Total Maximum Daily Load Development and Implementation: Models, Methods, and Resources

Book_5114_C000
Half Title
Title Page
Copyright Page
Contents
Foreword
Preface
TMDL Analysis and Modeling Task Committee
Blue-Ribbon Panel Reviewers
Contributing Authors
Acknowledgments
Executive Summary
Book_5114_C001
CHAPTER 1: Introduction
1.1 Definitions and History of the Total Maximum Daily Load Approach
1.1.1 Total Maximum Daily Load Definition and Approach
1.1.2 History of the Total Maximum Daily Load Approach to Water Quality–Based Management
1.2 Process of Determining Total Maximum Loads
1.2.1 Procedure for Total Maximum Daily Load Determination, Allocation, and Implementation Planning
1.2.1.1 Identification
1.2.1.2 Determination
1.2.1.3 Allocation
1.2.1.4 Implementation planning and compliance
1.2.2 Synoptic and Monitoring Data Requirements
1.3 Modeling Required to Determine a Total Maximum Daily Load
1.4 Benefits of the Total Maximum Daily Load Approach
1.5 Purpose of This Manual of Practice
1.6 Intended Audience for This Manual
1.7 Organization of This Manual
References
Book_5114_C002
Chapter 2: Watershed Models
2.1 Introduction
2.2 Brief Descriptions of the Selected Watershed Models
2.2.1 Agricultural Nonpoint-Source and Annualized Agricultural Nonpoint-Source Models
2.2.2 ANSWERS-2000
2.2.3 Dynamic Watershed Simulation Model
2.2.4 Gridded Surface and Subsurface Hydrologic Analysis
2.2.5 Generalized Watershed Loading Function
2.2.6 Hydrologic Engineering Center-Hydrologic Modeling System
2.2.7 Hydrological Simulation Program-Fortran
2.2.8 Kinematic Runoff and Erosion
2.2.9 Loading Simulation Program in C++
2.2.10 MIKE Système Hydrologique Européen
2.2.11 Soil and Water Assessment Tool
2.2.12 Storm Water Management Model
2.2.13 Watershed Assessment Model
2.2.14 Watershed Analysis Risk Management Framework
2.3 Analyses of Models: Suitability for Total Maximum Daily Loads
2.3.1 Key Characteristics and Capabilities of the Models
2.3.2 Hydrologic Simulations in the Models
2.3.3 Water Quality Simulations in the Models
2.3.4 Strengths and Limitations of the Models and Suitability for Total Maximum Daily Loads
2.4 Summary, Conclusions, and Recommendations
2.5 State-of-the-Art and State-of-the-Practice
References
Book_5114_C003
CHAPTER 3: Receiving Water Quality Models
3.1 Introduction
3.2 Receiving Water Quality Models for Total Maximum Daily Load Applications
3.2.1 Corps of Engineers Integrated Compartment Water Quality Model
3.2.1.1 Model Background and Capabilities. Corps of Engineers Integrated Compartment Water Quality Model (CE-QUAL-ICM), or simply referred to as ICM, is a multidimensional water quality model developed by US Army Corps of Engineers (USACE 2014)-En
3.2.1.2 Applicability to Total Maximum Daily Load Studies. The ICM can simulate water quality responses to point and nonpoint-source loads and can be used as part of TMDL modeling. The model has been applied in the environmental restoration project
3.2.2 Water Quality Analysis Simulation Program
3.2.2.1 Model Background and Capabilities. The Water Quality Analysis Simulation Program (WASP) was initially developed as a transport code with water quality subroutines. After Di Toro (1983) applied the WASP model to simulate nutrient cycling in
3.2.2.2 Applicability to Total Maximum Daily Load Studies. The WASP model is widely used in conjunction with other transport hydrodynamic models to simulate complex water quality processes in rivers, lakes, reservoirs, estuaries, and coastal waters
3.2.3 Environmental Fluid Dynamics Code
3.2.3.1 Model Background and Capabilities. The Environmental Fluid Dynamics Code (EFDC) is a surface water model with hydrodynamic and water quality modeling capabilities. The EFDC model was originally developed at the Virginia Institute of Marine S
3.2.3.2 Applicability to Total Maximum Daily Load Studies. The EFDC model has been widely used in more than 100 modeling studies of aquatic ecosystems around the world and in multiple TMDL studies. TMDL applications include the Peconic Bay in New Y
3.2.4 CE-QUAL-W2
3.2.4.1 Model Background and Capabilities. The CE-QUAL-W2 model is a two-dimensional (2D), laterally averaged hydrodynamic and water quality model. The hydrodynamic model capabilities include the simulation of water levels and depths, flow velocitie
3.2.4.2 Applicability to Total Maximum Daily Load Studies. The CE-QUAL-W2 model has been widely used as a management tool to evaluate effects from various stressors, including temperature, nutrients, and organic loads in waterbodies (Bowen and Hie
3.2.5 Hydrologic Engineering Center-River Analysis System
3.2.5.1 Model Background and Capabilities. HEC-RAS is a 1D and 2D hydraulic and water quality model for riverine ecosystems developed by the USACE Hydrologic Engineering Center (HEC). HEC-RAS is an extensively used model worldwide designed to perfor
3.2.5.2 Applicability to Total Maximum Daily Load Studies. The HEC-RAS water quality model has been used to support TMDLs and environmental impact statement studies. Recent studies include the lower Minnesota River (Zhang and Johnson 2014), Misso
3.2.6 Center for Computational Hydroscience and Engineering-1D/2D/3D
3.2.6.1 Model Background and Capabilities. The numerical models CCHE-1D/2D/3D have been developed by the National Center for Computational Hydroscience and Engineering at the University of Mississippi. This development was supported by the USDA Agri
3.2.6.2 Applicability to Total Maximum Daily Load Studies. CCHE-1D/2D/3D models are applicable to TMDL studies for nutrients, sediment, toxic chemicals in channel networks, rivers, lakes, and coastal waters. The CCHE-1D model has been applied to si
3.2.7 Environmental Protection Division-RIV1
3.2.7.1 Model Background and Capabilities. The EPD-RIV1 model is a 1D, cross-sectional-averaged, hydrodynamic, and water quality model for rivers and streams. The EPD-RIV1 model was originally developed for the Georgia Environmental Protection Divis
3.2.7.2 Applicability to Total Maximum Daily Load Studies. The EPD-RIV1 model can be used in 1D river systems subject to dynamic hydrodynamics. EPD-RIV1 provides time-varying simulations of water temperature and water quality with a primary focus o
3.2.8 QUAL2K
3.2.8.1 Model Background and Capabilities. The QUAL2K model is a 1D water quality model for river and stream networks. The model is based on the algorithms and routines originally included in the QUAL2E model with improvements in the representation
3.2.8.2 Applicability to Total Maximum Daily Load Studies. QUAL2K has been used to support WLAs and TMDL studies of rivers and streams. Typical applications are related to pollution caused by pathogens, excess nutrients such as nitrogen and phospho
3.2.9 MINTEQA2 and Visual MINTEQ
3.2.9.1 Model Background and Capabilities. The MINTEQA2 model (Allison et al. 1991) is a geochemical equilibrium-speciation model for the fate and transport of metals in aqueous systems. MINTEQA2 and Visual MINTEQ simulate the equilibrium and mass
3.2.9.2 Applicability to Total Maximum Daily Load Studies. Simulation of the fate and transport and speciation of dissolved metals, free metal ions, sorbed metals, metal precipitates, and metal complexes is a difficult task. Speciation is driven by
3.2.10 One-Dimensional Transport with Equilibrium Chemistry
3.2.10.1 Model Background and Capabilities. The One-Dimensional Transport with Equilibrium Chemistry (OTEQ) model is a reactive transport model that simulates the fate and transport of solutes and speciation and transport of metals in rivers and str
3.2.10.2 Applicability to Total Maximum Daily Load Studies. OTEQ can be applied to support TMDLs and WLA studies related to the fate and transport of metals in rivers and streams. The model has been previously used to support the evaluation of reme
3.2.11 MIKE 11
3.2.11.1 Model Background and Capabilities. MIKE 11 is a River Hydraulics and Sediment Transport model developed by the Danish Hydraulic Institute (DHI) Water and Environment which is currently in the MIKE+ platform (DHI 2021). The MIKE 11 model h
3.2.11.2 Applicability to Total Maximum Daily Load Studies. MIKE 11 is typically linked to ECO-Lab for water quality studies involving eutrophication of waterbodies, nutrient transport and cycling, and to support TMDL projects (e.g., Liang et al.
3.3 State-of-the-Art and State-of-the-Practice
References
Book_5114_C004
Chapter 4: Integrated Modeling Systems and Linked Models
4.1 Introduction
4.2 Integrated Modeling Systems
4.2.1 BASINS Modeling System
4.2.1.1 BASINS Model Releases. The BASINS version 1 was released in 1996 and consisted of various data sets (e.g., land use, water quality, digital elevation, river reach network, streamflow, and meteorological data), models such as HSPF, QUAL2E, an
4.2.1.2 Data and Supported Models in BASINS Modeling System. The BASINS modeling system has four data categories: base cartographic, environmental background, environmental monitoring, and point-source data. The base cartographic data include admin
4.2.2 Watershed Modeling System
4.2.2.1 Overview of Watershed Modeling System. The watershed modeling system (WMS) is a comprehensive graphical modeling environment for all phases of watershed hydrology and hydraulics, which was used to support TMDL development. It was developed b
4.2.2.2 Applications of the Watershed Modeling System. WMS was used in support of the Phosphorus TMDL for Conesus Lake within Livingston County, New York (USEPA Region 2, NYSDEC 2019). The nuisance aquatic plant growth is because of excess nutrie
4.3 Linked Model Applications
4.3.1 Linked Models in Chesapeake Bay Total Maximum Daily Load
4.3.2 LSPC-EFDC-WASP Models Applied to the Saugahatchee Creek Watershed (Alabama)
4.3.3 HSPF, UnTRIM, and CE-QUAL-ICM Applied to the Lynnhaven River Watershed (Virginia)
4.4 Performance Evaluations of Integrated Modeling Systems and Linked Models
4.5 Summary
4.6 State-of-the-Art and State-of-the-Practice
Acknowledgments
References
Book_5114_C005
Chapter 5: Simple Models and Methods
5.1 Introduction
5.2 Review of Simple Models and Methods
5.2.1 Simple Watershed Models and Methods
5.2.1.1 Simple Mass Balance. Simplified mass balances are typically performed in spreadsheet models that consider discrete water volumes containing a uniform concentration (or reliably represented by an average or median concentration) of a nonreact
5.2.1.2 Simple Method to Estimate Urban Stormwater Loads. The simple method (Schueler 1987) makes use of the calculated flux of the discharged pollutant to estimate the annual pollutant load or yield from a small urban watershed (catchment). Mode
5.2.1.3 Watershed Treatment Model. The Watershed Treatment Model (WTM) was developed by the Center for Watershed Protection (Caraco 2013) and provides estimates of annual nitrogen, phosphorus, total suspended solids, and bacterial loads from urba
5.2.1.4 Spreadsheet Tool for Estimating Pollutant Loads. The Spreadsheet Tool for the Estimation of Pollutant Load (STEPL) developed by Tetra Tech for USEPA (with the latest version as 4.4 in 2018) is an annual pollutant loading estimation tool. Th
5.2.1.5 Revised Universal Soil Loss Equation 2. The Revised Universal Soil Loss Equation 2 (RUSLE2) (USDA 2003) is an updated, advanced erosion calculation procedure that estimates sediment loading to receiving waterbodies, primarily headwater stre
5.2.1.6 Load Estimator. LOAD ESTimator (LOADEST) is a software program for estimating constituent loads in streams and rivers, which can be used a tool in supporting TMDL development. Given a time series of streamflow, additional data variables, an
5.2.1.7 Spatially Referenced Regressions on Watershed Attributes. The United States Geological Survey (USGS) has a long history of deriving regression equations for watershed yields of water (streamflow for ungauged watersheds) and pollutants based
5.2.1.8 Long-Term Hydrologic Impact Analysis. L-THIA was developed by Purdue University as a screening tool to use in assessing the long-term impacts of land-use changes. L-THIA results are intended to provide insight into the relative hydrologic i
5.2.1.9 Simple Transient Mass Balance Models. In some instances, standard models may not be suitable for TMDL development for certain pollutants. In these circumstances, customized model applications need to be developed. In most cases, these model
5.2.1.10 Geographical Information System Workflow Models. Increased use of geographic information systems (GISs) and high-resolution remote sensing analysis in support of TMDL modeling has led to the development of simple object-oriented modeling t
5.2.2 Simple Receiving Water Models and Methods
5.2.2.1 BATHTUB. The simplified lake eutrophication model BATHTUB was developed for the US Army Corps of Engineers (USACE) in 1985 to evaluate and manage USACE reservoirs (Walker 1985, 1986) prior to the development of USEPA and state TMDL progr
5.2.2.2 Stream Segment Temperature Model. One of the important water quality parameters in receiving water streams is temperature. The Stream Segment Temperature (SSTEMP) model is a scaled-down version of the Stream Network Temperature (SNTEMP) mod
5.2.2.3 Load–Duration Curve. The load–duration curve (LDC) approach allows the characterization of water quality concentrations for varied flow regimes. The pollutant load is expressed as a function of these diverse flow conditions, including criti
5.3 Summary
5.4 State-of-the-Art and State-of-the-Practice
Acknowledgments
References
Book_5114_C006
CHAPTER 6: Critical Condition Determination for Total Maximum Daily Load Modeling
6.1 Introduction
6.2 Methodology for Critical Condition Determination for Total Maximum Daily Load Modeling
6.2.1 Steady-State Models for Analyzing Impairment under Constant Flow
6.2.2 Dynamic Continuous Simulation Models for Analyzing Impairment under Variable Flow
6.2.3 Load–Duration Curves
6.2.4 Critical Flow-Storm Approach
6.3 Summary
6.4 State-of-the-Art and State-of-the-Practice
References
Book_5114_C007
CHAPTER 7: Model Data, Geographical Information Systems, and Remote Sensing
7.1 Model Data Requirements
7.1.1 Model Parameter Values
7.1.2 System Data
7.1.3 Total Maximum Daily Load Data Resources
7.2 Geographic Information System and Remote Sensing
7.2.1 Geographic Information System
7.2.2 Remote Sensing
7.2.2.1 Light Detection and Ranging. Light detection and ranging (LiDAR) is categorized as active remote sensing that records laser pulses that strike an object to detect the object and determine the distance between the instrument and the object (r
7.2.2.2 Multispectral Remote Sensing Imagery. Multispectral imagery is produced by sensors that capture the backscattered energy from multiple sections of the electromagnetic spectrum. Although multispectral sensors enable an analysis of extensive
7.2.2.3 Hyperspectral Remote Sensing Imagery. Hyperspectral data are composed of many contiguous spectral bands with narrow bandwidths. Hyperspectral sensors such as the compact airborne spectrographic imager (CASI), the airborne prism experiment (
7.2.2.4 Specialized Platforms. Apart from imaging platforms, several airborne and spaceborne platforms have been specifically designed to detect and track physical phenomena of interest. For example, the Ice, Cloud, and land Elevation Satellite 2 (
7.3 Geographic information system and Remote Sensing Applications in Total Maximum Daily Load Modeling
7.4 State-of-the-Art and State-of-the-Practice
References
Book_5114_C008
CHAPTER 8: Model Calibration and Validation
8.1 Introduction
8.2 Model Data and Sources of the Data
8.3 Data Management
8.4 Precalibration
8.5 Calibration
8.6 Manual Calibration and Automatic Calibration
8.7 Validation
8.8 Evaluation Criteria for Model Performance
8.9 State-of-the-Art and State-of-the-Practice
Acknowledgments
Disclaimer
References
Book_5114_C009
CHAPTER 9: Model Uncertainty Analysis and the Margin of Safety
9.1 Uncertainty Analysis
9.2 Total Maximum Daily Load Margin of Safety
9.3 Estimation of the Margin of Safety
9.3.1 Explicit Margin of Safety
9.3.2 Implicit Margin of Safety
9.3.3 Risk-Based Margin of Safety
9.4 Determination of the Margin of Safety in Case-Study Total Maximum Daily Loads
9.5 Uncertainty Analysis Methods
9.5.1 First-Order Variance Analysis
9.5.2 Monte Carlo Method
9.5.3 Bayesian Monte Carlo Analysis
9.5.4 Generalized Likelihood Uncertainty Estimation Method
9.5.5 Markov Chain Monte Carlo
9.5.6 Kalman Filtering
9.6 State-of-the-Art and State-of-the-Practice
Disclaimer
References
Book_5114_C010
Chapter 10: USEPA Total Maximum Daily Load Report Archive and Report Search Tool
10.1 Introduction
10.2 Accessing the Total Maximum Daily Load Database and Use of the Total Maximum Daily Load Report Selection Tool
10.2.1 USEPA ATTAINS Database
10.2.2 Features of the Total Maximum Daily Load Report Selection Tool
10.2.3 Total Maximum Daily Load Report Summary Statistics using the TMDL Report Selection Tool
10.2.4 Total Maximum Daily Load Model Development Workflow using the TMDL Report Selection Tool
10.3 State-of-the-Art and State-of-the-Practice
References
Book_5114_C011
Chapter 11: Model Selection and Applications for Total Maximum Daily Load Development
11.1 Introduction
11.2 Types of Models and Methods for Total Maximum Daily Load Determination
11.3 Model Selection for Total Maximum Daily Load Determination, ALLOCATION AND IMPLENTATION PLANNING Using a Holistic Approach
11.3.1 Fundamental Model Selection Principles
11.3.2 Technical and Practical Considerations in Model Selection
11.3.3 Stakeholder and Expert Engagements
11.4 Model Evaluation, Calibration–Validation, and Uncertainty Estimation
11.4.1 Performance Evaluation
11.4.1.1 Performance Evaluation during Model Selection. If model selection (Figure 11-1) and calibration fail to provide a reliable model, then the Nash–Sutcliffe efficiency (NSE) or Model Skill Score (MSS) is useful to determine if the best unrel
11.4.1.2 Performance Evaluation during Model Calibration and Validation. The best scientific practice in model calibration is to create the analytical likelihood from the synoptic data. Then, given a particular set of model parameters that need to
11.4.2 General Purpose Performance Measures for Qualitative Assessment
11.4.3 Model Calibration, Validation, and Sensitivity Analysis
11.4.4 Model Uncertainty Estimation
11.5 Modeling to Determine, Allocate, and Implement Total Maximum Daily Load
11.6 Future Innovation in Total Maximum Daily Load Modeling
11.7 State-of-the-Art and State-of-the-Practice
Appendix: Total Maximum Daily Load Implementation Planning Case Study (Alabama)
Acknowledgments
References
Book_5114_C012
Chapter 12: Modeling for Total Maximum Daily Load Implementation
12.1 Introduction
12.2 Choice of Models for Total Maximum Daily Load Implementation
12.2.1 Range of Pollutants and Pollutant Sources
12.2.2 Stormwater Best Management Practice Pollutant Control
12.2.3 Data Resources for Modeling
12.2.4 Expertise and Access
12.2.5 Availability and Support
12.3 Potential Models for Implementation
12.3.1 State or Watershed-Specific Models
12.3.1.1 Chesapeake Bay Program. The Chesapeake Bay Program, in conjunction with MDE and VA DEQ, developed the original version of the CAST, a model used by local jurisdictions to further BMP implementation planning (version 2019). The CAST encourag
12.3.1.2 Pennsylvania. Beginning in 1999, the Pennsylvania Department of Environmental Protection (PADEP), along with the Pennsylvania State University, developed a geographical information system (GIS)-based tool using the Environmental Systems Re
12.3.1.3 Wisconsin. The Wisconsin Department of Natural Resources (WI DNR) has issued guidance to Municipal Separate Storm Sewer System (MS4) permittees to assist them in the process of implementation planning and facilitating load-reduction activi
12.3.2 Spreadsheet Analysis
12.3.2.1 Watershed Treatment Model. WTM, developed by the Center for Watershed Protection (Caraco 2013), enables the analyst or entity responsible for TMDL implementation to predict the impact on pollutant loads over a wide range of land uses unde
12.3.2.2 Spreadsheet Tool for Estimating Pollutant Load. The spreadsheet tool for estimating pollutant load (STEPL) model was developed for the Office of Water of USEPA, with the latest version being 4.4b (USEPA 2020a). It has been used throughou
12.3.3 Custom Modeling Using a Geographic Information System Geodatabase
12.4 Summary
12.5 State-of-the-Art and State-of-the-Practice
Acknowledgments
References
Book_5114_A001
APPENDIX A: CONVERSION OF UNITS

APPENDIX B: Glossary
APPENDIX C: Abbreviations
APPENDIX D: Symbols
Book_5114_IDX