminos, Canales Y Puertos, Universitat Politecnica de Catalunya, Barcelona, Spain, pp.
199-220.
Huber, W.C. 2001. New Options for Overland Flow Routing in SWMM. American Society of Civil Engineers -
Environmental and Water Resources Institute, World Water and Environmental Congress, Orlando, FL.
Huber, W.C., and R.E. Dickinson. 1988. Storm Water Management Model Version 4, User's Manual. EPA 600/ 3-
88/ 00la (NTIS PB88-236641/ AS). U.S. Environmental Protection Agency, Athens, GA.
Ovbiebo, T., and N. She. 1995. Urban Runoff Quality and Quantity Modeling in a Subbasin of the Duwamish Rivet-
using XP-SWMM. Watershed Management: Planning for the 21st Century. American Society of Civil Engineers,
San Antonio, TX. pp.320-329.
Tshihrintzis, V. A., R. Hamid, and H. R. Fuentes. 1995. Calibration and verification of watershed quality model
SWMM in sub-tropical urban areas. In Proceedings of the First International Conference - Water Resources
Engineering. American Society of Civil Engineers. August 14-16. San Antonio, TX. pp 373-377.
Tsihrintzis, V. and R. Hamid, 1998. Runoff Quality Prediction from Small Urban Catchments using SWMM.
Hydrological Processes. 12 (2), pp. 311-329.
Warwick, J. J., and P. Tadepalli. 1991. Efficacy of SWMM application. Journal of Water Resources Planning and
Management. 117(3):352-366.
386
------- |