EPA/600/R-03/139
                                           September 2003
  Evaluation of Sediment Transport
Models and Comparative Application
        of Two Watershed Models
                         By
                       Latif Kalin
             Oak Ridge Institute for Science and Education
                   Cincinnati, Ohio 45268

                         and

                   Mohammed M. Hantush
            National Risk Management Research Laboratory
                Office of Research and Development
               U.S. Environmental Protection Agency
                   Cincinnati, Ohio 45268
            National Risk Management Research Laboratory
                Office of Research and Development
               U.S. Environmental Protection Agency
                    Cincinnati, Ohio 45268

-------
                                           Notice
    The U.S. Environmental Protection Agency through its Office of Research and Development funded
the research described here. It has been subjected to the Agency's peer and administrative review and has
been approved for publication as an EPA document.

    This research was supported in part by an appointment to the Post Doctoral Research Program at the
National Risk Management Research Laboratory, administered by the Oak Ridge Institute for Science and
Education through Interagency Agreement No DW89939836 between the U.S. Department of Energy and
the U.S. Environmental Protection Agency.  Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.

-------
                                          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.
                                                Lee A. Mulkey, Acting Director
                                                National Risk Management Research Laboratory
                                               in

-------
                                          Abstract

    Suspended solids and sediments are regarded as the two leading pollutants of nation's streams and
waterbodies. They serve as  carriers for various pesticides,  radioactive materials  and nutrients. Section
303(d) of the 1972 Clean Water Act requires states, territories, and authorized tribes to identify and list
impaired waters every two years and to develop Total Maximum Daily Loads (TMDLs) for pollutants in
these  waters.  Mathematical models  are widely  accepted, effective and powerful  tools for TMDL
development, and evaluating performances of Best Management Practices (BMP).  The rapid pace of
computer technology has been  a milestone for mathematical models in hydrology, hydrodynamics and
recently water quality. The high  demand on computer models resulted in development of many models and
placed a new burden on model users, that is model selection. The selection of the right model under certain
constraints requires a comprehensive knowledge of the capabilities and features of available models. This
report provides an overview  and evaluation of sediment models and compares two distributed, watershed
scale  models by  application to an experimental watershed.  A  probabilistic, risk-based mathematical
optimization framework is presented and proposed as a strategy for solving the  TMDL-BMP problem
involving multiple stressors in feature endeavors. Future modeling efforts may benefit from exploring the
use of system analysis approaches to obtain cost-effective, optimal load reductions using BMPs.

    The report is comprised of two parts. The first part evaluates and summarizes some of the key features
of the most widely cited watershed scale, hydrodynamic and water quality models with the emphasis on
TMDLs and BMPs.  Reviewed  models were selected based on minimum criteria.  Water quality models,
specifically those that can simulate nutrients in the environment are also considered since transport and fate
of sediments and nutrients are intimately related phenomena. Among the reviewed loading models SWAT
and AGNPS offer the most BMP alternatives at agricultural watersheds. For urban areas SWMM, and for
mixed  land uses,  i.e. rural and  urban, HSPF are identified  as the most suitable loading models.  These
models need to be used with hydrodynamic and water quality models for a complete TMDL analysis and
BMP  development.  BASINS and  MIKE-SHE are comprehensive watershed-water quality modeling
systems, with varying degrees of complexity. WMS offers a tractable watershed-modeling platform if fully
developed can be used for sediment TMDLs allocation. Available and potential model linkages between
loading, hydrodynamic and water quality models are also discussed. It is observed that most physically
based models are incapable for a complete BMP assessment. As a future need in modeling, enhancement of
such models to simulate more BMPs is recommended along with development of more linkages between
loading and hydrodynamic/water quality models.

    The  second part  of the report evaluates, by application to an experimental watershed, two promising
distributed watershed-scale sediment models in detail: KINEROS-2 and GSSHA. Sensitivity of KINEROS-
2 to model parameters was evaluated within a probabilistic framework using Monte Carlo simulations to
identify key model parameters  for calibration. It was  shown that the order of  parameter sensitivities
changes with the  quantity of interest (peak flow, total sediment yield, etc.). The  calibration/verification
procedure performed over KINEROS-2 has  shown that the  Manning's  roughness and  soil erosion
parameters show systematic seasonal variations. Both models were calibrated and verified and the results
clearly highlight the challenges modelers face when applying complex, distributed watershed models. The
results are discussed  and compared. They highlight the importance for numerical application of different
watershed models to gauged watersheds as means for  models evaluation.  Future efforts aiming  at the
evaluation of hydrologic and water quality  models should migrate from qualitative  analysis to  actual
comparative applications to real case studies.
                                               IV

-------
                                        Contents


Foreword	iii

Abstract	iv

Figures	vi

Tables	vii

1    Introduction	1
   1.1  Overview	1
   1.2  Total Maximum Daily Loading (TMDL)	3
   1.3  Mathematical Models	3
     1.3.1       Brief History of Sediment Modeling	4
   1.4  Risk Management Watershed Modeling	4

2    Model Classifications	8

3    Model Evaluation Criteria	10
   3.1  Screening Criteria	10
   3.2  Evaluation Criteria	10

4    Model Selection and Comparisons	12
   4.1  Model Selection	12
   4.2  Evaluation and BMPs Capabilities	12

5    Modeling of Sediment Yield in a Small Agricultural Watershed with KINEROS-2	20
   5.1  Model Background:	20
   5.2  Data and Model Parameters	21
   5.3  Sensitivity Analysis and MC Simulations	21
   5.4  Model Calibration, Validation	26
   5.5  Discussion:	29

6    Comparison of KINEROS-2 with GSSHA	31
   6.1  Model Features	31
   6.2  Approach	33
     6.2.1       Flow Simulations	33
     6.2.2       Erosion Simulations	36
   6.3  Long-Term Simulations  with GSSHA	38
   6.4  Discussion	41

7    Summary and Conclusions	43

8    References	45

Appendix: Model Summaries	48
   8.1  Loading Models	48
   8.2  Receiving Water Models:	63

-------
                                           Figures

Figure 1. Carbon cycle	1
Figure 2. Nitrogen cycle	2
Figure 3. Simplified schematics of sediment water interactions	2
Figure 4. Simplified schematic of various BMPs at the watershed scale  (adapted from USEPA
     2002a)	5
Figure 5. Flow of information during optimal BMP selection	6
Figure 6. Various waterbodies	9
Figure 7. Relationship between different model groups	9
Figure 8. Schematic of W-2 watershed	22
Figure 9. MC versus theoretical mean and std. of G	24
Figure 10. Rainfall events at 6/13/83 (left) and 8/26/81 (right)	24
Figure 11.  Probability of exceedance  of peak sediment discharge (kg/s),  total sediment yield
     (tons), and time to peak sediment discharge (min) for some selected parameters	25
Figure 12. Probability of exceedance of peak sediment discharge (kg/s) and total sediment yield
     (tons) for cf and cg parameters. Secondary axes are for cr2 and cg-2 (second event)	26
Figure 13. Effect of antecedent moisture condition on Ks sensitivity. Si is initial saturation, COVP
     and COVt are the coefficient of variations of peak sediment discharge and sediment yield,
     respectively	27
Figure 14. Computed and observed sedimentographs for selected events	28
Figure 15. Watershed conceptualization in GSSHA	32
Figure 16. Watershed conceptualization in KINEROS-2	32
Figure 17. Comparison of hydrographs generated with GSSHA (straight lines) and KINEROS-2
     (dashed lines) based on KINEROS-2 calibrated parameters.  Observed data is shown as
     hollow circles (Kalin and Hantush, 2003)	34
Figure 18. Comparison of hydrographs generated with GSSHA (straight lines) and KINEROS-2
     (dashed lines). GSSHA is recalibrated. Observed data is shown as hollow circles (Kalin and
     Hantush, 2003)	35
Figure  19.   Comparison of  sedimentographs  generated  with  GSSHA  (straight lines)  and
     KINEROS-2 (dashed lines). Observed data is shown as hollow circles (Kalin and Hantush,
     2003)	37
Figure 20. Rainfall histogram used in the long-term simulations of GSSHA	38
Figure 21. Observed (hollow circles) and simulated (straight line) hydrographs from the long-term
     simulations of GSSHA	39
Figure 22. Observed (hollow circles) and computed (straight line) sedimentographs from the long-
     term simulations of GSSHA	40
                                               VI

-------
                                         Tables

Table 1. Models selected for review after initial screening	12
Table 2. Loading Model Features	16
Table 3. Hydrodynamic Model Features	18
Table 4. Water Quality (Sediment/Nutrients) Model Features	19
TableS. Input parameters of KINEROS-2	22
Table 6. Summary statistics of G (cm) parameter for various soil types	23
Table?. Parameter set following calibration	27
TableS. Parameter sets used in KINEROS-2	33
Table 9. Total flows in m3 at the watershed outlet from observed data, and KINEROS-2 and
     GSSHA simulations with KINEROS-2 calibrated parameters	35
Table 10. Calibrated parameters with GSSHA	36
Table 11. Parameter values used in GSSHA long-term simulations	39
                                             vn

-------
                                  Acknowledgments
    The U.S.  Environmental Protection Agency through its Office of Research and Development funded
the research described here through in-house efforts and in part by an appointment to the Postgraduate
Research Program at the National Risk Management Research Laboratory administered by the Oak Ridge
Institute for Science and Education through an interagency agreement between the U.S. Department of
Energy and the U.S. Environmental Protection Agency. The report benefited from the constructive review
comments of Dr. Zhonglong Zhang, Dr. Gokmen Tayfur and Dr. Rao S.  Govindaraju.
                                            Vlll

-------
                                           1    Introduction
1.1     Overview

    Suspended solid and sediment (SSAS) yield has important implications for water quality and water resources. The
source of SSAS  can  be natural such as wind erosion,  upland erosion (detachment by rainfall and rill erosion),
stormwater runoff,  and bank erosion, or man-driven such as  wastewater  discharge, tillage, mining, construction,
silvicultural practices,  etc. Sediments may serve as carriers for pesticides, radioactive materials and nutrients giving rise
to water quality issues. Studies have shown that total suspended  sediment concentrations are positively related to total
phosphorus and nitrate concentrations. Nutrients, while essential for healthy aquatic systems, can have adverse effects at
low concentrations by increasing algal and macrophtye production and decreasing average dissolved oxygen. Stream
and waterbody water  quality is important not only for protection of fish and aquatic life, but it is often used as  an
indicator of the environmental health of a watershed.  Often,  SSAS in surface  waterbodies are contaminated  by
chemicals that tend to  sorb to fine-grained organic as well as inorganic soil particles. The sources of such contamination
can be from existing point or nonpoint sources (NFS) or from historical spills or discharges. When such contamination
exceeds critical levels,  they pose ecological and human health risks requiring appropriate  remedial actions.  Such
remedial actions take the form of either isolating the contaminated sediments, reducing their exposure  to other parts of
the ecosystem, complete removal of the contaminated sediment, or some combination of the above. Estimates of SSAS
yield are required for a wide spectrum  of problems dealing with dams and reservoirs, fate and transport of pollutants in
surface waters, design of stable channels, protection of fish and other aquatic life, watershed management and for
environmental impact statements. Figures 1  and 2 show typical processes  responsible for the transport and fate of
paniculate organic matter in waterbodies.

    Oxidation of organic matter occurs in the water column and in the bottom sediments. The deposition of algal mass
and paniculate organic matter on bottom sediments and decomposition therein exert sediment oxygen demand (SOD)
on the overlying water. Depletion of oxygen by oxidation of paniculate organic matter in the water column and by SOD
has undesirable environmental consequences, such as  loss of fishery. Figure  3 links the flux of paniculate organic
matter delivered to the sediments to SOD and sediment fluxes across the sediment-water interface.
Air-water V
int
i
erface
,

,


Alg


Labile
OC
r ,

^


ae-C
r


~
Other sources
ofPOC

T ir
Refra
O

Settling
r 1
*nn » DOC Respu

1
T
H J
SOD
r T ...

ation rn_



Sediment-water
                                           Sediment layer
                                                                                   interface
Figure 1. Carbon cycle.

-------
             Air-water
             interface
                                         Sediment layer
                                                                                interface
Figure 2. Nitrogen cycle.
i
1_
CD
-i— '
I,
Active sediment
1.
L
r
k
r

P(
flL
I

M
es

i
A K. //7/CW/C
Aerobic zone
Anaerobic zone
Production of NH^and CH4
SOD, h
CH4flu

|
H^and
es
1
Figure 3. Simplified schematics of sediment water interactions.
    The paniculate organic matters (POM) carried by water settles and within the anaerobic region decomposes to yield
dissolved CH4. The methane is later diffused upward to the aerobic zone and gets oxidized generating SOD. Similarly,
ammonification of organic N produces ammonium in the anaerobic zone which is later diffused to the aerobic zone
where it is nitrified to produce NO3" resulting in SOD.

    Changes in SSAS dynamics such as scour and erosion of channel bed and banks, deposition of fine particles, and
resuspension of solids in the suspended sediment load of the water column, can have significant effects on the aquatic
ecosystem health. Scouring and bank erosion may cause loss of habitat used for feeding,  reproduction, and cover by
fish, algae, birds etc. The consequences of deposition and resuspension are more obscure yet more significant (USEPA,
2002a). High suspended  sediment concentrations  increase the turbidity  in waterbodies that  can easily  alter  the
environment for phytoplankton and other aquatic flora from nutrient limited conditions to light limited conditions which
can eventually affect dissolved oxygen dynamics (Stanley 1994). The effects of high turbidity is more severe in the
more tranquil waters of lakes, reservoirs and estuaries than streams and rivers due to accumulation of suspended solids
in the water  column from multiple sources (USEPA, 2002a).

-------
    In the 1998 analysis of the U.S. impairment patterns, SSAS was determined as the leading cause of impairments of
rivers (USEPA, 2000). Further, in the same report sediment is listed as the third leading stressor in lakes, reservoirs and
ponds, where nutrients and metals were ranked first and second among other stressors. In a recent report known as "The
Twenty Needs Report", it is  stated that currently over 40 % of our assessed waters do not meet the water quality
standards set by states, territories and authorized tribes (USEPA, 2002b).
1.2     Total Maximum Daily Loading (TMDL)

    Section 303(d) of the Clean Water Act and Water Quality Planning and Management Regulations (40 CFR Part
130) are directly relevant to the total maximum daily load (TMDL) program as they interpret the statutory requirements
for states, territories and authorized tribes to list waterbodies that do not meet appropriate water quality standards. A
TMDL is defined as the maximum amount of pollutant that a waterbody can receive and still meet the water quality
standards. TMDLs include both the point source discharges and the nonpoint sources that arise from the watershed or
the environs of the watercourse  (Ward and Benaman, 1999). The Clean Water Act further  requires development of
TMDLs for all waters on the section 303 (d) list by developing restoration scenarios. The ultimate goal of a TMDL
development can be stated as removal  of the waterbodies from the 303(d)  list by attaining water quality standards.
Eventually, the list of impaired waterbodies and established TMDLs by states, territories and authorized tribes must be
approved by EPA.

    Since its introduction, there has been a tremendous amount of activity around  TMDL  programs. This, in turn,
brought many  opinions on the program's scientific needs from different sources including National Research Council
(NRC),  The EPA regional TMDL coordinators,  States  and  Tribes,  professional associations such  as the  Water
Environment Federation (WEF), non-governmental organizations and  private industry,  the Strategic Planning  and
Research Coordination (SPRC) research planners from EPA research and water offices, and  others (USEPA, 2002b).
The need to improve watershed and water quality modeling was among the recommended TMDL science needs in the
"Twenty Needs Report"  by the USEPA (2002b).
1.3     Mathematical Models

    Models are  extensively used by water resources planners, water quality managers, engineers and scientists to
evaluate the effectiveness of various control strategies. Mathematical models can help us understand the important
processes and interactions that affect the water quality of waterbodies. Further, they can be used in making decisions
regarding pollution control strategies by evaluating their effectiveness on water quality improvement and performing
cost-benefit analysis.

    It's worth noting that Novotny and Olem (1994) provide a diagram that compares the reliabilities of models of NFS
pollution. Based on that diagram, accuracy and reliability decrease with increased complexity and size  of the modeled
system. They list the hydrologic models simulating runoff from small, uniform and impervious surfaces as the most
accurate, and water quality models for large watersheds as the least reliable. The order of reliabilities of NFS models
decline as  follows: Hydrology with  impervious surface, hydrology,  sediment,  phosphates and metals, nitrogen and
organic chemicals, and bacteria.  The low uncertainty involved in the  hydrologic  and sediment transport models
compared to other processes, such as fate and transport of nutrients, definitely explains the high confidence associated
with them. In fact, this order of reliability becomes more discerning considering the fact that the physics used to
describe each process also decreases with the same order.

    The success in utilization of models in diverse fields has resulted in wide acceptance of models as an objective
evaluation tool and as a result they are often given higher credibility than what they actually deserve. Models  are only
approximate representations of the complex natural processes and due to time  and budget constraints involve many
assumptions made by the model creator who develops the  relationships and define the processes,  and  the  model
programmer who carries the model into computer platforms. Moreover, modelers usually simplify processes that are
seemingly not as important as other processes. Yet, this simplification might not be valid for other applications  due to
uniqueness of the problem and counter-intuitive results may be produced (AWWA, 2001). Modeling also  involves a
profusion of uncertainty. Macintosh et al.  (1994) defines two types of uncertainty: i) knowledge uncertainty and ii)
stochastic uncertainty.  The former is associated with measurement errors and inability  of the model to  accurately

-------
represent the physical, chemical and biological processes, and the latter arises from the random nature of natural
systems like rainfall and natural heterogeneity.  Any modeling application comprises both types of these uncertainties
implying that modeling cannot be deemed as  representing the absolute  truth. Therefore, care must be taken when
interpreting the results obtained through models. This clearly  calls for the need for implementing risk management
approaches to  TMDL allocation using Best Management Practices (BMP),  since model limitations, lack of perfect
knowledge of physicochemical  and biological processes, and inherent uncertainties  preclude  accurate,  risk-free
modeling approaches.  We elaborated on this later as we provide a probabilistic optimization framework as a proposal
for the solution of the BMP problem in general.
1.3.1   Brief History of Sediment Modeling

    Singh  and Woolhiser (2002)  provide  a historical perspective  of hydrologic modeling,  and  discuss  new
developments and challenges in watershed models. In that paper they date the origin of mathematical modeling back to
the rational method developed by Mulvany (1850) and an event model by Imbeau (1892) that relates the peak runoff
rate to rainfall intensity. The work of Streeter and Phelps (1925) may be treated as the first effort in water quality
modeling where the authors tried to address the relationship between dissolved oxygen in rivers and streams, and input
from domestic wastewater. The works of Velz (1938) and O'Connor (1960, 1962) are among the other early attempts in
water quality modeling. The earliest attempts in sediment modeling originated from relating soil loss from field plots to
slope  and steepness (Zingg, 1940). This work is extended by several researchers (Smith, 1941; Browning et al., 1947)
which led to the development of the famous Universal Soil Loss Equation (USLE) (Wischmeier and Smith 1958; 1965;
1978). Early models were based on simple one-dimensional, steady-state conditions. Advances in the theory of flow and
transport phenomena, and in computer technology elevated the art of sediment transport and water quality modeling as
time constraint was not a factor anymore. Development of fully dynamic, steady state, and three-dimensional water
quality models became feasible. The computational capability allowed the coupling of water  quality  models  with
watershed and hydrodynamic models. As a result, varieties of models have become available, and the choice of the right
model became a challenge. Selecting the right model  for a specific application depends on factors like type of the
stressors considered,  economic constraints such as time and labor, hardware, personal  experience and preferences,
hydrologic considerations, and scientific rigor and data availability.

    In the following sections  we classify sediment and nutrient water quality models and evaluate them based on
selected criteria. We use previously published material (eg. Shoemaker et al., 1997; USEPA, 1999; Ward and Benaman,
1999;  Tetratech,   2000;  SAAESD,  2001;  WERF 2001)   and related web  sites (eg.  USGS-SMIC  database:
http:smig.usgs.gov/smic,  Water Ways  Experiment Station (WES) models: http://www.wes.armv.mil/el/elmodels.
Register of Ecological Models  (REM) meta-database: http://eco.wiz.uni-kassel.de/ecobas.html) to  synthesize necessary
information. The goal of the evaluation process is  to provide  a list and summary of widely used sediment and nutrients
models and their ability to simulate for BMPs.
1.4     Risk Management Watershed Modeling

    The  Twenty  Needs Report  (USEPA, 2002b)  stresses improved ability to evaluate the effectiveness of Best
Management Practices (BMP) to manage,  among other stressors, suspended  solids  and sediments.  BMPs reduce
pollutant concentrations and loads in runoff by infiltration into the soil, physical infiltration by grass or other vegetation,
adsorption  on to  soil and plants, bacterial decomposition, plant uptake, and  sediment deposition (Komor,  1999).
Varieties of BMPs are available to trap sediments and control nutrients at the watershed scale varying from structural
such as wet and dry ponds, vegetative filter strips, riparian buffers, and wetlands to non-structural such as conservation
tillage, and improved fertilizer and animal-waste management (Figure 4).

    Models developed with BMP components are capable for allocating TMDLs in watersheds. The common practice
in the use of models for TMDL allocation is to evaluate alternative BMP scenarios using simulations based on trial and
error. There is no guarantee, however, that this approach can yield optimal results, as there is often frustratingly large
number of feasible solutions. Even when combined with efficient techniques and enormous computational effort, the
result may lead to a solution that is still far from the best possible. With increasingly powerful computers, an alternative
approach is to implement a system analysis in which the BMP problem can be cast in terms of an objective function
(e.g.,  cost of design and maintenance of BMPs) subject to TMDLs, physical,  legal,  technical, financial, and other

-------
constraints. In this case, the solution for the BMP selection problem involves the identification of several design and
operating variables related to the ensemble of alternative BMPs. These variables are referred to as decision variables
whose optimal values, which optimize the objective function (e.g., minimum cost), are to be determined (Louks et al.,
1981). A few studies, however, exist in the  literature which developed  methodologies to identify the optimal BMP
scenarios (eg. Udoyara et al., 1995; Mostaghimi et al., 1997; Zhen and Shaw, 2001; Srivastava et al., 2002). Most of
these studies rely on coupling a water quality model with an optimization algorithm. Mathematically, the optimal
solution for the BMP selection problem may be cast in this optimization framework
    Objective Function:
    Subject to
                                              Min  V C, (x)
                                                V    -^"^
 (1)



(2a)

(2b)
where Q is the cost corresponding to 1th BMP; x is the set of decision variables x1 associated with BMPs, both structural
and nonstructural; m is the total number of BMPs (structural and non-structural), gj(x) is the model generated value; and
aj and bj, respectively, are the upper or lower limits of the constraint j  (e.g., TMDL of sediment); and N is the total
number of constraints. Pollutants can have either lower or upper TMDL limits. For instance, sediment yield has an
upper limit, whereas total dissolved oxygen has a lower limit.
Figure 4. Simplified schematic of various BMPs at the watershed scale (adapted from USEPA 2002a).
    Model limitations and technical, economic, social, and political uncertainties pose a formidable challenge to the
application of suspended solids and sediment models, in fact, any other models, to risk management, especially at the
watershed scale. The above optimization problem is rigid because it requires strict validation of the constraints (2a and
2b). A more realistic,  risk-based approach is to acknowledge model imprecision, inherent uncertainties due to temporal
variability and spatial heterogeneity, and lack of precise knowledge of TMDL targets. In light of the uncertainties, strict
enforcement of the constraints (2a and 2b) may be redundant, perhaps too stringent of a requirement for realistic BMP

-------
planning problems. Instead, the approach should be a probabilistic one; that is, we acknowledge the uncertainties and
accept the risk involved in violating a  given constraint with a prespecified probability. The probability  that each
constraint would be violated constitutes an acceptable level of risk, whose value may be determined by water quality
managers,  regulatory agencies,  and other  stakeholders.  Probabilistically, the  above  optimization  model  can be
reformulated as follows
    Objective Function:
    Subject to
             Min   ]T C, (x)



Pr^J(x)aJ,
                                                                                                           (3)



                                                                                                          (4a)

                                                                                                          (4b)
where the model related function gj(x) is deterministic; and Aj and Bj are random variables whose distribution functions,
respectively, FAj (aj) and FBj (bj) are known. This problem is also referred to as chance constrained optimization (Louks
et al., 1981; Hantush and Marino,  1989). The chance constraint (4a) requires that the function gj(x) be no greater than
the random variable Aj with at least probability otj. Conversely, the chance constraint (4b) requires that the function
gj(x) be no less than the random variable Bj with at least probability Pj.
    The risk involved in satisfying condition (4a) is 1- otj, and for (4b) the risk is 1-
this chance-constrained problem can be shown to be
                                                                               . The deterministic equivalence of
    Objective Function:
    Subject to
                                              Min  Yc;(x)
                                                V    -^"^
                                 gj(x)
-------
    The above risk-based optimization approach may be suitable for problems involving multiple stressors (e.g., flow,
sediments,  dissolved  oxygen, and nutrients), where  multiple BMPs,  both structural and  nonstructural,  can be
implemented to achieve TMDL targets. To the best of the authors' knowledge no evidence exists in the literature which
suggests that this approach has been implemented for the solution of the BMP selection problem. Future research may
explore the use  of probabilistic, constraint optimization for the management of pollutant loads reduction, because such
an approach lends itself to risk-based management of stressors in watersheds.

-------
                                      2    Model Classifications
    Hydrology constitutes the most important component of any water quality model. For a water quality model, flow
distribution, both in time and space, is required. A model can have a hydrologic module and solve for the flow itself, or
the flow distribution can be supplied externally as input through  another hydrologic model. In either case, hydrologic
models  play a crucial role. Hydrologic models can be classified into various categories. For instance, they  can be
distinguished  as empirical vs.  physically based, deterministic vs.  stochastic (randomness),  lumped  vs. distributed
(spatial  variation), steady  state  vs. dynamic (time variation), and linear vs. non-linear. Empirical models are  usually
based on statistical relationships obtained through regression analysis of observed data. The problem  with  empirical
models  is that they are usually suitable for conditions under which the relationships have been developed. In other
words,  such models become less reliable under the  conditions outside the limit of the original environment  and
generally are not suitable for predictions under different conditions.  Physically based models, in contrast to  empirical
models, are based on physical principles such as conservation of mass and momentum. The input parameters of
physically based models can usually be obtained through field measurements. Deterministic models do not consider the
randomness involved in the data and always produce the same result for a given input parameter set, whereas stochastic
models  reflect the uncertainty in the data and may produce different output from the same input parameter set. Chow et
al.  (1988)  state this difference by  calling deterministic  models  as  forecasters  and stochastic models as predictors.
Lumped models usually consider the system as a black box and everything is spatially averaged over that single system.
Distributed models, to some extent, take into account heterogeneities by dividing the system into smaller units,  such as
cascade of planes in case of a  watershed. Such models assume  that the  model parameters and initial conditions are
uniform within each unit. Steady state models do not consider the variation of flow with time, contrary to dynamic
models. Linear models, such  as  the unit  hydrograph theory,  are based  on two  simple  principles:  principle of
proportionality and principle of superposition. The former can be stated as; if f(x) is a solution of a system, then c-f(x) is
also a solution of the same system with c being a constant.  The latter principle implies that if fi(x)  and f2(x) are both
solutions of the same system, then fi(x) + f2(x) is also a solution of the same system.

    Based on  how they function, suspended solids and sediments, and nutrients water quality models  can be broadly
categorized into three groups:

1.  Loading models: Models in this  group simulate field or watershed scale hydrologic processes  and determine the
    generation and transportation of SSAS and nutrients from source in the upper lands to the receiving  water. Loading
    models can be distinguished into agricultural, urban, or mixed categories based on land use.

2.  Receiving water models:  Again based on  the functionality, receiving water models can be divided into  two
    subclasses: hydrodynamic  and water quality models. Hydrodynamic models solve for  the hydraulics of water
    quality  models including transport, deposition, circulation and the stratification processes. Water Quality  models
    simulate the movement of SSAS  in the water column and determine the fate and transport of nutrients,  including
    eutrophication, in surface waters.  Sediments and paniculate organics are delivered to receiving models by loading
    models. Based  on the waterbody (Figure  6)  receiving  water models can be further subdivided into three
    subcategories:
        a)  Rivers and streams
        b)  Lakes and reservoirs
        c)  Estuaries

3-  Eutrophication/Ecological models: These models are a subclass of receiving water models. They relate biomass
    production (algae, crops, riparian vegetation) to nutrient loading.  Eutrophication models relate algal  production and
    growth in the waterbody to nutrient loading and photosynthesis.  They also include the sediment flux model. Refer

-------
    to Figures 2 and  3 and Chapter 1 for more  details. Figures 2 and 3 depict processes  typically  modeled  in
    eutrophication models.
                     Watershed
                  (Loading models)
                 Lake/reservoir
                              (Receiving water
                                  models)
Figure 6. Various waterbodies.
    The relationship between these groups of models is depicted in Figure 7. Models in each group can be stand alone
or they  may be coupled with other models. Often, hydrodynamic  and pollutant models are  integrated under the same
modeling system. This is called direct or internal linkage. If not under the same system, the output of the hydrodynamic
model such as water velocity, temperature, salinity, etc., may be fed externally into the pollutant model as input, called
indirect or external linkage. A detailed discussion on this topic is given in WERF (2001).
                                        Receiving Water Model
                                         - Rivers/Streams
                                         - Lakes/Reservoirs
                                           Estuaries/Bays
                                                                                    Eutrophication
                                                                                     /Ecological
                                                                                        Model
- Rivers/Streams
- Lakes/Reservoirs
- Estuaries/Bays
Figure 7. Relationship between different model groups.

-------
                                  3    Model Evaluation Criteria
3.1     Screening Criteria

    Transport and fate of sediments and nutrients  are intimately related phenomena, because suspended solids and
sediments (SSAS) include paniculate organic matter and serve as carriers for highly adsorbed phosphor. We therefore
consider in our evaluation water quality models, specifically those that can simulate nutrients in the environment. These
models are evaluated based on various criteria listed in the next section. A vast number of hydrologic and water quality
models is available ranging from heavily used ones to models with no users at all. We limit the focus of this evaluation
to models related to SSAS and nutrients. The following minimum criteria are used:
    1.   Capability of modeling SSAS
    2.   Good model documentation and model support
    3.   Proven record of application with sufficient history

    The first criterion limits the focus of this report to SSAS models. Those models that do not simulate SSAS were
excluded. The second minimum requirement is strong model support and a well-documented manual. Modelers should
be able to access the corresponding user manual and, preferably get technical assistance. The last constraint in the initial
screening is the acceptability of the model. The history of successful applications is a measure of acceptability of a
model.
3.2     Evaluation Criteria

    The models passing the initial screening are further appraised in detail based on the following criteria
    1.   Level of analysis:  screening or management
    2.   Rigor of processes i.e. level of sophistication
    3.   Spatial and time scale
    4.   Ease of use: preprocessing, post processing (GIS-GUI)
    5.   Hardware/software requirements
    6.   Data requirements
    7.   Linkage capabilities, adaptability
    8.   Model availability and cost
    9.   BMP evaluation, BMP costs

    Screening models are  relatively simple models and usually do not require much modeling expertise. They don't
account for spatial or temporal variability. They are mostly useful for a preliminary evaluation and can be used for
deciding whether a more  thorough  evaluation of the problem is  required or not. Default values usually suffice for
screening models and hence an extensive calibration/verification procedure is not justified. They are usually preferred in
the absence of data. On the other hand, planning and management models are much more  complex than screening
models. If the scope of a water quality  problem is identified, more complex management models can  provide a
comprehensive, more detailed analysis. They are preferred over screening models to answer 'what if' scenarios. Though
not necessarily, most of them can handle spatial and temporal variability.

    Rigor of processes refers to the soundness behind the theory used to develop the model. As described under model
classification  section, physically based models do rely on the physical laws and empirical models are usually derived
from observed data by  regression techniques.  Although subject to argument, the general consensus is that physically
based models are superior to empirical models, at least during the planning phase. For instance, Woolhiser  (1996)
                                                     10

-------
cautioned against overselling models. By referring to physically based models, he states that "...we should be able to
estimate the parameters a priori or measure them in the field, yet such estimates have a great deal of uncertainty.
Further, it is more difficult to calibrate physically based models because they are overparameterized'. There is no fully
physically based  sediment transport model. Wherever applicable, the accuracy  and stability of numerical solution
schemes used in the models form another basis for evaluating model robustness.

    Another norm used during the assessment of models is the spatial and temporal scales. Field scale models run over
a single overland plane. Watershed models require both overland flow planes and channels. On the other hand, the detail
of representation  of channels may vary from small  to large watershed models as channels  dominate flow in  large
watersheds,  whereas in small watersheds hydrology is still governed by overland flow. Models also differ in terms of
temporal scales. Some models only provide annual averages. For instance the USLE formulation is based on annual
sediment yield.  Some models are event based requiring very small time steps, sometime on the order of seconds. Large
time steps,  commonly  a day, usually suffice for  continuous  models, but not always. For example, when the full
Richards'  Equation option is employed in the GSSHA model, the required time step is well less than a minute if the size
of the grid meshes is small (less than 30 m).

    The required effort in using  a model depends on several factors. The first, and perhaps the most important factor, is
the  complexity  involved in the model. The availability of a Graphical User Interface (GUI) can drastically reduce the
input effort from a modeler's perspective. GUIs can help the user both in pre- and post-processing stages. Most models
nowadays offer GIS (Geographical Information Systems) interfaces which help extraction of model parameters  from
digital maps such as DEMs (Digital Elevation Models), soil maps, land use maps, etc. They can also be utilized in
interpreting model results visually.

    While computer cost has dropped drastically in the last decade, the hardware requirement may still be an issue for
the  user. For instance, some models only run on a UNIX platform which is generally available only in universities and
research institutes. This puts a severe limitation on number of potential users. Some models heavily rely on computer
power as they solve for full partial differential equations using numerical techniques, disregarding simplifications. This
necessitates  computers with fast processors (CPU) and large memories (RAM). Simple screening models can run on
almost any computer.

    The amount and type of required data might play  a significant role in model selection. In case measured data is not
available,  often input data can be gathered from literature for physically based models. On the other hand, it is hard to
make initial  guesses for empirical models and an exhaustive calibration/verification effort may be required.

    Model linkage is important for a comprehensive watershed analysis, especially for the evaluation of alternative
BMP scenarios. For instance, a water quality model which runs only on UNIX platform can only be linked to models
designed for UNIX platforms.  Similarly,  the output data of a loading  model must be compatible  with the  input
requirements of a hydrodynamic model.  The same is true between a hydrodynamic model and a water quality model. If
the  outputs  of the supplier model do not  involve all the  inputs of the receiving  model then  they can not be linked.
Examples of successful  model linkages  are  given  in the Tetra Tech  2000 report which summarizes sediment-
contaminant transport models. For example,  it  is reported that the water  quality model  CE-QUAL-ICM/TOXI is
designed to  be linked to hydrodynamic model CH3D-WES. Further EFDC can be linked to CE-QUAL-ICM and
WASPS.

    Model availability is a significant criterion in model selection. Some  models (most EPA  and USDA models) are
available free to public, yet some proven models such as MIKE-SHE require purchase of a license which may not be
affordable for some users.

    Last but probably the most desired feature of the  listed models within the context of this report is the capability of
simulating BMPs. Since this report focuses on review of models for risk management purposes, having a BMP
component is a preference.
                                                     11

-------
                           4    Model Selection and Comparisons
4.1     Model Selection

    Models or systems of models selected for review after the initial screening are listed in Table 1. Some models are
included in the list because of their promising futures despite short application histories. Some models appear multiple
times in the table,  since they have more than one component such as hydrodynamic and water quality (eg. MIKE-11
falls into all categories). Some of the models listed below are only Graphical User Interfaces (GUIs) which integrate
various models under the same umbrella and provide the linkages between them. BASINS and WMS are such modeling
systems.

Table 1. Models selected for review after initial screening.
                                                                Receiving
                Loading
        Hydrodynamic
          Water Quality
        (sediment/nutrient)
   AGNPS, AGWA, ANN-AGNPS,
   ANSWERS, ANSWERS-2000*,
   BASINS, EPIC, DWSM*, GLEAMS,
   GSSHA, GWLF, HSPF, KINEROS-2,
   MIKE-11, MIKE-SHE, OPUS, PRMS,
   REMM*, SWAT, SWMM,
   VFSMOD*, WEPP,
   WMS(HSPF,GSSHA)
CE-QUAL-RIV1, CE-QUAL-
W2, CH3D-WES, DELFT3D,
DYNHYD5, EFDC, MIKE-11,
MIKE-2LMIKE-3
CE-QUAL-ICM, CE-QUAL-
ICM/TOXI, CE-QUAL-R1, CE-
QUAL-RIV1, CE-QUAL-W2,
CH3D-SED, DELFT3D, EFDC,
HSPF, MIKE-11, MIKE-21, MIKE-
3, QUAL2E, WASP5
 Models having insufficient application history but are very promising

    Features of each model are summarized in a tabular format in Tables 2, 3 and 4. These tables provide a summary of
each model's attributes. SC in the tables, under the platform category, refers to availability of the source code which can
be compiled on any platform and used accordingly. Model linkages in Table 2 are divided into two categories i) Linked:
means such a link already exists, and ii) Potential: means either work is under progress for model linkages or the models
are compatible and can be linked in future. Description of each model's  features  and capabilities are given in the
Appendix. It should be noted that model summaries are based on model manuals and other available literature (eg.
Shoemaker et al., 1997; USEPA 1999; Ward and Benaman, 1999; Tetratech 2000; WERF 2001, SAAESD 2001).
4.2     Evaluation and BMPs Capabilities
    Table 2 lists capability of models to simulate BMP features. Among the models reviewed the USDA's AGNPS
model appears to offer the most comprehensive BMP  simulation capability  (agricultural practices, ponds,  grassed
waterways,  irrigation, tile drainage,  vegetative  filter strips  and riparian buffers) to  the user.  Tillage  effects, soil
consolidation, residue decomposition etc. are considered within the Revised Universal Soil Loss Equation (RUSLE).
The impoundment module uses a modified sediment deposition algorithm. It is modified to reflect the simplifications
associated with small impoundments with restricted pressurized outflow and/or some permanent pool storage. These
simplifications are i) constant transport discharge equal to a constant outflow; ii) zero sediment transport capacity for all
                                                  12

-------
sediment sizes; and iii) dilution of the incoming water-sediment mixture by the permanent pool storage. AGNPS is
suited for agricultural watersheds. Its major drawback, however, is its semi-empiricism. It can be used for both event
and continuous simulations.  Numerous applications of AGNPS are found in literature, perhaps due to its ability to
model various BMPs.

    SWAT is another widely accepted continuous simulation model suitable for large agricultural watersheds (>100
km2), however it is also semi-empirical.  It has the ability of simulating surface flow, subsurface flow, sediment, and
nutrients in addition to various BMPs (agricultural practices, ponds, tile drains). Management practices are handled
within the Modified Universal Soil Loss Equation (MUSLE). SCS curve numbers can also be varied throughout the
year to  taker into  account variations in the management conditions.  SWAT  divides the watershed into Hydrologic
Response Units (HRU) that has uniform properties. Edge-of filter strips may be defined in an HRU. The filter strip
trapping efficiency for sediment is calculated empirically as a function of the width of the filter strip. When calculating
sediment movement through a water body, SWAT assumes the system is completely mixed. Settling occurs only when
the sediment concentration in the water body exceeds the equilibrium sediment concentration specified by the user. The
sediment concentration at the end of a day is determined based on an exponential decay function. SWAT also simulates
the buildup  and washoff mechanisms similar to SWMM model. SWAT has  its own GIS interface and currently
integrated into USEPA's  BASINs and USDA's AGWA modeling systems. SWAT is also linked to the  water quality
model QUAL2E.

    The WEPP model probably  has the most mechanistic sediment transport  conception, but it has received little
application outside the National Soil Erosion Research Laboratory  staff. It  can simulate various BMPs including
agricultural practices (e.g. tillage, contouring,  irrigation, drainage, crop rotation, etc.), ponds, terraces, culverts, filter
fences and check dams. Soil erosion is represented in two ways for WEPP overland flow profile applications: i) soil
particle  detachment by raindrop impact and transport by sheet flow on interrill areas (interrill delivery rate), and ii) soil
particle  detachment, transport and deposition by concentrated flow in rill areas (rill erosion). Effect of different
agricultural  management practices is reflected with soil  detachment parameters.  Deposition  of sediments  in
impoundments is  calculated by  assuming  complete  mixing and  later  adjusted to account for stratification, non-
homogeneous concentrations and the impoundment shape. It is  applicable to very small watersheds. SWAT, AGNPS
and WEPP are all available free to public.

    The DHI's MIKE-SHE watershed model is physically based, comprehensive with a history of applications in peer
reviewed journals.  MIKE-SHE includes virtually all of the processes in the land phase of the  hydrologic cycle with
several  BMP  options including wetlands,  nutrient and  pesticide  management, etc. MIKE-SHE can be used  in
combination with MIKE-11 for river hydraulics. This modeling package, however, is proprietary.

    For urban areas, the  most complete loading model is the widely used SWMM model. Modelers can simulate all
aspects  of the urban hydrologic and quality cycles,  including rainfall, snowmelt, surface and subsurface runoff, flow
routing through the drainage network, storage and treatment. SWMM is structured in the form of blocks. Infiltration can
be computed by Green-Ampt or Horton's  equations. Kinematic wave  routing is used in the transport block. For
hydraulic flow routing complete Saint Venants' equations are used. Detention basin simulations and street cleaning are
the available BMP  alternatives. Using SWMM  requires high expertise. SWMM outputs can be directed to  the USEPA's
WASP6 receiving water model.

    For large watersheds comprised of both urban and rural areas HSPF is the most suitable model to address the
sediment and nutrient TMDL problems.  The BMP components of HSPF can be  listed  as: nutrient and pesticide
management, urbanization and ponds. HSPF  employs the same  algorithms for sediment transport in  reservoirs  as
rivers/streams. Deposition or scour of cohesive sediment is calculated based on the bed shear stress. Whenever shear
stress is less than  the user-supplied critical shear stress for deposition, deposition  occurs; whenever shear stress is
greater than the user-supplied critical shear stress for scour, scouring of cohesive bed sediments occurs. The rate of
deposition is given by simplified Krone's equation (1962) which is a function of settling velocity (user defined), current
sediment concentration, shear stress and critical shear stress. Like SWMM, HSPF is freely available to the public.

    GLEAMS can be utilized for simple screening analysis over field  scale agricultural areas where different
agricultural practices, irrigation and ponds can be simulated as  alternative BMPs. Hydrology, erosion/sediment yield,
pesticide transport  and nutrients are the four major components of GLEAMS. USLE formulation is implemented for
computation of erosion. It is publicly available.
                                                     13

-------
    The KINEROS-2 model is suitable for event based studies over small watersheds. It is one of the two models in the
AGWA modeling system. Model performances reported in the literature (see model detail in appendix) are impressive.
Different agricultural practices, detention basins and culverts can be listed as the BMP options available in KINEROS-
2. Effect  of different agricultural management practices on the sediment transport is reflected by splash and hydraulic
erosion parameters. Pond sedimentation in KINEROS-2 is similar to that for tank sedimentation Particle fall velocities
and flow-through velocities are used to find the trajectories that intersect the reservoir bottom. Particle fall velocities are
calculated for each particle size  class. Suspended and slowly falling particles are subject to molecular diffusion and
dispersion. With the addition of an evapotranspiration component, it can be used for continuous time simulations.

    GSSHA is another promising model. Its flow component is fully physically based and has a proven applications
trackrecord (see references given in model details in the appendix), whereas the sediment component is semi-empirical.
On the other hand, the sediment component is currently being reformulated based on physics based sediment transport
concepts.  In its current version, the sediment transport formulation is based on the USLE  soil parameters.  Thus,
agricultural management practices can be listed as the GSSHA's BMPs.  US Army Waterways Experiment Station
(WES) supports the model and it is incorporated into the WMS modeling system.

    Most of the agricultural areas with low slopes especially in the Midwest contain tile drains. In addition to SWAT
and AGNPS, the newly developed DWSM model presents a promising future for development of BMPs  in tile-drained
watersheds. It is a physically based and event model capable of simulating surface and subsurface flows,  sediments and
agrochemicals in tiled-drained agricultural watersheds. Detention basins, alternative ground covers and tile drains can
be listed as its BMP component. The source code is in FORTRAN and is freely available.

    REMM and VFSMOD are two  field scale models being able to route flow and sediment through riparian buffers
and vegetative filter strips, respectively. REMM is suitable  for long-term simulations and VFSMOD is event based.
REMM simulates movement and storage of water within riparian buffer systems by a process-based, two-dimensional
water balance operating on a daily time step. Sediment transport is simulated both in channels and overland flow areas,
but channel erosion or detachment is not simulated. Because of the roughness of the riparian buffers, it is assumed that
sediment transport is primarily of suspended particles. Upland loadings are assumed to be  provided as input  to the
REMM. Overland flow erosion is based on the USLE equation. Five classes of sediment are considered: sand, large
aggregate, small aggregate, silt and clay. Sediment load computations are performed for each of these classes.  Steady
state continuity  equation is used to  compute the  sediment at the downslope edge. VFSMOD considers that during a
rainfall/runoff event, field runoff reaches the upstream edge of the filter with  time dependent flow rate and sediment
load. The vegetation produces a sudden increase in hydraulic resistance that slows the flow, lowers its transport capacity
and produces deposition of the coarse material (particle diameter dp >0.0037 cm) carried mostly as bed load transport.
The trapped bedload forms a trapezoidal shape.  Suspended load zone follows this  zone. The calculation procedure
utilizes a modified  Manning's open channel flow equation, continuity equation, and Einstein's  sediment bed load
transport  function. The sediment trapping algorithm for the suspended load zone follows Tollner et al. (1976) equation
based on a probabilistic approach to turbulent diffusion for non-submerged flow. REMM and VFSMOD can be  linked
to appropriate  watershed models to analyze  sediment transport and potential trapping through  riparian buffers  or
vegetative filter strips  in detail. REMM is already being linked to ANNAGNPS and has the potential to be linked to
SWAT. VFSMOD can potentially be linked to KINEROS-2.  The  receiving  water  models CE-QUAL-RIV1, CE-
QUAL-W2, DELFT3D,  EFDC, MIKE-21 and MIKE-3 have both hydrodynamic and water quality components, and
they can  be run as standalone programs  if they are linked to a loading model.  Within these models DELFT3D and
MIKE models are proprietary.

    In spite of its one-dimensional, steady-state flow component, QUAL2E is a widely used water quality model for
streams and rivers. Although it is not suited for sediment transport, it simulates for paniculate organic matter; therefore,
can be linked to watershed loading models to evaluate the impact of BMPs on transport and fate of nutrients in surface
waterbodies. QUAL2E is relatively simple  and easy  to  use. This  model is integrated into the USEPA's BASINS's
system where it is coupled with a watershed model which provides flow data to  QUAL2E. A linkage between QUAL2E
and SWAT is also available. CE-QUAL-W2, a 2-D model, has a complete eutrophication module which is suitable for
deep lakes and reservoirs. If linked to  a loading model, CE-QUAL-W2 can be  used to assess impacts of various BMP
scenarios on the state of eutrophication in surface waterbodies.

    For large, complex waterbodies where 3-D consideration is  important, EFDC or WASP6  can be used for sediment
and nutrient analysis. Momentum and conservation equations form the basis of governing hydrodynamic equations of
EFDC. The sediment routine used in EFDC is relatively unsophisticated. Both cohesive and non-cohesive sediments
                                                    14

-------
can be simulated. User is given the option to select number of sediment size classes. Problems that have been studied
using WASP6 include biochemical oxygen demand, dissolved oxygen dynamics, nutrients/eutrophication,  bacterial
contamination, and toxic chemical movement.  The WASP6 system consists  of two stand-alone computer programs,
DYNHYD5 and WASP6 that can be run in conjunction or separately. WASP has been linked to the hydrodynamic
models DYNHYD5, EFDC and CH3-WES. The SWMM outputs can be directed to the WASP6 as well.

    The HSPF model is a full-scale simulation model that can be applied to large watersheds containing both urban and
rural areas, streams, rivers, lakes and reservoirs to assess the effects of land-use change, reservoir operations, point or
nonpoint source treatment alternatives, flow diversions, etc. It has been widely used for TMDL studies and watershed
planning. However, it is a very complex model requiring high level of knowledge of watershed processes.  The source
code written in F-77 is freely available and can be compiled and used on any platform. It is also part of the USEPA's
BASINS modeling system and has been incorporated into the WMS modeling environment. MIKE-11 is another full-
scale and complex simulation model capable of simulating, among others, sediment transport in estuaries, rivers, and
other inland waters. It has a module for automated model calibration that uses the state of the art global optimization
routine called the  Shuffled Complex Evolution (SCE). MIKE-11 has a fully integrated interface in the Arc View GIS
that facilitates input data  preparation  and output visualization. The inclusion of MIKE-11 by  The US Federal
Emergency Management Agency (FEMA) on their list of hydraulic models  accepted for use  in the National Flood
Insurance Programme (NFIP) shows its credibility. Like other DHI products, license purchase is necessary.

    USEPA's BASINS is a complete modeling system which has loading (SWAT and HSPF), and  stream and river
water quality (QUAL2E and HSPF) models.  The system provides the  linkages between  these models within an
Arc View environment to simulate for sediments and nutrients. EPA is  also working on expanding BASINS to include
the 3-D water quality model EFDC. WMS is another modeling system which incorporates HSPF and GSSHA models at
this stage. WMS is an effective and easy to apply modeling system for  runoff and sediment yield analysis. AGWA is a
GIS-based hydrologic modeling tool. It is an  Arc View  3.X extension within which spatially-distributed  data are
collected and used to prepare model input files and evaluate model results for SWAT and KINEROS models. For event-
based studies over  small watersheds (<100 km2) KINEROS is recommended and for long-term,  continuous-time
simulations over large watersheds (>100 km2) SWAT is utilized.

    The information given thusfar can be used to select group of candidate models based on qualitative comparisons.
To further decide on the optimal model a more quantitative comparison might be necessary. In the following two
chapters such an exercise  is presented. Two  distributed,  hydrologic and sediment transport models, the Kinematic
Erosion Model (KINEROS) and GSSHA, are applied to an experimental watershed. We conduct sensitivity  analysis,
calibrate and verify both models, and evaluate their performances. Both models are commonly used and are promising
with many applications in peer  reviewed literature. GSSHA is supported by Waterways  Experiment Station and is
embedded into the WMS modeling system. KINEROS is developed by USDA scientists and is one of the two models
under the AGWA modeling system which is supported by both USDA and USEPA.
                                                    15

-------
Table 2. Loading Model Features.

AGNPS/AnnAGNPS
AGWA (KINEROS-2)
AGWA (SWAT)
ANSWERS
ANSWERS-2000*
BASINS (HSPF)
BASINS (SWAT)
DWSM*
EPIC
GLEAMS
GSSHA**
GWLF
HSPF
KINEROS
MIKE-11
MIKE-SHE
OPUS
PRMS
RE MM*
SWAT
SWMM
VFSMOD*
WEPP
WMS (HSPF)
WMS (GSSHA)
Field-F
Agricultural
watershed-A
Urban
watershed-U
A
A, U
A
A
A
A, U
A
A
F
F
A, U
A
A, U
A, U
A
A
F
A
F
A
U
F
A
A, U
A, U
Level of Analysis
Screening-S
Detailed-D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
D
S
S, D
S, D
S
D
D
S, D
S, D
S, D
S, D
D
S, D
S, D
D
Rigor
Empirical-E
Semi-Empr.-S
Phys. Based-P
E
P
S
P
P
P
S
P
E
E
P
E
P
P
P
P
P
P
S
S
P
P
P
P
S
Spatial Scale
Lumped-L
Distributed-D
D
D
D
D
D
D
D
D
L
L
D
L
D
D
L
D
D
D
L
D
D
D
D
D
D
Temporal Scale
Event-E
Continuous-C
E, C
E
C
E
C
C
C
E
C
C
E, C
C
C
E
E, C
E, C
C
E, C
C
C
E, C
E
E, C
C
E, C
Level of Effort
Low-L
Medium-M
High-H
M-H
M-H
M
M-H
M-H
M-H
M-H
M
M
M
H
M
M-H
M-H
H
H
M
M-H
M
M
H
M
M
M-H
M-H
Platform"
WIN/SC, AV
WIN, AV
WIN, AV
DOS/SC
WIN, AV
WIN, AV
WIN, AV
SC
DOS/UNIX
DOS/SC
DOS
WIN, AV
DOS
DOS/WIN/SC
WIN, AV
WIN, AV
DOS
DOS/UNIX/SC
WIN
WIN, AV
DOS/SC
DOS/WIN/SC
Wl IN/DOS
WIN
WIN
Availability
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Proprietary
Public
Public
Public
Proprietary
Proprietary
Public
Public
Public
Public
Public
Public
Public
Proprietary
Proprietary
        * Models having insufficient application history but are very promising
        x SC = Source Code, AV =ArcView, AI = Arclnfo, WIN = WINDOWS
                                                                               **
Flow is physically based, sediment transport is semi empirical
                                                                        16

-------
Table 2. Loading Model Features (continued).

AGN PS/An nAGN PS
AGWA
ANSWERS
ANSWER2-2000
BASINS
DWSM
EPIC
GLEAMS
GSSHA
GWLF
HSPF
KINEROS-2
MIKE-11
MIKE-SHE
OPUS
PRMS
REMM
VFSMOD
SWAT
SWMM
WEPP
WMS
Linkage
Linked

KINEROS, SWAT


SWAT, HSPF, QUAL2E

GLEAMS
EPIC
WMS

BASINS,WMS
AGWA
MIKE-SHE
MIKE-11




AGWA, QUAL2E, BASIN


HSPF, GSSHA
Potential
REMM



EFDC





CE-QUAL-W2
VFSMOD




AGNPS, SWAT
KINEROS-2
REMM
WASP


BMP
Agricultural practices, ponds, grassed waterways, tile drainage, vegetative filter strips, riparian buffers
See SWAT and KINEROS-2
Agricultural management, ponds, grassed waterways, tile drainage
Agricultural management, ponds, grassed waterways, tile drainage
See SWAT and HSPF
Detention basins, alternative ground covers, tile drains
Agricultural practices
Agricultural practices, ponds, irrigation
Agricultural practices
Agricultural practices, septic systems, manured areas
Nutrient and pesticide management, ponds, urbanization
Agricultural practices, detention basins, culverts

Agricultural and forest practices, wetlands, nutrient and pesticide management, irrigation, drainage
Terraces, contours, furrows, grassed buffer-strips or waterway, and farm ponds

Agricultural practices, riparian buffers
Vegetative filter strips
Agricultural practices, ponds, tile drains
Detention basins, street cleaning
Agricultural practices, ponds, terraces, culverts, filter fences, check dams
See HSPF and SWAT
* Agricultural practices may include: tillage, irrigation, drainage, nutrient and pesticide management, crop management, crop rotation, grazing etc.
                                                                             17

-------
Table 3. Hydrodynamic Model Features.

CE-QUAL-RIV1
CE-QUAL-W2
CH3D-WES
DELFT3D
DYNHYD5
EFDC
MIKE-11
MIKE-21
MIKE-3
Dimension
1-D
2-D
3-D
3-D
1-D
3-D
1-D
2-D
3-D
Waterbody
Stream-S
River-R
Lake/Res.-LR
Estuary-E
Coastal-C
S, R
S, R, LR, E
S, R, E, C
S, R, LR, E, C
S, R, E
S, R, LR, E, C
S, R, E
R, LR, E, C
R, LR, E, C
Level of
Analysis
Screening-S
Detailed-D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
Rigor
Empirical-E
Phys. Based-P
P
P
P
P
P
P
P
P
P
Stead y-S
Unsteady-U
U
U
U
U
U
U
U
U
U
Level of Effort
Low-L
Medium-M
High-H
M-H
H
H
M-H
M
M-H
M-H
M-H
M-H
Platform
(SC=Source
Code
available)
&GIS
sc
sc
UNIX
WIN
DOS/WIN
SC
WIN, AV
WIN
WIN
Availability
Public
Public
Public
Proprietary
Public
Public
Proprietary
Proprietary
Proprietary
Water Quality Model
Linkage


CE-QUAL-ICM,
WASPS

WASP6
WASP6, CE-QUAL-
ICM



                                                                            18

-------
Table 4. Water Quality (Sediment/Nutrients) Model Features.

CE-QUAL-ICM
CE-QUAL-ICM/TOXI
CE-QUAL-R1
CE-QUAL-RIV1
CE-QUAL-W2
CH3D-SED
DELFT3D
EFDC
HSPF
MIKE-11
MIKE-21
MIKE-3
QUAL2E
WASP6
Dimension
3-D
3-D
1-D
1-D
2-D
3-D
3-D
3-D
1-D
1-D
2-D
3-D
1-D
3-D
Waterbody
Stream-S
River-R
Lake/Res.-LR
Estuary-E
Coastal-C
S, R, LR, E, C
S, R, LR, E, C
LR
S, R
S, R, LR, E
S, R, E, C
S, R, LR, E, C
S, R, LR, E, C
S, R, LR
S, R
S, R, LR, E, C
S, R, LR, E, C
S, R
S, R, LR, E, C
Level of
Analysis
Screening-S
Detailed-D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
S, D
Rigor
Empirical-E
Phys. Based-P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
Steady/
Unsteady
U
U
U
U
U
U
U
U
U
U
U
U
U
U
Level of Effort
Low-L
Medium-M
High-H
M-H
H
M
M-H
H
M-H
M-H
M-H
M
M-H
M-H
M-H
L-M
M-H
Platform
(SC=Source
Code
available)
&GIS
DOS/SC
DOS/SC
DOS/WIN/SC
SC
WIN/SC
UNIX
WIN
DOS/SC
DOS/WIN
WIN, AV
WIN
WIN
DOS/WIN/SC
DOS/WIN
Availability
Public
Public
Public
Public
Public
Public
Proprietary
Public
Proprietary
Proprietary
Proprietary
Proprietary
Public
Public
Hydrodynamic Model
Linkage
EFDC, CH3D-WES
EFDC, CH3D-WES





WASPS, CE-QUAL-
ICM
BASINS, WMS



BASINS, SWAT
CH3D-WES,
DYNHYD5, EFDC,
RIVMOD, SWMM
                                                                                 19

-------
      5    Modeling of Sediment Yield in a Small Agricultural Watershed with
                                              KINEROS-2
    Distributed models are favored over lumped ones for detailed TMDL developments and BMP implementations.
The availability of high power computers has relaxed the burden of long simulation times. Among the distributed
models, the physically-based ones are generally preferred over empirical ones, since model parameters have physical
meaning and can be measured in the field. When measurements are not available, model parameters can be still be
deduced from published data in literature based on topography, soil and land use maps. Where flow is concerned, to our
knowledge three  models seem to be the  most physically based with proven history, and  separate themselves from
others: GSSHA (Downer and Ogden 2002), KINEROS-2 (Smith et al. 1995) and MIKE-SHE (Refsgaard and Storm
1995).

    Calibration is a very time  demanding process  and is a  prerequisite  before using complex models with many
parameters. Most physically based and distributed models require enormous amount of input data. Although some
parameters play crucial roles, some have minimal effect on model results. Therefore, it is a common practice to perform
sensitivity analysis before calibrating model parameters. In doing so, the number of parameters to be calibrated can be
reduced drastically and only most sensitive parameters are calibrated while average values can be used for the rest of the
parameters. The sensitivity of KINEROS-2 to various input parameters was evaluated in this section through Monte
Carlo (MC) simulations. Based on the sensitivity analysis, the model parameters were calibrated and then validated over
several events.  In the following chapter we examine and compare KINEROS-2 and GSSHA for their performances on
modeling flow and sediment movement.
5.1     Model Background:

    KINEROS-2 is a distributed, event-oriented, physically based model describing the processes of surface runoff and
erosion from small agricultural and urban watersheds (Woolhiser et al., 1990). The watershed is represented by cascade
of planes and channels, in which flow and sediments are routed  from one plane to the other and, ultimately, to the
channels. The elements (planes or channels) allow rainfall, infiltration, runoff, and erosion parameters to vary spatially.
This model may be used to determine the effects of various artificial features such as  urban development, small
detention reservoirs, or lined channels on flood hydrographs and sediment yield.

    When rainfall rate approaches the infiltration capacity, Hortonian overland flow begins. KINEROS-2 assumes one-
dimensional  flow in each plane and solves  the kinematic  wave approximation of the overland and channel flow
equations using finite differences. The flow rate is related to the  channel flow cross-sectional area or overland flow
depth through Chezy and Manning flow resistance relationships.  In these  relationships  the  channel or bed slope
approximates the friction slope.
    Sediment transport equation is described by the following mass balance equation:
                              |-(AC)+JL(QC)-e(x,t) = qs(x,t)                                       (7)
                              <3t       ox

in which C is the volumetric sediment concentration [L3/L3]; A is the channel cross section area [L2]; for overland flow
it is equal to the flow depth h for a unit flow width [L]; Q is the is the channel discharge [L3/T]; for overland flow it is
equal to the discharge per unit width [L2/T]; e is sediment erosion rate [L2 /T] given below; and qs is the rate of lateral
                                                   20

-------
sediment inflow for channels [L3/T/L]. In KINEROS-2 Sediment erosion/deposition rate e is  composed of rainfall
splash erosion rate gs and hydraulic erosion rate gh:

                                           e = gs+gh                                                    (8)

Rainfall splash erosion is given by (Woolhiser et al., 1990)

                                    gs =cf e~Chh rq;    q>0                                            (9)
                                      = 0;              q<0

in which cf is a positive constant [T]; h is flow depth [L]; ch is damping coefficient for splash erosion [I/1]; r is rainfall
rate [L/T]; q is excess rainfall (rainfall rate minus interception minus infiltration) [L/T]. The exponential term represents
the reduction in splash erosion caused by increasing depth of water (Smith et al. 1995). In channel flow, this term is
usually equal to zero: the accumulating water depth absorbs nearly  all the imparted  energy by the raindrops. The
hydraulic erosion represents the rate of exchange of sediment between the flowing water and the soil  over which it
flows. Such interplay between shear force of water on the loose soil or channel bed and the tendency of the soil particles
to settle under the force of gravity may  be described by this first-order rate expression:

                                        gh=cg(C*-C)A                                                (10)

where C* is the volumetric concentration at equilibrium transport capacity [L3/L3]; cg is a transfer rate coefficient [T"1].
For sheet flow A = h. This relationship assumes  that if C exceeds equilibrium saturation, C*, deposition occurs. cg is
usually very high for fine, noncohesive material, and very low  for cohesive material. Several expressions for C* are
available from literature (see, e.g., Woolhiser et al. 1990). In our analysis, we used  Engelund and Hansen (1967)
formula.

    Successful applications of KINEROS-2 and its older version KINEROS to gaged watersheds has been reported in
the literature (Osborn and Simanton 1990, Goodrich et al. 1994, Smith et al.  1999, Ziegler et al. 2001, Kalin et al. 2003,
and Kalin and Hantush 2003 etc.).
5.2     Data and Model Parameters

    A small USDA experimental watershed (W-2) located near Treynor, Iowa having an area of 83 acres was employed
in this study (Figure 8). Measurements of runoff and sediment load are available. There are two rain gauges (115 and
116) around the watershed. W-2 has a rolling topography defined by gently sloping ridges,  steep side  slopes, and
alluvial valleys with incised channels that normally end at an active gully head, typical of the deep loess soil in MLRA
107. Slopes usually change from 2  to 4 percent on the ridges and valleys and 12 to  16 percent on the side slopes. An
average slope of about 8.4 percent is estimated, using first-order soil survey maps. The major soil types are well drained
Typic Hapludolls, Typic Udorthents, and Cumulic Hapludolls (Marshall-Monona-Ida and Napier series), classified as
fine-silty, mixed, mesics. The surface soils consist of silt loam (SL)  and silty clay loam (SCL) textures that are very
prone to erosion, requiring suitable conservation practices to prevent soil loss. Corn has been grown continuously on W-
2 since 1964.
5.3     Sensitivity Analysis and MC Simulations

    Sensitivity of KINEROS-2 was performed over the parameters  listed in Table 5. In the table Ks is saturated
conductivity, A, is pore size distribution index, *¥b is bubbling pressure,  G is net capillary drive, § is porosity,  S; is initial
saturation, nch and np are channel and plane Manning's  roughness, respectively, I is the interception depth, CAN is
canopy percentage, cg is the transfer rate coefficient, cf is rainsplash coefficient and d50 is the mean particle diameter.
One  thousand random values were generated for each parameter. The ranges of parameters from which the random
numbers were generated are shown in Table 5 for two soil types (SL  and SCL). KINEROS manual (Woolhiser et al.,
1990) suggests values and puts limits for cg and cf. During calibration, however, we found values outside the margins. In
                                                     21

-------
a similar study, Smith et al.  (1999) estimated even larger values for these two parameters during the calibration of
Catsop Catchment. After confirming with one of the model developers (C. Unkrich, personal communication) it was
decided not to limit ourselves to the values given in the manual. The random values for the parameters Ks, X, *Pb and $
were generated from log-normal distributions  using IMSL  routine,  where the corresponding  mean and  standard
deviations are given respectively in parentheses in Table 5. The parameter *Pb is not required by KINEROS-2 but used
here to generate random G values as described below. The rest of the parameters were generated from uniform
distributions.
Figure 8. Schematic of W-2 watershed.
Table 5. Input parameters of KINEROS-2.

SL
SC
L



Ks (mm/hr) a
log(4.5,12.3)
log(0.7,1.9)


SL
SCL
Xb T
log(0.23,0.13) lo|
log(0.18,0.14) loj

V ie
0.01-1 0-3

b (cm) c
5(51,59)
5(70,74)

CANe
0-1.0

G (cm) d
0.2-694
0.7-7380

cge
0.01-1
0.01-1

log(0
log(0


.00
.00
i b
4:
.50,0.08)
.47,0.05)

cfe
100-1000
100-1000
Si"
0.03-0.97
0.08-0.92

b
dso (M-in)
3-50

nche
0.01-1




                US EPA/600/R-93/046, 1993. PRIZM-2 Users Manual for Release 2.0
                KINEROS Manual (Woolhiser et al., 1990)
                Rawlsetal., 1982
                From G=xFb(2+3V(l+3X)
                Randomly decided
                                                    22

-------
    The net capillary drive parameter, G is defined as
     u

G=  J[K(H/)/KJdH/
                                                                                                        (U)
Using the Brooks-Corey soil characteristic relation for unsaturated conductivity  K(y) = Ks (\|/b /\\/)2+3X leads to the
simple expression
Rawls et al. (1982) indicated that \\ib and A, are log-normally distributed; they provided the arithmetic and geometric
mean values with the corresponding standard deviations for both parameters, for different  texture class.  Over the
reported range of values for X, we have this approximation (Hantush and Kalin, 2003)
and
                                                                  (13)


                                                                  (14)
Thus, G is lognormally distributed, with the mean of InG (i.e., geometric mean) given by (14) and variance of InG
CThn|/  ' which is the variance of ln\|/b. A, is the geometric mean of X. Rawls et al. (1982) provide values of cr^   and A,

for different soil textures. Table 6 (Hantush and Kalin, 2003) provides the arithmetic mean and standard deviations of G
for different soil  textures  obtained from  the  lognormal approximation  and by  performing 10000  Monte Carlo
simulations, using the statistics of the lognormally distributed \\ib and A,  (Rawls et al.  1982). It is striking that the
suggested G values in the KINEROS-2 manual are much smaller than the values shown in Table 6.
Table 6. Summary statistics of G (cm) parameter for various soil types.
Arithmetic
mean
Soil Texture
Sand
Loamy sand
Sandy loam
Loam
Silt loam
Sandy clay
Clay loam
Silty clay loam
Sandy clay
Silty clay
Clay
theoretical
39
41
64
105
158
181
129
195
219
209
242
MC
40
44
62
112
156
180
129
183
224
204
232
std.
theoretical
118
131
186
475
563
864
364
601
909
666
770
Geometric
(MC)
MC
156
156
153
493
544
800
309
561
937
583
689
mean
9.9
12.3
22.1
17.9
33.5
44.1
42.3
55.0
48.6
59.0
64.1
std.
5.3
4.8
4.3
6.9
5.8
5.0
4.5
4.9
5.9
4.9
5.0
    Figure 9 plots the theoretical arithmetic mean (analytical) and standard deviation versus those obtained by MC
simulations. The comparison shows that the lognormal approximation of G is valid over different soil textures.
                                                     23

-------
                         o
300


200 -


100 -


  0
                                      Mean
1000

 750

 500

 250 -I

   0
                                                                  StDev
                                 0     100    200    300

                                        analytic

Figure 9. MC versus theoretical mean and std. of G.
                                0   250  500  750  1000

                                       analytic
    A rainfall event was randomly selected. It occurred on 6/13/1983 with a total rainfall depth of 48 mm (Figure 10).
MC simulations are performed with this event for each parameter by running KINEROS-2  (Kalin and Hantush, 2003).
Peak flow (qp), cumulative flow (qt), time to peak flow (tpf), peak sediment discharge (qsp), total sediment yield (qst) and
time to peak sediment discharge (tps) values were recorded. Figures 11 and 12 show results from the MC simulations.
Since our focus is on sediment, only results related to sediment are shown. The vertical axis in each figure shows the
exceedance probabilities (1-CDF). Results for less sensitive parameters are not shown. A sudden drop from 1 to 0 in the
exceedance probability implies no variation of the  model output with respect to the particular parameter uncertainty,
whereas the more  gradual the transition  from 1 to 0,  the more sensitive the model output to  the parameter. Only
parameters shown in Figure 12 are directly affecting sediment transport. In other words, parameters shown in Figure 11
determine the shape  of the  hydrograph and since sediment discharge is  a  function of flow, they indirectly affect
sedimentograph. MC simulations were performed for an additional, smaller event (8/26/81) with a total rainfall depth of
17 mm for cf and cg (Figure  10). The secondary axes in Figure 12 correspond to this event. From Figure 11 it is clear
that the order of sensitivity is Ks, np,  G, A, (with  \\ib fixed  at its geometric mean),  S; and  ^ when peak sediment
discharge, qsp, is concerned. When total sediment  yield,  qst, is concerned  Ks is by far the most sensitive parameter
followed by G, Si, np, and X.  Time to peak sediment discharge, tps, is most sensitive to n^ and np. Ks and G are the next
most  sensitive parameters. Although A, affects model  output only through the  G  parameter, allowing \\ib  to vary
randomly, but independently, with A, explains the more gradual transition from 1 to zero of the probability exceedance
curve for  G than that for X,  indicating a greater uncertainty of the model output with respect to  the former. Order of
sensitivities may differ depending on the size and the nature of the rainfall event and quantity  of interest. For instance,
interception depth may play a significant role during small events. However, the general picture is  the same. The model
sensitivity to cf and cg are again event dependent as shown in Figure 12.  It is more sensitive to cg than cf during large
events. This mode of sensitivity  is reversed for smaller  events, where  rain splash erosion dominates model output
uncertainty (Kalin  and Hantush, 2003). The time to  peak sediment discharge, tps, is insensitive  to cf and cg. During
calibration, since flow parameters have to  be calibrated  first, Manning's roughness should be estimated initially to
match hydrograph timings. Next, Ks, G and St should be calibrated to adjust the volume of hydrographs. The parameter
St depends on the antecedent moisture condition and should be adjusted for each event.
^.u
^ 1*
E 15
o
5 10
'w
C c
0 O
"c
n



n
i Ir

JU






ia^^
                                                           20

                                                           15
                                                        t1°
                                                        |   5
                                                        "c
                                                            0
                        0    50   100  150 200  250
                             time (min)
Figure 10. Rainfall events at 6/13/83 (left) and 8/26/81 (right).
                                      50  100 150  200  250
                                        time (min)
                                                     24

-------
                               0.0
                                                               100

                                                         peak Qs (kg/s)
                                                                             150
                                                                                           200
                            Q.

                            
-------
                                               0.0002
                                                              0.0004
                                                                             0.0006
                                            200         400         600

                                             peak sediment discharge (kg/s)
                                                                                 0.0
                                                                              800
                                 0.000   0.001    0.002   0.003    0.004   0.005    0.006
                                1.0
                                                     6        9

                                                  sediment yield (tons)
Figure 12. Probability of exceedance of peak sediment discharge (kg/s) and total sediment yield (tons) for cf and cg parameters.
         Secondary axes are for cr2 and cg-2 (second event).


    The antecedent moisture condition has a significant effect on the sensitivity results. For instance, Figure 13 shows
the effect of initial saturation (SO on the sensitivity of peak sediment discharge and sediment yield to Ks. It is clear that
both the peak sediment discharge and sediment yield become more sensitive to Ks as the antecedent moisture condition
becomes dryer. A small perturbation in Ks results in significant differences as indicated by the large coefficient of
variations (COV) of peak sediment discharge and sediment yield. COV is a measure of deviation from the mean and is
computed by dividing standard deviation to the mean. This signifies that, under dry  conditions, model is sensitive to
more parameters and calibration is more difficult.
5.4     Model Calibration, Validation

    Three events for model calibration and 4 events for model validation were selected. Calibrations were performed
manually by comparing computed and observed hydrographs and sedimentographs (Kalin and Hantush, 2003). Average
values were used for G (20,35 cm), A, (0.6,0.6), O (0.50,0.47) and d50 (7 |am). First values in parenthesis are for silt loam
(SL) and second values are for silty clay loam (SCL). Table 7 shows calibrated parameters. The first three events are for
calibration and the rest is for validation purposes. At the end of each row the Nash-Sutcliffe statistics are given for both
flow and sediment. The sensitivity results indicate that peak  sediment discharge and sediment yield are very sensitive to
plane roughness (rip), but almost insensitive to channel roughness (ric). Time to peak sediment discharge is equally
sensitive to n- and np. Therefore, we calibrated for rip and used the same value for  ric This simplifies calibration as well.
Considering the agricultural nature of W-2, np and n^ are allowed to vary by time of the year due to growing crops.  It is
                                                      26

-------
assumed lowest at the beginning and largest at the end of the growing season. Si was allowed to vary from event to
event.  Si  values were  calibrated by  taking precipitation  fallen during the previous  five days into  account.  Since
KINEROS-2 does not model evapotranspiration losses, these losses were incorporated into  the interception depth I,
which was also allowed to vary by event and seasonally. The soil erosion parameters cg and cf are known to vary from
event to event due to sediment availability (Ziegler et al, 2001) and seasonally  due to tillage practices, freeze-thaw
processes and change in vegetation (Smith et al., 1999). Therefore, they  were allowed to decay exponentially from
highest values at beginning of the growing season to lowest at the  end of the growing season.  They were highest in
5/30/1982 and lowest in 8/26/1981. Negligible differences in Ks values were observed during calibration.
  -  0-
  '1  e
  <"  !s
  0  .!2
  Q.  -D
12


10


 8


 6


 4


 2
        50
        40
   S J1 30
        20
        10
                          Si=0.20
                    COV =4.25, COV,=4.11
         0.0
                    0.5        1.0
                          Ks
                                        1.5
                             Si=0.90
                       COVp=0.45, COV,=0.59
300


250

    O)
200 "^
    2
    .0
150 ^
    £
100 |
    0
    01
50


0
                                                2000
                                                1600




! .
c
0

M
a)
Q.







W)
^)
^
0"
E
to
T3



£-*J

20


15

10


5

n

Si=0.55
I COVp=2.94, COV,=2.91
\
|
I

I
1
\
\
\
JVja
/ UU

600
500

400

300

200

100
n



0)

•o
.0)

£
^
0
to



                                         1200
                                                400
                                               0
                                               '>,
                                               'c
                                               0
                                               E
          0.0
                    0.5       1.0
                          Ks
                                        1.5




- »
£ s1
E .
T3 o)
0) (0
^ •=
g »
Q. T3




0.0 0.5 1.0 1.5
Ks
60 -3nnn
50
40


30


20

10
o
«
\^^^^--^
*» ^

" -• - ^


_
Si=1.00
COVp=0.10, COV,=0.22

2500
2000


1500


1000

500
n
0.0 0.5 1.0 1.5
Ks




2
—^
—
•>,
^
0
E
0




Figure 13. Effect of antecedent moisture condition on Ks sensitivity. Si is initial saturation, COVp and COVt are the coefficient of
          variations of peak sediment discharge and sediment yield, respectively.
Table 7. Parameter set following calibration.

5/30/1982
6/13/1983
8/26/1981
6/12/1980
7/8/1981
8/1/1981
8/29/1975
n
0.04
0.055
0.08
0.055
0.08
0.02
0.09
KsSL
(mm/hr)
6
6.5
7
6.5
16
13
9
KSSCL
(mm/hr)
1.5
1.8
2.0
1.8
5.0
3.0
2.5
Inter (mm)
0.0
2.0
1.0
2.0
3.5
4.0
2.5
SisCL
0.86
0.27
0.60
0.27
0.20
0.20
0.20
SiSL
0.90
0.44
0.84
0.44
0.24
0.24
0.34
C9
0.250
0.150
0.050
0.150
0.080
0.015
0.010
Cf
200
160
100
160
130
100
90
Nashflow

0.92
0.99
0.87
0.96
Nashsed

0.83
0.91
0.84
0.93
                                                       27

-------
    Two different strategies can be followed for model validation purposes. The first technique is based on employing
the parameters, estimated with calibration, at the validation stage and comparing the performances of predicted and
observed hydrographs/sedimentographs. In the second method, parameters are recalibrated so as to have good matches
between observed and predicted model outputs. Then, recalibrated parameters are  compared to the expected values
obtained through calibration. In this study we utilized the latter method. Parameters estimated using the validation
events are, in general,  in good agreement with calibrated parameters  (Table 7).  There are acceptable amount  of
variations in Ks values considering the nature of Ks which has very high coefficient of variations in most soils (eg. 2.73
for SL). The only unexpected result is with the n value of the event 8/1/1981. A value of 0.02 is estimated in contrast to
an expected value of 0.08 to  accommodate the early response observed in measured data. Based on rainfall records, the
soil is expected to be very dry prior to this event. Therefore S; is kept minimum, and since it is the month of August, I
can not be zero. Possible explanations might be i) potential measurement errors, or ii) even at this small scale spatial
variation of rainfall may play an important role. The computed and observed sedimentographs are shown in Figure 14.
               5/30/1982



               o  observed

                  computed
                                                            400
                                                       -22
                                                        O)
                                                        D)
                                                        ro
                                                        -C
                                                        o
                                                            300
                               £   200
                                                       OT
                                                            100
                                                                                                6/13/1983
                                                                                            o   observed
                                                                                           	computed
                                                                                        _oo  n  _
50        100
      time (rrin)
                                        150
                          200
50
                                   90
                                    time (min)
                                       130
170
              8/26/1981



             o  observed
                computed
                                                            400
                                                         o)  300
                                                         ro
                                                         '?,  200
                                                         c
                                                         CD

                                                         8  100
                50
100      150
 time (min)
200
250
                                                                                            06/12/1980
                                                                   o   observed
                                                                  	computed
                                     30             60
                                       time (rrin)
                                                                                                          90
Figure 14. Computed and observed sedimentographs for selected events.
                                                     28

-------
                                                           120
                            CD
                            O)

                            |   80
                            o
                            T3
                                                         CD
                                                            40
                                                                                            08/01/1981
                                                                                            o  observed
                                                                                               computed
       150
                     300
                                                                             30            60             90
                                                                                time (rrin)
0
£10
-C
o
^
•E
0
I   5
0
en
                                                              08/29/1975
                                                               o  observed
                                                               	computed
                                  0          50        100        150        200
                                                     time (min)

Figure 14 (continued). Computed and observed sedimentographs for selected events.
5.5     Discussion:

    The calibration and validation exercise performed over the W-2 watershed with KINEROS-2 show that channel
roughness, ric, plane  roughness, np, and soil erodibilities cg  and cf, show seasonal variations.  This is due  to  the
agricultural nature of W-2. During calibration it is recommended that np and n^ be calibrated first to adjust hydrograph
timings. Average values suggested in the literature can be used for ric, as the sensitivity results indicate that KINEROS-2
is more sensitive to np than n^ when peak sediment discharge and sediment yield are concerned. The time to peak
sediment  discharge is almost equally sensitive to both parameters. The  saturated hydraulic conductivity, Ks and
effective capillary drive parameter, G can be  calibrated next by focusing more on Ks to match the flow volumes. The
soil erosion parameters cg and cf can be calibrated next, to adjust the computed sediment yield to the observed.

    Beven (1989) states that calibration to match  a single event  is not difficult where a loss function and a routing
function are all that is needed. However, the calibrated data set has to be verified over additional events. The difficulty
lies under the estimation of initial  soil moisture content which depends primarily on prior  rainfall events. Like all
physically-based models, KINEROS-2 requires the initial estimation of soil moisture which  is usually not available.
Figure 13  shows how  important the  selection of the initial soil moisture content is in the KINEROS-2 model. The best
way to overcome the effect of the initial soil moisture is performing continuous simulations where none of the critical
processes  are ignored in the water balance and soil moisture is redistributed between the storms, i.e. during rainfall
hiatus. Although KINEROS-2 considers soil moisture redistribution, it ignores evapotranspiration. Therefore it is not
suitable for continuous simulations  since a true water balance is not possible. In the next section the GSSHA model
having both event and continuous simulation capabilities is investigated. The flow and sediment results are compared to
                                                      29

-------
KINEROS-2 by running the event module of GSSHA with the same events employed in KINEROS-2 simulations.
Later, long-term, continuous-time simulations are performed over the same watershed with GSSHA and results are
discussed.
                                                 30

-------
                       6    Comparison of KINEROS-2 with GSSHA


    In this chapter KINEROS-2 and GSSHA (Downer and Ogden, 2002) models are compared quantitatively based on
their performances on modeling flow and sediment movement. Each model has a different watershed conceptualization
(Figures 15 and 16). GSSHA divides the watershed into cells, and flow and sediments are routed through these cells in a
cascading fashion. Conversely, KINEROS-2 divides the watershed into  sub-watersheds or transects and channel
segments  having  uniform properties.  GSSHA  may require  much  longer simulation times  depending on what  is
simulated. KINEROS-2, on the other hand, entails relatively less data and effort. Simulations were performed with each
model over the W-2 watershed. Both models were calibrated using the same events and the differences in estimated
parameters were discussed. Both models have resulted in different calibration parameters. The differences in model
behaviors are discussed. Model descriptions and features are given in the previous sections. For full model descriptions
users can refer to the references given.
6.1     Model Features

    Features of KINEROS-2 model, with emphasis on the sediment component, was described in the previous chapter.
Here, the properties of the GSSHA model are presented with the focus on the sediment formulation.

    GSSHA is a reformulation and enhancement  of the hydrologic model  CASC2-D  (Ogden and Julien,  2002).
However, the  sediment components are exactly the same. GSSHA  can perform single  event  and  continuous time
simulations. Watershed is divided into cells and water and sediment is routed from one cell to another in two principle
dimensions. It uses one and two-dimensional diffusive wave flow routing at channels and overland planes, respectively.
Although only Hortonian flows were  modeled by  employing Green-Ampt (G-A) infiltration model in the initial
versions, GSSHA considers other runoff generating mechanisms such as lateral saturated groundwater flow, exfiltration,
stream/groundwater  interaction etc.  GSSHA  offers three options for  computation of infiltration:  G-A, G-A with
redistribution (Ogden and Saghafian 1997) and the full Richards' equation.

    Modified Kilinc and Richardson equation (Julien 1995) is used to compute sediment transport capacity at plane
cells. The potential sediment transport rate is computed in x and y directions as


                                                                                                     (15)
                                                           0.15

where qs is sediment unit discharge (ton/m/s), q is unit flow discharge (m2/s), Sf is friction slope, and K (soil credibility
factor), C  (cropping factor) and P (conservation factor) are the USLE (Universal Soil Lois Equation) soil parameters.
The index i represents the two principal directions, x and y, therefore  sediment transport capacity is computed in both
directions.

    Each cell can either be eroded or aggraded depending on the sediment in suspension and potential sediment rates.
This determination is made for three particle sizes: silt, clay and sand. If sediments in suspension are unable to satisfy
the potential transport rate, erosion occurs. If the potential transport rate is unable to transport the sediment already in
suspension, deposition occurs. A trap efficiency measure is used to determine how much material is deposited (Johnson
etal.,2000).
                                                    31

-------
'.; =l-e
                                                                                                          (16)
where TEj is the trap efficiency for the f1 particle size ranging from 0 to 1, Ax is the grid cell size (m), Wj is the fall
velocity of the j* particle size (m/s), u is the overland flow velocity (m/s) and y is the overland flow depth (m). The use
of trapping efficiency allows deposition of larger particles before the smaller ones.
Figure 15. Watershed conceptualization in GSSHA.
(
1

9 -»>






12 *

r

* 8

I 11
t » ;
2 f « 7

n 10
o
                                            L
                      w = A/L

       L = average length of overland flow ; A = watershed area

Figure 16. Watershed conceptualization in KINEROS-2.
    Yangs' unit stream power method (1973) is used for routing sand size particles in stream channels. This routing
formulation is limited to trapezoidal channels. Silt and clay particles are assumed to be always in suspension and

-------
therefore transported as wash load. More details on theory and equations used can be found in Downer and Ogden
(2002).

    Many applications of the GSSHA model and its predecessor CASC2D can be found in peer reviewed literature (eg.
Johnson et al. 2000; Molnar and Julien, 2000; Senarath et al., 2000; Ogden and Heilig, 2001; Downer and Ogden,
2003a; Downer and Ogden, 2003b).]

    The watershed conceptualization employed in GSSHA seems more realistic than the realization used in KINEROS-
2. The use of diffusive wave approximation to the full Saint Venant equations in GSSHA is an improvement over the
kinematic wave approach utilized in KINEROS-2. KINEROS-2 is limited to Hortonian flow and is not suitable for
long-term simulations because it lacks evapotranspiration (ET) component which is important for the mass balance of
the water cycle. On the other hand,  GSSHA can handle various runoff generating mechanisms. In general, the flow
component of GSSHA can be expected to perform better than the flow component of KINEROS-2 since it involves less
simplification. Contrary to  flow, the sediment formulation of GSSHA is not  as strong.  KINEROS-2 has a better
sediment transport formulation. GSSHA's sediment component is based on semi-empirical relationships, whereas
KINEROS-2 employs a more physically based-approach.
6.2     Approach

    KINEROS-2 was already calibrated for W-2 watershed in the previous section using 3 rainfall events. The fixed
parameters are net capillary  drive, G(20,35 cm), pore size  distribution index, 1(0.6,0.6), porosity, c|>(0.50,0.47), and
median particle size diameter, d50(7 |am). The two values given in parentheses represent different soil types, silt loam
(SL) and silty clay loam (SCL), respectively. Table 8 lists the parameter sets used after calibration of KINEROS-2. In
the table, n is Manning's roughness, Ks is saturated hydraulic conductivity, I is interception depth, Si is initial saturation,
cg is soil cohesion coefficient and cf is rain splash coefficient. The sensitivity results in chapter 5 indicated that peak
sediment discharge and  sediment yield are very sensitive to plane roughness (np), but almost insensitive to channel
roughness (ric). Time to peak sediment discharge is equally sensitive  to nc and np. Therefore, we calibrated for np and
used the  same value for ric. Since corn has been grown on W-2, the parameters ric, np, cg and cf were allowed to vary
with season where cg  and cf were assumed to decay exponentially  with the growing season.  This  assumption was
justified over 4 independent verification events (see previous  section).


Table 8. Parameter sets used in KINEROS-2.
event
6/1 3/1 983
5/30/1 982
8/26/1981
n
0.055
0.04
0.08
Ks
(mm/hr)
(6.5,1.8)
(6.0,1.5)
(7.0,2.0)
I
(mm)
2
0
1
Si
(0.27,0.44)
(0.86,0.90)
(0.60,0.84)
Cg
0.15
0.25
0.05
c,
160
200
100
6.2.1   Flow Simulations

    GSSHA was run with the above events. KINEROS-2 values were directly substituted for parameters common to
both models i.e. X, c|>, n, I and Si. Other parameters were adjusted accordingly. The infiltration scheme in GSSHA is the
Green-Ampt (G-A) model, whereas KINEROS-2 uses Smith-Parlange infiltration model, which is a generalization of
the former. G-A capillary head (*Ff) needs to be provided in GSSHA. We approximated *Ff as equal to G in KINEROS-
2. We used the Ks values given in Table 8 for the G-A hydraulic conductivity (KQ.A). Figure 17 shows the comparison
of the  simulation results for flow with two models. It is clear that both models perform differently when similar
parameter sets are used as  inputs. The most striking observation is that, in all cases GSSHA generates later responses
and lower peak flows than KINEROS-2. For instance, the difference in time to peaks for the event 8/26/81 is around 25
minutes which is very significant considering the fact that the base time is around 150 minutes. Similarly, the peak flow
generated by KINEROS-2 is about 45 % larger than the peak flow generated by  GSHHA. One  possible rationale to this
might be the different watershed conceptualizations involved in each model. Flow routing in GSSHA is only in  x-y
directions (Figure 15).  In  other words, flow from a cell is allowed  only in the four principal directions. Diagonal
neighboring cells can not be receivers which well might be the reality. This  results in overestimation of the travel
                                                    33

-------
lengths of water particles which might be up to 41 %. On the other hand, the travel paths used to compute the average
travel lengths of each element in KINEROS-2 were determined based  on the D-8 methodology using the TOPAZ
algorithm (Garbrecht and Martz 1999) which allows flow in 8 directions. Considering the fact that flow in the study
watershed is mostly diagonal, the overestimation of travel lengths by GSSHA resulted in longer travel time leading to
more resistance to flow, and consequently lower and retarded peaks.
                            3 -
                         CO
                         ,§

                         o
2 -
                            1  -
                             60
                                                     6/13/83
                     kineros-2
                     -gssha
                     observed
                  90               120
                       time (min)
                                150
                            0.4 -
                               30
                  80            130
                     time (min)
                          180
                            0.20 -
                            0.15 -
                            0.10 -
                         o
                            0.05 -
                            0.00
                                30
80              130
    time (min)
                                                    180
Figure 17. Comparison of hydrographs generated with GSSHA (straight lines) and KINEROS-2 (dashed lines) based on KINEROS-2
            calibrated parameters. Observed data is shown as hollow circles (Kalin and Hantush, 2003).
                                                    34

-------
    Total flows at the watershed outlet for observed data and KINEROS-2 and GSSHA simulations are shown in Table
9. The differences between the flow volumes of KINEROS-2 and GSSHA do not seem to be significant. With this set of
parameters KINEROS-2 seems to simulate events having multi-modal shapes, such as the one in 5/30/82, better than
GSSHA. In fact GSSHA completely misses the first and second humps in 5/30/82 (at 48 and 61 minutes, respectively)
as opposed to KINEROS-2. KINEROS-2, to some extent, performs better than GSSHA in simulating the small hump
seen on the observed data of 8/26/81.


Table 9. Total flows in m3 at the watershed outlet from observed data, and KINEROS-2 and GSSHA simulations with KINEROS-2
      calibrated parameters.
                                         6/13/83
               5/30/82
              8/26/81
                     OBSERVED
                     KINEROS-2
                     GSSHA
3801
3435
3509
1042
679
602
317
335
318
    It is important to keep in mind that all these observations are based on simulations with the parameters calibrated
for KINEROS-2. Therefore, we recalibrated the GSSHA parameters for the same events. This time each event was
calibrated individually and parameters were compared to KINEROS-2 calibrated parameters. We accept that we did not
follow the traditional model calibration/verification methodology. However, we need to mention that the aim of this
study is basically a comparison of the two models rather than a model calibration effort. Keeping this in mind, we kept
I, S; and the overland plane roughness (np) same and recalibrated channel roughness (nj and KG.A. Figure 18 shows the
hydrographs after calibration. For the event 6/13/83 both model performs equally. For 5/30/82 GSSHA is still
underestimating the first and second humps (at 48 min  and 61 min, respectively). Although KINEROS-2 could not
simulate the first hump (the smallest hump in the figure) GSSHA was able to generate all the humps. Finally, when we
look at the last event we see that GSSHA almost perfectly reproduces the observed hydrograph shape while KINEROS-
2 does a poorer job of simulating the first peak.

    The  recalibrated parameters for GSSHA are  summarized in Table  10.  In the table C is the USLE cropping
management factor which will be discussed later. The value of ric had to be decreased dramatically for each event which
is clearly expected from Figure 17 as GSSHA generated later responses in each case. One remarkable observation is that
ric values are very close to each other which confirms the comments of Larry Kramer (personal communication) who
has extensive  experience on Treynor watersheds. He stated that channels are covered with bromegrass and they are
cultivated such a way that channel roughness can be assumed  invariable year around. KG_A values are very close to
KINEROS-2 Ks values. Rawls and Brakensiek (1983) recommends KG-A=Ks/2 based on Bouwer's (1966) findings.
                                            90            120
                                                time (min)
                              150
Figure 18. Comparison of hydrographs generated with GSSHA (straight lines) and KINEROS-2 (dashed lines). GSSHA
          is recalibrated. Observed data is shown as hollow circles (Kalin and Hantush, 2003).
                                                   35

-------
                              0.4 -
                              0.3 -
                           CO
                           ,§

                           o
                                  30
80            130
  time (min)
           180
                             0.20 -
                             0.00
                                 30
 80
 time (min)
130
180
Figure 18 (continued). Comparison of hydrographs generated with GSSHA (straight lines) and KINEROS-2 (dashed lines). GSSHA
                  is recalibrated. Observed data is shown as hollow circles (Kalin and Hantush, 2003).
Table 10. Calibrated parameters with GSSHA.
                                      event
          KG-A
        (mm/hr)      C
                                    6/13/1983   0.025    (7.7,2.0)    0.042
                                    5/30/1982   0.020    (6.0,1.5)    0.150
                                    8/26/1981   0.025    (6.5,1.8)    0.050
6.2.2   Erosion Simulations

    GSSHA requires silt and sand percentages for sediment computations. The default values used in the GSSHA
model for D50 are 0.25 mm for sand, 0.016 mm for silt and 0.003 mm for clay. Based on these values compositions of
each soil class were determined as sand % (25,10) and silt % (61,56) so that the overall average D50 is 7 mm, which is
the value used in KINEROS-2. Again, the values in the parentheses are for silty loam (SL) and silty clay loam (SCL),
respectively. The sediment routine  in GSSHA  is empirical and based on the USLE concept  that requires  three
parameters:  K (soil credibility factor), C  (cropping management factor) and P (conservation practice factor). It is  not
practical to  infer estimates of these  parameters from the KINEROS-2 soil parameters;  i.e., cg and cf.  Therefore, by
keeping KP product constant C was calibrated for each event, since it is only the product of K, C, and P that matters.
The values of K and P are (0.37,0.48) and (0.01,0.01), correspondingly. The estimated C values are listed in Table  10.
                                                      36

-------
The pattern observed in KINEROS-2 is that credibility decreases with the growing season, but is not observed between

the C values here. The C values obtained for the event 8/26/1981 is unexpectedly high, even higher than the value of

6/13/83. Figure 19compares the sedimentographs obtained by KINEROS-2 and GSSHA. The general observation is that

GSSHA generates narrower sedimentographs than KINEROS-2 generates. This may be attributed to the fact that unlike

the physically based sediment  component  in KINEROS-2,  GSSHA utilizes  empirical  relationships for  sediment

transport. Further, this cannot be attributed to flow, since such a behavior is not reflected in Figure 19.
                            400
                          (/)

                          ^3)
                             300
                          1200
                             100
                                                                6/13/1983
                                         kineros-2

                                         -gssha

                                         obs
                                                           ?.0j>. 0__0 o
                                60
                             25
                          2 20
E>
ro
.c
o

T3




I

T3
                             15 -
                             10 -
                              5 -
                                40
                     90             120
                         time (mm)
 150
                                                                5/30/1982
                     80  ,.   ,   . .   120
                         time (mm)
160
                              5

                            E?
                            ro
                           -C o
                            o 3
                            
-------
    It is interesting to note that the erosion parameters, cf and cg, found after calibration for KINEROS-2 are well above
the recommended values given in Woolhiser et al. (1990) and the calibrated C parameters for GSSHA are well below
the literature values. This implies that when  literature values are  used GSSHA overestimates erosion compared to
KINEROS-2. Slope  is an important factor in both models' erosion formulation. The smaller the computational element,
which is the grid size for GSSHA and the average  length of overland flow  planes in KINEROS-2, the greater the
erosion. This occurs because, as the element size increases the tendency of smoothing the topography increases, and this
results in loss of areas with steep slopes meaning reduction in erosion. KINEROS-2 uses far less elements than GSSHA,
thus leading to loss of local slope information in the former. This probably elucidates the difference in estimates of soil
erosion. A detailed discussion on this topic can be found in Rosalia (2002).
6.3     Long-Term Simulations with GSSHA

    Here we investigate the long term simulation capabilities of the GSSHA model over the W-2 watershed. In order to
perform long-term simulations in GSSHA, in addition to rainfall data, hydrometeorological data are required for the
entire period of the simulation. The required data are hourly values of barometric pressure, relative humidity, total sky
cover, wind speed, dry bulb temperature, direct radiation and global radiation. These data can be  supplied in three
different formats to GSSHA: WES, SAMSON and NOAA/NCDC surface airways format. WES is the simplest and the
preferred format, while the last one is the  least recommended. SAMSON data is used in this  study which can be
purchased from National Climatic Data Center (NCDC) in a CD-ROM. The closest station to the W-2 watershed was in
Omaha, NE.

    GSSHA offers two options for infiltration  calculations during long-term simulations:  Richards' equation (RE)
(Richards,  1931) and  Green-Ampt with redistribution  (GAR)  (Ogden  and  Saghafian, 1997) which is basically
simplification of RE.  In  Hortonian basins GAR method produces  comparable results to RE (Downer and  Ogden,
2003a).  However, when Hortonian flow is not the dominant stream flow generating mechanism, GAR may produce
erroneous results, and RE should be used (Downer and Ogden, 2003a). Since W-2 is a Hortonian watershed we used
GAR to simulate a period from 5/17/1984 to 6/17/1984.

    The precipitation data used in this long-term simulation is shown in Figure 20.  The last rainfall event before
5/17/84 is on 5/6/84. Therefore, we assumed dry initial condition with initial moisture content of 0.1 for both soil types
(SL and SCL).  In fact, we considered the first 7 days of the simulation as warm up period and  thus disregarded the
results in that period to reduce the effect of initial moisture content.
u
3
O
^ 5
£
o
J 10
c
1 15
f 20
'ra
X
f
1

-


-

-
-
. . 1 . . 1 ,
X X ^










1 1 1 1 1 1 1 1
\%h ^ ,A\^
(/Z-3 fe\
1










X
jl ' F










rt^ pja^ M%^
»* iT* e>^
I










^
' r ^









:

Figure 20. Rainfall histogram used in the long-term simulations of GSSHA.


    The parameters used in the simulation and their values are shown in Table 11. In the table KG_A corresponds to G-A
hydraulic conductivity, ©r is residual water content, ©w is wilting point water content, ©; is initial water content, and *Ff
is wetting front capillary pressure head. Other parameters  are as defined before. The values listed in the table are
selected in a way that they are close to the values listed in Tables 9 and 10 for the event 5/30/82, since 5/30 seasonally
falls in the middle of the simulation period (5/17-6/17). The only significant difference is in the KG_A values. We
recalibrated KG_A values for the first two events occurring on 5/25/84 and 6/2/84 (5/19 and 5/25 are discarded as they
                                                    38

-------
are in the warm up period). These calibrated KG_A values are smaller than the values given in Table 10. In single-event
calibrations, the  initial moisture content has to be estimated more realistically.  Any  overestimation of initial water
content results in overestimation  of hydraulic conductivity  and vice versa.  In continuous long-term simulations,
however, effect of initial water content is more considerable at earlier stages, and decays with time. Therefore, obtaining
different KG_A values within tolerable ranges from event and continuous simulations is reasonable.
Table 11. Parameter values used in GSSHA long-term simulations.
K ff 1 sand
nD nc (mm/hr) O 0r 0W &t (cm) X (mm) K C P %
SL 0.04 0.02 3.5 0.486 0.015 0.133 0.10 20 0.23 1.0 0.48 0.15 0.01 25
SCL 1.0 0.432 0.040 0.208 0.10 35 0.18 0.37 10
silt
%
61
56
    Nine  different  events  are  recorded between the periods 5/25/85 and 6/17/82 in W-2.  Figure 21 shows the
hydrographs of the  first seven events. Last two events occurring on 6/16/84 and 6/17/84 are not shown in the graph
since GSSHA estimated no flow during those two events, although significant flows are observed in both events (peak
discharge is 0.39 mVs on 6/16/84, and 0.42 m3/s on 6/17/84). Events on 6/4/84 and 6/5/84 are shown on the same graph
(Figure 21). First two events are the calibration events where only G-A hydraulic conductivity (KG.A) was calibrated.
Rest are validation  events.  Estimated and observed flow hydrographs from calibration and validation events conform
well as can be seen  in Figure 21. Interestingly, validation events produce even better results than calibration events. As
mentioned earlier, GSSHA  did not generate runoff for the events happening on 6/16/04 and 6/17/04. Simulations were
performed with the  RE option, by adjusting the parameters accordingly (results not shown) to explore if this might be
linked  to the infiltration routine used. GSSHA was  still unable to generate any flow during the last two events. The
observed flows in both events are smaller  than the observed flows  of the other events. Thus, either  GSSHA has
difficulty in generating small events, or there is an anomaly in the rainfall data during that  time interval, such as
inappropriate representation of the rainfall pattern due to spatial variation.

    Figure 22 shows the observed and GSSHA generated sedimentographs. Sediment data was not available for 6/2/84.
The overall performance is poor. However, in 6/12/84 and 6/14 the falling limbs are well represented.
     1
1.0 -

0.8 -

0.6 -



0.2 -

0.0
                                         '0000
                                                            2.0
                                                            1.5 -
                                                          CO


                                                          o
1.0 -
                                                            0.5 -
                                                            0.0
                                                                                     o ooooo   Q
                                              ,<5'
                              ,T>~
Figure 21. Observed (hollow circles) and simulated (straight line) hydrographs from the long-term simulations of GSSHA.
                                                     39

-------
2.5
_ 2.0 -
I1-5'
:g 1.0 -
0.5 -
n n -



.
rl ^


!



*
i
L^
                                                               1.4
                                                               1.2
                                                               1.0
                                                               0.8
                                                               0.6
                                                               0.4
                                                               0.2
                                                               0.0
   00
   ,§
    I
4.0
3.5

2.5
2.0
1.5
1.0
0.5
0.0
                                                           CO
                                                                                           **
Figure 21 (continued). Observed (hollow circles) and simulated (straight line) hydrographs from the long-term simulations of
                   GSSHA.
   120
B) 10° "
D)   8° "
TO
o   eo ^
(/>
I   «H
                                                g 200 -
                                                
-------
                                                       2500
                                                    'in
                                                    9> 2000 -
                                                       1500 -
                                                    =6  1000 -

                                                    0)
                                                    .1   500 -
                a

                   
-------
to get more realistic results. This suggests that the sediment routine in KINEROS-2 is more robust than the routine used
in GSSHA. In fact, there is a contract between US Army Corps of Engineers' Engineering Research and Development
Center and University of Connecticut to completely reformulate the sediment routine of GSSHA (Downer and Ogden,
personal communications). It would be interesting to redo this whole exercise once that project is completed.

    Long-term, continuous simulations performed over W-2 with GSSHA using the Green-Ampt with redistribution
(GAR) infiltration option produced hydrographs comparable to observed data except for two events which are at the end
of the simulation period. GSSHA was unable to generate runoff during those two events, though observed data indicate
considerable flow. The performance of sediment results was poor.  In some events, however, the falling limbs of the
sedimentographs were well represented.
                                                   42

-------
                                 7   Summary and Conclusions


    As required by the 1972 Clean Water Act, states, territories,  and authorized tribes are required to develop TMDLs
for sediments which  is the leading stressor of nation's  streams. Water quality managers  and  stakeholders  are
increasingly relying on hydrologic and water quality models  as  cost-effective  tools for preliminary and detailed
watershed planning, including TMDL development and BMP performance evaluations. BMPs  are important parts of
risk management studies since they are used to reduce pollutant loading and achieve TMDL targets. A large amount of
models are available for users to select from. The process of selecting the right model given the needs is not an easy
one,  entailing familiarity with the available models.  Several studies exist  in the literature  assessing  models and
summarizing their features and capabilities, all based on different perspectives. This report presented an evaluation of
the most widely used suspended solids and sediment transport models and related  nutrients water quality models. The
report addressed the capability of the models to simulate for BMPs, both structural and nonstructural. A probabilistic,
risk-based mathematical optimization framework was  presented and was proposed as a strategy for solving the TMDL-
BMP problem involving  multiple  stressors. Although, the framework  was presented  in  general mathematical
formulation it may guide  future  model applications  to the management of sediment  and  nutrients  in  complex
watersheds. Future modeling efforts should be directed toward applying system analysis approaches to solve the BMP
problem in an optimal fashion.

    The models evaluated in this report had a proven track record of applications and documentation, and were cited in
numerous reports. However, some of the models that have a less visible track record and applications may be promising.
Models  were  selected  after an initial phase of screening, based  on their suspended solid or sediment modeling
capability,  strong model documentation and/or  support, and proven record  of  application with  sufficient history.
Relatively new and promising models were also added to the list for future considerations. The latter models have short
history and some are still in the beta versions, but have been cited in peer reviewed  publications. Models  were reviewed
under two basic categories: loading or watershed models, and receiving  water models. Features of each  model were
summarized in a tabular form. Detailed description of the model features was included in the Appendix.

    Among the loading models that have capabilities to simulate  sediment and nutrient load reductions by management
practices were AGNPS (ANNAGNPS for continuous time simulations) and SWAT. Both models are widely used in
agricultural watersheds. The latter has  its  own  GIS interface and currently integrated into USEPA's BASINs and
USDA's AGWA modeling systems. It is also linked to the water quality  model, QUAL2E. For urban areas, the most
comprehensive sediment loading model is the widely used SWMM model. An urban watershed-receiving waterbody
modeling system can  be formulated by linking SWMM to the USEPA's WASP.  The latter has a  eutrophication
component. For large watersheds comprised of both urban and rural areas HSPF is the  most suitable model to address
the sediment and nutrient TMDL problems. HSPF can be run under BASINS and WMS modeling systems. The DHFs
MIKE-SHE watershed model is probably the most physically based, comprehensive "modeling system", especially in
agricultural watersheds, with a history  of applications in peer  reviewed journals. It is equipped with several BMP
simulations capabilities including wetlands, nutrient and pesticide management. This modeling package,  however, is
proprietary. USEPA's BASINS is another complete modeling system and has been applied for TMDLs. It has loading
(SWAT and HSPF), and stream and river water quality (QUAL2E and HSPF) models. EPA  is also  working on
expanding BASINS to include 3-D hydrodynamic and water quality model EFDC.  It not only simulates for sediments,
but also simulates transport and fate of  many other pollutants. However,  it is less physically based than MIKE-SHE.
The  system provides the linkages between these models within  an Arc View environment. The WMS  is  a watershed
modeling system into which the GSSHA and HSPF models have been integrated. It is  an effective, user-friendly
package for simulating sediment yield from watersheds. If linked to QUAL2E and WASP, it has the potential to be a
formidable watershed analysis tool for suspended solids, sediments, and nutrients.
                                                    43

-------
    In conclusion SWAT and ANNAGNPS are suitable for sediment and nutrient BMP simulations analysis in
agricultural areas. SWIMM is preferable for development of sediment TMDLs and BMP strategies in urban areas, and
HSPF is the recommended model for large watersheds with mixed land use containing both rural and urban areas. To
our knowledge  MIKE-SHE and BASINS are the only comprehensive modeling systems  for TMDL allocation and
sediment and nutrients load reduction assessment of BMPs. If fully developed for water quality and eutrophication,
WMS can be a promising, user-friendly watershed modeling system capable of a complete sediment TMDL analysis.

    Unless extra, often time consuming, effort is made, current watershed and water quality models can not be used for
comprehensive sediment TMDL allocation and reductions. Future efforts should focus on state-of-the-science in terms
of processes improvement,  and on the state-of-the-art by  further developing efficient, user-friendly  modeling
frameworks. A suggested enhancement would be developing more model linkages. Widely used receiving water quality
models either have their  own hydrodynamic components, or are linked to other hydrodynamic models. However, there
appears to be a big gap between loading models and hydrodynamic models. Developing modular modeling frameworks
that provide selective linkages between loading models and hydrodynamic models, or complete modeling systems is
worthwhile. Most mechanistic  models  that are based  on sound physical  principles lack  comprehensive BMP
components due to the fact that the original objectives during model developments were  not geared toward TMDL
development and assessment of BMPs. Enhancement of such physically based models with additional BMP capabilities
would benefit TMDL developments and evaluation of diverse BMP options. Further,  most BMP models rely on
empirical relationships and are functional only at the local field scale. Future efforts should focus on developing
process-oriented, mechanistic models for both structural and nonstructural management practices, and should develop
techniques  to take processes at the  local management scale and scale them up to the watershed scale.  For instance,
REMM and VFSMOD can be linked to loading models to simulate sediment transport in riparian buffers and vegetative
filter strips, respectively.

    The second part of the report addressed numerical evaluation of two physically based runoff and sediment transport
models, KINEROS-2 and GSSHA.  The purpose of the second part was demonstration of a strategy for quantitative
model comparison. The  models were applied to an USDA experimental, agricultural watershed.  Both models  are
promising, distributed hydrologic loading models. KINEROS-2 is suitable for small agricultural watersheds (<100 km2)
and is one of the two models in the newly developed AGWA modeling system which is supported by both USEPA and
USDA. It is suitable for event-based simulations since it does not have a complete soil moisture accounting component.
The sediment component is physically based and has a track record of successful applications in literature (see model
summary).  The sensitivity analysis performed  over  KINEROS-2 with Monte Carlo showed that among the flow
parameters  the most sensitive parameters in descending order are Ks, np, G, X, Si and  ric when peak sediment discharge
is concerned. For total sediment yield, Ks is by far the most sensitive parameter followed by G, Si, np, and X. Time to
peak sediment discharge is most sensitive to nch and np. The soil erosion parameters cg and cf have mixed effects. For
large storms cg is the dominant parameter, whereas results  are more affected by cf in smaller events. Model is sensitive
to more parameters as the antecedent moisture condition get dryer. KINEROS-2 was calibrated for 3 events and the
calibrated parameters were verified for 4 events. The overall model performance was good.  Results  indicated that the
Manning's  roughness and soil erosion parameters show seasonal variations. In future applications, it is recommended
that Manning's  roughness should be estimated initially to match hydrograph timings. Next, Ks, G and Si should be
calibrated to adjust the volume of hydrographs. The parameter Si depends on the antecedent moisture condition and
should be adjusted for each event.

    Both models, KINEROS-2 and GSSHA were calibrated and verified. The results indicated that the flow component
of the latter over performed the former. Conversely KINEROS-2 was more robust in simulating erosion  and sediment
transport. GSSHA, however, has both event-based and continuous  simulation capabilities, whereas KINEROS-2 is
essentially event based. At this stage both models lack nutrient components, and their capability to simulate for BMPs is
limited. Future efforts concerned with watershed model evaluation may benefit from migrating from qualitative analysis
to quantitative evaluation using real watershed data. The  limits and merits of models can only be identified through
numerical evaluation on selected watersheds.
                                                    44

-------
                                           8    References
AWWA. 2001. Water Resource Planning Manual (M50). Edited by W.O. Maddaus, American Water Works
    Association, AWWA, Denver, CO.
Beven, K.J. 1989. Changing ideas in hydrology - the case of the physically-based models. Journal of Hydrology,
    105:157-172.
Borah, D.K. 2002. Watershed scale non-point source pollution models: mathematical bases. 2002 ASAE Annual
    International Meeting/CIGR World Congress, Chicago, IL. Paper no. 022091.
Bouwer, H. 1966. Rapid field measurement of air entry value and hydraulic conductivity of soil as significant
    parameters inflow system analysis. Water Resources Research, 2:729-738.
Browning, G.M., C.L. Parish, and J.A. Glass.  1947. A Method for Determining the Use and Limitation of Rotation and
    Conservation Practices in Control of Soil Erosion in Iowa. Soil Science Society of America Proceedings.  23:246-
    249.
Chapra, S.C. 1997. Surface-Water Quality Modeling. McGraw-Hilll, Inc.
Chow, V.T., D.R. Maidment, L.W. Mays. 1988. Applied Hydrology. McGraw-Hill, Inc.
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, Engineering Research and Development Center, Vicksburg,
    MS.
Downer, C.W., and F.L. Ogden. 2003a. Prediction of runoff and soil moistures at the watershed scale: effects of model
    complexity and parameter assignment. Water Resources Research, 39(3):1.1-1.13
Downer, C.W., and F.L. Ogden. 2003b. Appropriate vertical discretization of Richards' equation for two-dimensional
    watershed-scale modeling. Hydrological Processes, in press.
Engelund, F., and E. Hansen.  1967. A monograph on sediment transport in alluvial streams. Teknisk Vorlag,
    Copenhagen.
Garbrecht, I, and L.W. Martz. 1999. An automated digital landscape analysis tool for topographic evaluation, drainage
    identification, watershed segmentation, and sub-catchment parameterization. RepJ GRL 99-1, Grazinglands
    Research Laboratory, USD A, Agricultural Research Service, El Reno, OK.
Goodrich, D.C., T.J. 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, 30:1393-
    1405.
Hantush, M. M., and L. Kalin. 2003.  Modeling uncertainty of runoff and sediment yield using a distributed hydrologic
    model. Proceedings of the First Interagency Conference on Research in the Watersheds, Benson, AZ, 27-30
    October.
Hantush, M. M.  and M. A. Marino. 1989. Chance-constrained model for management of a stream-aquifer system. J
    Water Resour. Plan. Manag, ASCE,  115(3), 259-277.
Imbeau, M.E. 1892. La Durance: Regime, Crues et Inundations. Ann. Fonts Chaussees, Mem. Doc. Ser., 3(I):5-18 (in
    French).
Johnson, B.E., P.Y. Julien, D.K. Molnar,  and C.C.  Watson. 2000. The two dimensional upland erosion model CASC2D-
    SED. Journal of American Water Resources Association,  36:31-42.
Julien, P.Y. 1995. Erosion and sedimentation. Press Syndicate of the University of Cambridge, New York, NY.
Kalin, L., and M.M. Hantush. 2003. Probabilistic modeling of sediment yield in a small agricultural watershed using
    Kineros-2. In Conference Proceedings,  AWRA Specialty  Conference, Kansas City, MO, May 2003.
Kalin, L., and M.M. Hantush. 2003. Assessment of two physically-based watershed models based on their performances
    of simulating water and sediment movement", In Conference Proceedings, 1st Interagency Conference on Research
    in the Watersheds (ICRW), Benson, AZ, 27-30 October 2003.
Kalin, L., R.S. Govindaraju, and M.M. Hantush. 2003. Effect of geomorphologic resolution on runoff hydrograph and
    sedimentograph. J. Hydrol. 276:  89-111.
                                                   45

-------
Komor, S.C. 1999. Water Chemistry in a Nutrient and Sediment Control System near Owasco, New York. Water-
    Resources Impact, 1(6): 19-21.
Krone, R.B. 1962. Flume studies of the transport of sediment in estuarial shoaling processes. Hydraulic Engineering
    Laboratory and Sanitary Engineering Research Laboratory, University of California, Berkeley, CA.
Louks, D. P., J. Stedinger, and D. A. Haith. 1981. Water Resource Systems Planning and Analysis. Prentice-Hall, New
    Jersey.
Macintosh, D.L., G.W. Suter, and P.O. Hoffman. 1994. Uses of Probabilistic Exposure Models in Ecological Risk
    Assessments of Contaminated Sites, Risk Analysis, 14(4):405-419.
Molnar, D.K., and P.Y. Julien. 2000. Grid size effects on surface water modeling. Journal of Hydrologic Engineering
    5:8-16.
Mostaghimi, S., S.W. Park, R.A. Cooke, and S.Y. Wang. 1997. Assessment of Management Alternatives on a Small
    Agricultural Watershed. Water Research, 31(8): 1867-1878.
Mulvany, T.J. 1850. On the Use of Self-Registering Rain and Flood Gauges. Proc. Int. Civ. Eng., 4(2): 1-8.
Novotny, V., and H. Olem.  1994. Water Quality: Prevention, Identification, and Management of Diffuse Pollution. Van
    Nostrand, Reinhold, New York, NY.
O'Connor, D.J. 1960. Oxygen Balance of an Estuary. Journal of Sanitation Engineering Division, 86(3):35-55.
O'Connor, D.J. 1962. Organic Pollution of New York Harbor: Theoretical Considerations. Journal of Water Pollution
    Control, 34(9):905-919.
Ogden, F.L., and A. Heilig, 2001, Two-dimensional watershed-scale erosion modeling with CASC2D. In: Landscape
    Erosion and Evolution Modeling, (R.S. Harmon and W.W. Doe III, eds.), Kluwer Academic Publishers, New York,
    ISBN 0-306-4618-6, 535pp.
Ogden, F.L., and P.Y. Julien. 2002. CASC2D: A two-dimensional, physically-based, hortonian, hydrologic model. In:
    Mathematical Models of Small Watershed Hydrology and Applications, V. J. Singh and D. Freverts, eds., Water
    Resources Publications, Littleton, CO, ISBN 1-887201-35-1, 972 pp.
Ogden, F.L., and B. Saghafian. 1997. Green and Ampt infiltration with redistribution. Journal of Irrigation and Drainage
    Engineering, 123:386-393.
Osborn, H.B., and J.R.  Simanton. 1990.  Hydrologic modeling of a treated rangeland watershed. Journal of Range
    Management 43:474-481.
Rawls, W.J., D.L. Brakensiek, andK.E.  Saxton.  1982. Estimation of soil water properties. Trans. ASAE 25:1316-1320.
Rawls, W.J., D.L. Brakensiek. 1983. Green-Ampt infiltration parameters from soils data. Journal of Hydraulic
    Engineering, 109(1):62-70.
Refsgaard, J.C., andB. Storm. 1995. MIKE-SHE. Singh, V.J. (Ed), Computer Models of Watershed Hydrology, 809-
    846, Water Resources Pub., Highlands Ranch, CO.
Richards, L.A. 1931. Capillary conduction of liquids in porous mediums, Physics, 1:318-333.
Rosalia, R.S. 2002. GIS-based upland erosion modeling, geovisualization and grid size effects on erosion simulations
    with CASC2D-SED. Ph.D. Thesis, Colorado State University, Fort Collins, CO.
SAAESD 2001. Agricultural non-point source water quality models: their use and application. Southern Cooperative
    Series Bulletin #398, edited by J.E.  Parsons et al., ISBN: 1-58161-398-9.
Senarath, S.U.S., Ogden, F.L., Downer,  C.W., andH.O.  Sharif. 2000.  On the calibration and verification of two-
    dimensional, distributed, Hortonian, continuous watershed models. Water Resources Research, 36:1495-1510.
Shoemaker, L., M. Lahlou, M. Bryer, D. Kumar, and K.  Kratt. 1997. Compendium of tools for watershed assessment
    and TMDL development. Report EPA841-B-97-006, Office of Water, U.S. Environmental Protection Agency,
    Washington, D.C.
Singh, V.J., and D.A. Woolhiser. 2002. Mathematical Modeling of Watershed Hydrology. Journal of Hydrologic
    Engineering, 7(4):270-292.
Smith, D.D. 1941. Interpretation of Soil Conservation Data for Field Use. Agricultural Engineering, 22:173-175.
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.
Smith, R.E., D.C. Goodrich, and C.L. Unkrich. 1999. Simulation of Selected Events on the Catsop Catchment by
    KINEROS2: A Report for the GCTE Conference on Catchment Scale Erosion Models. Catena 37:457-475.
Srivastava, P., J.M. Hamlett, P.O. Robillard, and R.L. Day. 2002. Watershed optimization of best management practices
    using AnnAGNPS and a genetic algorithm.  Water Resources Research, 38(3):1-13.
Stanley, D.W. 1994. Regulation of Primary Production and Hypoxia in the Pamlico River Estuary, NC, by Factors
    Other Than Nutrients: Implications  for Management Strategies to Control Eutrophication. Program and Abstracts,
    IAGLR, Buffalo, NY.
Streeter, H.W., and E.B. Phelps. 1925. A Study of the Pollution and Natural Purification of the Ohio River. Ill: Factors
    Concerned in the Phenomena of Oxidation and Reareation, Bull. No.  146. US. Public Health Service.
                                                    46

-------
Tetra Tech Inc., 2000. Overview of Sediment-Contaminant Transport and Fate Models for Use in Making Site-Specific
    Contaminated Sediment Remedial Action Decisions. Prepared for US. EPA OERR, Tetra Tech Inc., Fairfax, Va.
Tollner, E.W., BJ. Barfield, C. Vachirakornwatana, and C.T. Haan. 1977. Sediment deposition patterns in simulated
    grass filters. Transactions of ASAE. 20(5):940-944.
Udoyara, S.T., J. Robert, and H.H. Liao. 1995. Impact of Landscape Feature and Feature Placement on Agricultural
    Non-Point-Source Pollution Control. Journal of Water Resources Planning and Management,  121(6):463-469.
U.S. EPA. 1999. Selection of Eutrophication Models for Total Maximum Daily Loads Analyses Based on Regulatory
    Environmental Modeling Guidelines. NERL, US. EPA.
U.S. EPA. 2000. The Quality of Our Nation's Waters. A Summary of the National Water Quality Inventory: 1998
    Report to Congress. Office of Water, Washington, D.C.  841-S-00-001.
U.S. EPA. 2002a. Risk Management Research Plan for Suspended Solids and Sediments. EPA600-R-02-002, Office of
    Research and Development, Cincinnati, OH. (91 pp).
U.S. EPA. 2002b. The Twenty Needs Report: How Research Can Improve the TMDL Program. EPA841-B-02-002,
    Office of Water, Washington, D.C. (43 pp).
Velz, CJ. 1938. Deoxygenation and Reoxygenation. ASCE Transactions, 560-572.
Ward, G.H., and J. Benaman. 1999. Models for TMDL Application in Texas Watercourses: Screening and Model
    Review. Report submitted to Texas Natural Resource Conservation Commission, Austin, TX. Online report
    CRWR-99-7.
WERF. 2001. Water Quality Models: A survey and Assessment. Edited by J.  Fitzpatrick, J. Imhoff, E. Burgess, and R.
    Brashear. Water Environment Research Foundation,  Project 99-WSM-5, Alexandria, VA.
Wischmeier, W.H., and D.D. Smith. 1958. Rainfall Energy and Its Relation to Soil Loss. American Geophysical Union
    Transactions.  39: 285-291.
Wischmeier, W.H., and D.D. Smith. 1965. Predicting Rainfall-Erosion Losses from Cropland East  of Rocky Mountains:
    Guide for Selection of Practices for Soil and Water Conservation USDA Agricultural Handbook, No 282.
Wischmeier, W.H., and D.D. Smith. 1978. Predicting Rainfall Erosion Losses: A Guide to Conservation Planning.
    USDA Agricultural Handbook, No  537.
Woolhiser, D.A. 1996. Search for physically based runoff model - A hydrologic El Dorado? Journal of Hydraulic
    Engineering. 3:122-129.
Woolhiser, D.A., R.E. Smith, and D.C. Goodrich. 1990. KINEROS-A kinematic runoff and erosion model:
    Documentation and user manual, USDA-ARS, ARS-77, 130 pp.
Yang, C.T. 1973. Incipient motion and sediment transport. Journal of Hydraulic Division, ASCE, 99, No. HY10:1679-
    1704.
Zhen, J.X., and L.Y. Shaw. 2001. Development of a Best Management Practice (BMP) Placement Strategy at the
    Watershed Scale. In Proceedings of the 2001 Wetland Engineering and River Restoration Conference, Aug 27-31,
    2001, Reno, NV.
Ziegler, A.D., T.W. Giambelluca, andR.A. Sutherland. 2001. Erosion Prediction on Unpaved Mountain Roads in
    Northern Thailand: Validation of Dynamic Erodibility Modeling Using KINEROS2. Hydrological Processes
    15:337-358.
Zingg, A.W. 1940. Degree and Length of Land Slope as It Affects Soil Loss in Runoff. Agricultural Engineering,
    21:59-64.
                                                   47

-------
Appendix: Model Summaries


    The model summaries provided here are mostly from model web sites (if available), model manuals and other sited
literatures. The URLs of the model web sites are given at the end of each summary, if exists.
8.1   Loading Models


AGNPS (AGricultural NonPoint Source pollution model) & AnnAGNPS (Annualized AGNPS): AGNPS, supported
by USDA-ARS-NRCS, was a single event model initially. The current version refers to system of modeling components
and is geared toward continuous simulations (daily time steps) of sediment and nutrient transport from agricultural
watersheds. The set of computer programs consist of: i) input generation & editing as well as associated databases, ii)
the "annualized"  science & technology pollutant loading model for agricultural-related watersheds (AnnAGNPS), iii)
output reformatting & analysis,  and iv) the integration of more comprehensive routines (CCHE1D) for the stream
network processes, v) a stream corridor model (CONCEPTS),  vi) an instream water temperature model (SNTEMP),
and vii) several related salmonid models (SIDO, Fry Emergence, Salmonid Total Life Stage, & Salmonid Economics).
Not all of the models are electronically linked but there are paths of common input/output that, with the use of standard
text editors, can be linked. The input programs include: i) a GIS-assisted computer program (TOPAZ with an interface
to AGNPS)  to develop terrain-following cells with all the needed hydrologic  & hydraulic parameters  that  can be
calculated from readily available DEM's, ii)  an input editor to initialize, complete, and/or revise the input data,  and iii)
an AGNPS-to-AnnAGNPS converter for the input data sets of the old single-event versions of AGNPS (4.03 & 5.00).
Watershed is divided into cells to reflect landscape spatial heterogeneity.  Several BMPs can be modeled including
ponds, vegetative filter strips, riparian buffers  and different management practices. AGNPS can be classified as an
empirical model. Runoff generation is based on unit hydrograph theory with total runoff being computed from SCS
curve number and peak discharge from TR-55. Sediment mobilized is calculated from RUSLE and sediment delivery is
based on  HUSLE. The latest version  of AnnAGNPS includes tile drainage, multiple climate file capabilities  and
enhanced lateral subsurface flow options. The basic model outputs are runoff volume, peak runoff rate, sediment yield,
sediment concentration, sediment particle size distribution, upland erosion, amount of deposition (%), enrichment ratios
by particle size, delivery ratios by particle size, nitrogen, phosphorus,  and chemical oxygen demand. Efforts are going
on to integrate REMM (Riparian Ecosystem Management Model) to AGNPS system.
BMPs: Agricultural practices, ponds, grassed waterways, tile drainage, vegetative filter strips, riparian buffers.
URL:  http://www.sedlab.olemiss.edu/agnps.html

Application and Model References:
Bingner, R., C. Murphree, and C. Mutchler. 1989: Comparison of sediment yield models on watershed in Mississippi.
    Trans. ASAE, 32(2):529-534.
Bingner, R. L., andF. 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, 25-29 March 2001. p. VII-1 - VII-8.
Fisher, P., R. Abrahart, and W. Herbinger. 1997. The sensitivity of two distributed non-point source pollution models to
    the spatial arrangement of the landscape. Hydrological Processes, ll(3):241-252.
McCool, D.K., M.T. Walter, andL.G. King.  1995. Runoff index values for frozen soil areas of the Pacific Northwest.
    Journal of Soil and Water Conservation,  50(5):466-469.
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 Resources Research, 38(3):1-13.
                                                   48

-------
Yuan, Y., Dabney, S., and Bingner, R. L. 2002. Cost/benefit analysis of agricultural BMPs for sediment reduction in the
    Mississippi Delta. Journal of Soil and Water Conservation 57(5): 259-267.
Wu, T., J. Hall, and J. Bonta. 1993. Evaluation of Runoff and Erosion Models. Journal of Irrigation and Drainage
    Engineering, 119(2):364-382.
Zhen, J.X., andL.Y. Shaw. 2001. Development of a best management practice (BMP) placement strategy at the
    watershed scale. In Proc. Wetlands Engineering and River Restoration Conference, August 27-31, 2001, Reno,
    Nevada.


AGWA (Automated  Geospatial Watershed Assessment): This is a  GIS interface developed  by  The USDA-ARS
Southwest  Watershed Research Center, in cooperation with the U.S. EPA Office of Research and Development to
facilitate the data preparation efforts of two USDA models: SWAT for large watersheds  and term simulations, and
KINEROS-2 for small watersheds (<100 km2) for event based studies (see corresponding model descriptions below for
details on SWAT and KINEROS-2). AGWA is designed as a tool for performing relative assessment (change analysis)
resulting from land cover/use change. Areas  identified through large-scale assessment with SWAT as being most
susceptible to change can be evaluated in more detail at smaller scales with KINEROS-2. Data used in AGWA include
Digital Elevation Models (OEMs), land cover grids, soils data, and precipitation data. It is built on Arc View version 3.X
and the interface is similar to USEPA's BASINS. There are five major tasks: i) watershed delineation, ii) land cover and
soils parameterization, iii) writing a precipitation file for model input, iv) writing parameter files and running the chosen
model, and v) viewing results. To use AGWA,  ARcView version 3.1 or later of Arc View and version 1.1 of the Spatial
Analyst extension is required.
URL: http://www.tucson.ars.ag.gov/agwa


ANSWERS (Areal Nonpoint Source Watershed Environment Response Simulation) & ANSWERS-2000: ANSWERS
is an event based, distributed parameter, physically-based, watershed scale, upland planning model  developed for
evaluating the effectiveness of agricultural and urban BMPs in reducing sediment and nutrient delivery to streams in
surface runoff and leaching of nitrogen through the root zone. The model is  intended for use by planners on ungaged
watersheds where data for model calibration is not available. It divides the area into uniform grid squares (less than 1
hectare), where all properties are assumed homogeneous. ANSWERS-2000 is the continuous version of the model. Both
versions simulate  interception; surface retention/detention; infiltration; percolation; sediment detachment and transport
of mixed particle size classes  in rills, interrill areas, and channels. The continuous version,  in addition, simulates  crop
growth, evapotranspiration,  soil moisture redistribution, plant uptake of nutrients; N and P dynamics in the soil; nitrate
leaching; and losses of nitrate, ammonium, total Kjeldahl nitrogen, and P in surface runoff. Event based version  uses
Holton model to  simulate infiltration, whereas Green-Ampt model  is employed in  the continuous version. A  GIS
interface of the event version with GRASS is available. The continuous version has an Arc View based user interface,
QUESTIONS, that facilitates data file creation and manipulation. Model documentation and user support is very limited
for the continuous version.  The model is currently only suitable for use by expert modelers with a good knowledge of
upland hydrology and agriculture. The current version of the model makes heavy use of relationships derived from the
WEPP and EPIC models.
BMPs: Agricultural practices, ponds, grassed waterways, tile drainage.
URL: http://dillaha.bse.vt.edu/answers/index.htm

Application and Model References:
Beasley, D.B., L.F.Huggins, and E. J. Monke. 1980. ANSWERS: A model for watershed planning.  Trans, of the ASAE
    23(4):938-944.
Beasley, D.B., and L.F. Muggins. 1991. ANSWERS Users Manual, 2nd Ed. Agricultural Engr. Dept, Coastal Plain
    Experiment Station, Univ. of Georgia, Tifton, GA.
Bouraoui, F., and T. A. Dillaha. 1996. ANSWERS-2000: Runoff and sediment transport model. Journal of
    Environmental Engineering, ASCE 122(6):493-502.
Bouraoui, F., and T.A. Dillaha. 2000.  ANSWERS-2000: Nonpoint source nutrient transport model. J. of Environmental
    Engineering, ASCE 126(11): 1045-1055.
Bouraoui, F., G. Vachaud, R. Haverkamp and B. Normand. 1997. A distributed physical approach for surface-
    subsurface water transport modeling in agricultural watersheds. J. of Hydrology 203:79-92.
Fisher, P., R. Abrahart, and W. Herbinger.  1997. The sensitivity of two distributed non-point source pollution models to
    the spatial arrangement of the landscape. Hydrological Processes, ll(3):241-252.
                                                    49

-------
Montas, H.J., C.A. Madramootoo. 1991. Using ANSWERS model to predict runoff and soil loss in Southwestern
    Quebec. Transactions of the ASAE 34(4):1752-1762.
Srinivasan, R, and J. Arnold. 1994. Integration of a basin-scale water quality model with GIS. Water Resources
    Bulletin, 30(3):453-462.
Storm, D. E., T. A. Dillaha, S. Mostaghimi, and V. O. Shanholtz. 1988. Modeling phosphorus transport in surface
    runoff. Transactions of the ASAE 31(1): 117-127.


BASINS (Better Assessment  Science  Integrating point  and  Nonpoint  Sources):  BASINS  is  a multipurpose
environmental analysis system for use by regional, state, and local agencies in performing watershed and water quality
based studies. The heart of BASINS is its suite of interrelated components essential for performing watershed and water
quality analysis. These components are grouped into several categories:
    •   Nationally derived environmental and GIS databases (the 48 continuous states and the District of Columbia)
    •   Assessment tools (TARGET, ASSESS, and DATA MINING) for evaluating water quality and point source
        loadings at a large or small scales
    •   Utilities including local data import and management of local water quality observation data
    •   Two watershed delineation tools
    •   Utilities for classifying elevation (DEM), land use, soils, and water quality data
    •   An in-stream water quality model (QUAL2E)
    •   A  simplified GIS based nonpoint source annual loading model (PLOAD)
    •   Two watershed loading and transport models (HSPF and SWAT)
    •   A  postprocessor  (GenScn) of model data and scenario generator to visualize,  analyze, and compare results
        from HSPF and SWAT
    •   Many mapping, graphing, and reporting formats for documentation.
BASINS' databases  and assessment tools are directly integrated within an Arc View GIS environment.  The simulation
models run in a Windows environment, using data input files generated in Arc View. EPA is working on expanding
BASINS system to include three dimensional water quality model EFDC.
URL: http://www.epa.gov/OST/BASINS


'DWSM (Dynamic  Watershed Simulation Model): DWSM  was developed at the  Illinois State Water Survey. It
simulates surface and  subsurface  flow, upland soil  erosion, sediment  transport,  and  agrochemical transport in
agricultural and rural watersheds. It is a one dimensional, event based model. Rainfall excess at overland flow planes
can be computed in two ways:  i) Curve number method, ii) Smith-Parlange infiltration model. Kinematic  Wave
equations are solved using analytical and an approximate shock fitting solutions to compute runoff over planes and
channels. Flows in reservoirs are based modified pulse method. Subsurface flow is a combination of interflow, tile drain
flow and base flow.  Soil erosion is based on raindrop detachment and hydraulic erosion. Scour and deposition of user
defined particle sizes is computed based  on sediment transport capacity. Approximate analytic solution of temporal and
spatially varying continuity equation is employed. All sediments entering the reservoirs are assumed trapped.  Nutrients
and pesticides are simulated in dissolved and adsorbed phases with water and sediment respectively. The watershed is
divided into overland planes, channel segments, and reservoir units. 18 applications of the model or its components are
available in the literature. All these applications are performed by the model developers.
BMPs: Detention basins, alternative ground covers, tile drainage.

Application and Model References:
Borah, O.K. and M. Bera.  2003. Watershed scale hydrologic and nonpoint source pollution models: review of
    mathematical bases. Transactions of the ASAE. Uner review.
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, 113-166.  Singh and D.K. Frevert eds.  Water Resources
    Publications, LLC, Highlands Ranch, CO.
Borah, D.K, R.Xia, and M. Bera. 2002. Watershed model to study hydrology, sediment, and agricultural chemicals in
    rural watersheds. In Surface Water Hydrology Vol-1, 343-358. V.P Singh, M. Al-Rashed, and M.M. Sherif eds.
    A.A. Balkema Publishers, Lisse/Abingdon/Exton (PA)/Tokyo.
                                                    50

-------
EPIC (Erosion-Productivity Impact Calculator): EPIC was developed to assess the effect of soil erosion on soil
productivity. EPIC is a continuous simulation model that can be used to determine the effect of management strategies
on agricultural production and soil and water resources. The drainage area considered by EPIC is generally a field-sized
area, up to 100 ha (weather, soils, and management systems are assumed to be homogeneous). 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.  Recently, most of the EPIC model development has
been focused on  problems involving water quality and global  climate/CO2 change. Example additions include the
GLEAMS (Leonard et al., 1987) pesticide  fate component, nitrification and volatilization submodels, a new more
physically based wind erosion component, optional SCS technology for estimating peak runoff rates, newly developed
sediment yield equations, and mechanisms for simulating CO2 effects on crop growth and water use. These and other
less significant developments extend EPIC's  capabilities to deal with a wide  variety  of agricultural management
problems. Example applications include:
    •   1985 RCA analysis
    •   1988 drought assessment
    •   soil loss tolerance tool
    •   Australian sugarcane model (AUSCANE)
    •   pine tree  growth simulator
    •   global climate change analysis (effect of CO2, temperature, and precipitation change on runoff and crop yield)
    •   farm level planning
    •   five-nation EEC assessment of environmental/agricultural policy alternatives
    •   Argentine assessment of erosion/ productivity
    •   USD A-Water Quality Demonstration Project Evaluation
    •   N leaching index national analysis.
BMPs: Agricultural practices.
URL: http://www.brc.tamus.edu/epic

Application and Model References:
Benson, V.W., K.N. Potter, H.C. Bogusch, D. Goss, and J.R. Williams. 1992. Nitrogen leaching sensitivity to
    evapotranspiration and soil water storage estimates in EPIC. J. Soil and Water cons. 47(4):334-337.
Leonard, R.A., W.G. Knisel, andD.A.  Still. 1987. GLEAMS: Groundwater loading effects on agricultural management
    systems. Trans. ASAE 30(5): 1403-1428.
Vijay, P. S., J.R. Williams. 1995. The EPIC model. Computer Models of Watershed Hydrology, chapter 25. Water
    Resources Publications, Highlands Ranch, Colorado.
Williams, J. 1995. The EPIC model. Chap. 25, Computer models of watershed hydrology (V.P. Singh, ed.), pp. 909-
    1000. Highlands Ranch, CO: Water Resources Publications.
Williams, J.R., J.R. Kiniry, and V.W. Benson. 1991. Water quality sensitivity to EPIC crop growth parameters. ASAE
    Paper No. 91-2075.
Williams, J.R., C.A. Jones, and P.T. Dyke. 1990. The EPIC model. Chapter 2, pp. 3-92. In: A.N. Sharpley and J.R.
    Williams (eds.) EPIC-Erosion/Productivity Impact Calculator: 1. Model Documentation. USDA Tech. Bull. No.
    1768.
Williams, J.R., P.T. Dyke, W.W. Fuchs, V.W. Benson, O.W. Rice, and E.D. Taylor. 1990. EPIC Erosion/Productivity
    Impact Calculator: 2. User Manual. In : A.N. Sharpley and J.R. Williams (eds.) USDA Tech. Bull. No. 1768.


GLEAMS (Groundwater  Loading Effects of Agricultural Management Systems):  GLEAMS is a continuous
simulation, field  scale model, which was developed  as  an extension of the  Chemicals,  Runoff and Erosion from
Agricultural Management Systems (CREAMS) model. GLEAMS assumes that a field has  homogeneous land use, soils,
and precipitation. It consists  of four major  components: hydrology, erosion/sediment yield, pesticide transport, and
nutrients. GLEAMS was developed to evaluate the impact of management practices on potential pesticide and nutrient
leaching within, through, and below the root zone. It also estimates surface runoff and sediment losses from the field.
GLEAMS was not developed as an absolute predictor of pollutant loading. It is a tool for comparative analysis of
complex pesticide chemistry, soil properties, and climate. 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. Application rates, methods, and timing can be altered to account for these  systems and to reduce
the possibility of  root zone leaching. The model also accounts for varying soils and weather in determining leaching
potential. GLEAMS can also  be useful in simulations for pesticide screening of soil/management. The model tracks
                                                    51

-------
      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 products. Flow is determined by SCS curve number method. Erosion in overland flow areas is
      estimated using modified USLE. 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.
      BMPs: Agricultural practices, ponds.
      URL: http://arsservO.tamu.edu/nrsu/glmsfact.htm. http://www.cpes.peachnet.edu/sewrl/Gleams/gleams y2k update.htm

      Application and Model References:
      Knisel, W.G., and J.R. Williams. 1995. Hydrology components of CREAMS and GLEAMS models. In: V. J. Singh
          (Ed.) Computer Models of Watershed Hydrology. Chapter 28. pp. 1069-1114.
      Knisel, W.G., and E. Turtola.  1999. GLEAMS model application on a heavy clay soil in Finland. Agricultural Water
          Management, 43(3):285-309.
      Leonard, R.A., W.G. Knisel, and P.M. Davis. 1995. Modeling pesticide fate with GLEAMS. Eur. J. Agron.  4(4):485-
          490.
      Morari, F., and W.G. Knisel. 1997. Modifications of the GLEAMS model for crack flow. Trans., Amer. Soc. of Agric.
          Engrs., 40(5): 1337-1348.
      Michael, J.L., M.C. Smith, W.G. Knisel, D.G. Neary, W.P. Fowler, and D.J. Turton. 1996. Using a hydrologic model to
          determine the most environmentally safe windows for herbicide application. New Zealand J. of Forestry Science,
          26:288-297.
      Shirmohammadi, A., B. Ulen, L.F. Bergstrom, and W.G. Knisel. 1998. Simulation of nitrogen and phosphorus leaching
          in a structured soil using GLEAMS and a new submodel, "PARTLE". Trans. Amer. Soc. Of Agric. Engrs.,
          41(2):353-360.
      Sugiharto, T; T. Mclntosh, R. Uhrig, and J. Lardinois. 1994. Modeling alternatives to reduce dairy farm and watershed
          nonpoint source pollution. Journal of Environmental Quality, 23(l):18-24.


      GSSHA (Gridded Surface Subsurface Hydrologic Analysis): This is  a reformulation and enhancement  of CASC2D
      (Downer and Ogden 2002). The CASC2D model was initiated at Colorado State University by Pierre Julien as a two
      dimensional  overland flow routing model. In its final form, it is a distributed-parameter, physically-based watershed
      model. Both single event  and continuous simulations are possible.  The US Army  Waterways Experiment  Station
      considered this model as very promising and therefore fully incorporated this model into WMS (Watershed Modeling
      System). Watershed is divided into cells and water and sediment is routed from one cell to another. It uses one and two-
      dimensional diffusive wave flow routing at channels and overland planes, respectively. Although only Hortonian flows
      were modeled by employing Green-Ampt (G-A) infiltration model in the initial versions, GSSHA considers other runoff
      generating mechanisms such as lateral saturated groundwater flow, exfiltration, stream/groundwater interaction etc.
      GSSHA offers two options for simulations: G-A with redistribution (Ogden and Saghafian 1997) and the full Richards'
      equation. The latter requires tremendous amount of simulation time and is very sensitive to time step and horizontal and
      vertical cell  sizes. Modified Kilinc and Richardson equation (Julien 1995) is used  to compute sediment  transport
      capacity at plane cells. A trap efficiency measure is used to determine how much material  is transported  from the
      outgoing cell. Details on theory and equations used can be found in Julien et al. 1995, Johnson  et al. 2000, and Downer
      and Ogden 2002. GSSHA is currently available under the WMS suite of models which significantly reduces burden on
      input preparation.
      Contact info:
        Fred L. Ogden,                                        Brian E. Skahill
        309 F.L. Castleman Building                           Watershed Systems Group
        Civil and Environmental Engineering, U-37               Coastal and Hydraulics Laboratory
        University of Connecticut                              Engineer Research and Development Center
        Storrs, CT 06269-2037                                 ATTN: CEERD-HC-HW
        Phone: (860) 486-2771                                 3909 Halls Ferry Road
        Fax: (860) 486-2298                                   Vicksburg, MS 39180-6199
        ogden(@,engr.uconn.edu                                Phone: 601-634-3441
                                                             Fax: 601-634-4208
                                                             BriaaE. Skahill@erdc.usace.army.mil
BMPs: Agricultural practices.
                                                          52

-------
Application and Model 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, Engineering Research and Development Center, Vicksburg, MS.
Downer, C.W., and F.L. Ogden. 2003. Prediction of runoff and soil moisture at the watershed scale: effects of model complexity
    and parameter assignment. Water Resources Research, 39(3): 1045.
Julien, P.Y. 1995. Erosion and Sedimentation, Press Syndicate of the University of Cambridge, New York, N.Y.
Julien, P.Y., B. Saghafian, and F.L. Ogden. 1995. Raster-based hydrologic modeling of spatially-varied surface runoff. Water
    Resources Bulletin 31:523-536.
Johnson, B.E., P.Y. Julien, O.K. Molnar, and C.C. Watson. 2000. The two-dimensional upland erosion model CASC2D-SED.
    Journal of American Water Resources Association 36:31-41.
Ogden, F.L., and A. Heilig, 2001, Two-dimensional watershed-scale erosion modeling with CASC2D, in Landscape Erosion and
    Evolution Modeling, (R.S. Harmon and W.W. Doe III, eds.), Kluwer Academic Publishers, New York, ISBN 0-306-4618-6,
    535 pp.
Senarath, S., F.L. Ogden, C.W. Downer, andH.O. Sharif. 2000. On the calibration and verification of two-dimensional,
    distributed, Hortonian, continuous watershed models. Water Resources Research, 36(6): 1495-1510.


GWLF (Generalized Watershed Loading Functions): GWLF model was developed by Haith and Shoemaker (1987). The GWLF
model provides the ability to simulate runoff, sediment, and nutrient (N and P) loadings from a watershed given variable-size
source areas (i.e., agricultural, forested, and developed land). It also has algorithms for calculating septic system loads, and allows
for the inclusion of point source discharge data. It is a continuous simulation model which uses daily time steps for weather data
and water balance calculations. Monthly calculations are made for sediment and nutrient loads, based on the daily water balance
accumulated to monthly values.  GWLF is  considered to be a combined distributed/lumped parameter  watershed model. For
surface loading, it is distributed  in the sense that it allows multiple land use/cover  scenarios, but each area is assumed to be
homogenous in regard to various attributes considered by the model.  Additionally, the model does not spatially distribute the
source areas, but simply aggregates the loads from each area into a watershed total; in other words there is no spatial routing. For
sub-surface loading, the model acts as a lumped parameter model using a water balance approach. No distinctly separate areas are
considered for sub-surface flow contributions. Daily water balances are computed for an unsaturated zone as well as a saturated
sub-surface zone, where infiltration is simply computed as the difference between precipitation and snowmelt minus surface
runoff plus evapotranspiration. With respect to the  major processes simulated, GWLF models surface runoff using the SCS-CN
approach with daily weather (temperature and precipitation) inputs.   Erosion and sediment yield are estimated using monthly
erosion calculations based on the USLE algorithm (with monthly rainfall-runoff coefficients)  and a monthly composite of KLSCP
values for each source area (i.e., land  cover/soil type  combination).   A sediment delivery ratio based on watershed size and a
transport capacity based on average daily runoff are then applied to the calculated erosion to determine  sediment yield for each
source area. Surface nutrient losses are determined by applying dissolved N and P coefficients to surface runoff and a sediment
coefficient to the yield portion for each agricultural  source area. Point source discharges can also contribute to dissolved losses
and are specified in terms of kilograms per month. Manured areas,  as well as septic systems, can also be considered.  Urban
nutrient inputs are all assumed to  be solid-phase, and the model uses an exponential accumulation and washoff function for these
loadings. Sub-surface losses are calculated using dissolved N and P coefficients for shallow groundwater contributions to stream
nutrient loads, and the sub-surface sub-model only considers a single, lumped-parameter contributing area. Evapotranspiration is
determined using daily weather data and a cover factor dependent upon land use/cover type. Finally, a water balance is performed
daily using supplied or computed precipitation, snowmelt, initial unsaturated zone storage, maximum available zone storage, and
evapotranspiration values. An Arc View interface of the model is available called AVGWLF.
BMPs: Agricultural practices, septic systems, manured areas.
l/^Z: http://www.avgwlf.psu.edu/AVGWLFmanual.htnrfGWLFModel

Application and Mode I References:
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.R., R. Mandel, and R.S. Wu, 1992.  GWLF: Generalized Watershed Loading Functions User's Manual, Vers. 2.0.
    Cornell University, Ithaca, NY.
Howarth, R., J. Fruci, andD.  Sherman.  1991. Inputs of sediment and carbon to anestuarine ecosystem: influence of land use.
    Ecological applications 1:27-39.
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(4):833-847.
                                                         53

-------
HSPF (Hydrological Simulation Program): HSPF simulates for extended periods of time the hydrologic, and associated water
quality, processes on pervious and impervious land surfaces and in streams and well-mixed impoundments. It is supported by
both USEPA and USGS. It is incorporated into the BASINS and WMS modeling  systems. The model contains hundreds of
process algorithms developed from theory, laboratory experiments, and empirical relations from instrumented watersheds. There
are three basic modules: PERLND and IMPLND watershed loading models with former for pervious surfaces  and latter for
impervious surfaces. RCHRES is a one-dimensional stream model  serving  as the receiving water model. It is  based on the
Stanford Watershed  Model, ARM (Agricultural Runoff Management) and NFS (NonPoint Source) models. It uses simple storage
based equations for flow routing. Flows in streams are one-dimensional. It is one of the few comprehensive models of watershed
hydrology and water quality that allows the integrated simulation of land and soil contaminant runoff processes with in-stream
hydraulic  and sediment-chemical  interactions.  HSPF uses continuous rainfall  and other  meteorologic records to  compute
streamflow hydrographs and pollutographs. HSPF simulates interception soil moisture, surface runoff, interflow, baseflow,
snowpack depth and water content, snowmelt, evapotranspiration, ground-water recharge, dissolved oxygen, biochemical oxygen
demand (BOD), temperature, pesticides, conservatives, fecal coliforms, sediment detachment and transport, sediment routing by
particle size,  channel  routing,  reservoir  routing,  constituent routing,  pH,  ammonia,  nitrite-nitrate,  organic nitrogen,
orthophosphate, organic phosphorus, phytoplankton, and zooplankton. Program can simulate one or many pervious or impervious
unit areas  discharging to one or many river reaches or reservoirs. Frequency-duration analysis can be done for any time series.
Any time step from  1 minute to 1 day that divides equally into  1 day can be used. Any period from a few minutes to hundreds of
years  may be simulated. HSPF is generally used to assess the effects  of land-use change, reservoir operations, point or nonpoint
source treatment alternatives, flow diversions, etc. Programs, available separately,  support data preprocessing and postprocessing
for statistical and graphical analysis of data saved to the Watershed Data Management (WDM) file. The major application of
HSPF is the Chesapeake Bay Project.
BMPs: Nutrient and  pesticide management, ponds.
URL:  http://water.usgs.gov/software/hspf.html

Application and Mode I References:
Bicknell, B.R., J.C. Imhoff, J.L. Kittle, A.S. Donigian, and R.C. Johanson. 1997. Hydrological Simulation Program - FORTRAN.
    User's Manual for Release 11. EPA/600/R-97/080. U.S. EPA Environmental Research Laboratory, Athens, GA.
Chen, Y.D., DJ. Norton, and J.P. Craig. 1996. Enhancement and Application of HSPF for Stream Temperature Simulation in
    Upper Grande Ronde Watershed, Oregon. Published in Proceedings. Watershed '96. US Environmental Protection Agency.
    June 8-12.
Donigian,  A.S., Jr., B.R. Bicknell, and J.C. Imhoff. 1995. Hydrologic Simulation Program -FORTRAN (HSPF). Chapter 12 in
    Computer Models of Watershed Hydrology, V.P. Singh, Ed., Water Resources Publications, Littleton, CO.
Donigian,  A.S., J.C.  Imhoff, B.R. Bicknell and J.L. Kittle. 1984. Application Guide for Hydrological Simulation Program -
    FORTRAN (HSPF). EPA- 600/3-84-065. Office of Research and Development, U.S. Environmental Protection Agency,
    Athens, GA.
Donigian,  A.S., B.R. Bicknell, A.S. Patwardhan, L.C. Linker, C.H. Chang, and R. Reynolds. 1994. Chesapeake Bay Program -
    Watershed Model Application to Calculate Bay Nutrient Loadings: Final Findings and Recommendations (FINAL
    REPORT). Prepared for U.S. EPA Chesapeake Bay Program, Annapolis, Maryland.
Fontaine, T., and V.  Jacomino. 1997. Sensitivity analysis of simulated contaminated sediment transport. J. Amer. Water Res.
    Assn., 33(2):313-326.
Jacomino, VMF; Fields, DE. 1997. A critical approach to the calibration of a watershed model. Journal of the American Water
    Resources Association, 33(1): 143-154.
Laroche, A., J. Gallichand, R. Lagace, and A. Pesant. 1996.  Simulating atrazine transport with HSPF in an agricultural watershed.
    J. Envir. Engr,  122 (7):622-630.


KINEROS-2 (KINematic EROSion model): This is the improved version of KINEROS (Woolhiser et al., 1990). It is event based
since  it lacks a true soil moisture redistribution formulation for long rainfall hiatus and more importantly  it does not consider
evapotranspiration (ET) losses.  This model is primarily useful for predicting surface runoff and erosion over small agricultural
and urban watersheds. Smith et al. 1995 suggest watershed  size smaller than 1000 ha for best results. Runoff is calculated based
on the Hortonian approach using a modified version of Smith- Parlange (Smith and Parlange 1978) infiltration model. KINEROS-
2 requires the watershed divided into homogeneous overland flow planes and  channel segments, and routs water movement  over
these  elements in a  cascading fashion. Mass balance and the kinematic wave approximations to the Saint Venant equations are
solved with implicit finite difference numerical scheme in a  1-D framework. KINEROS-2 accounts for erosion resulting from
raindrop energy and by flowing water separately. A mass balance equation is  solved to describe sediment dynamics at any point
along a surface flow path. Erosion is based on maximum transport capacity  determined by Engelund-Hansen equation (1967).
The rate of sediment transfer between soil and water is defined with a first order uptake rate. KINEROS-2 can be used under the
                                                        54

-------
AGWA system which provides a GIS interface for data preparation and visualization of results. A detailed description of the
model  and  the  equations  used  can be found   in  Smith  et  al.  1995  and  at  the  official URL  of the  model:
http://www.tucson.ars.ag.gov/kineros.
BMPs: Agricultural practices, detention basins, culverts.

Application and Model References:
Kalin, L., R.S. Govindaraju, M.M. Hantush. 2003. Effect of geomorphologic resolution on runoff hydrograph and
    sedimentograph. J. Hydrol., 276:89-111.
Kalin, L., and M.M. Hantush. 2003. Modeling of sediment yield in a small agricultural watershed with KINEROS-2. In J. D.
    Williams and D. W. Koplin, ed., American Water Resources Association 2003 Spring Specialty Conference on Agricultural
    Hydrology & Water Quality, Kansas City, MO, CD-ROM.
Lane, L. J., D.A. Woolhiser, and V. Yevjevich. 1975. Influence of simplifications in watershed geometry in simulation of surface
    runoff. Hydrology paper No. 81, Colorado State University, Fort Collins, CO. 27 pp.
Smith, R.E., D.C. Goodrich, D.A. Woolhiser and C.L. Unkrich. 1995. A kinematic runoff and erosion model. Singh, V.J. (Ed),
    Computer Models of Watershed Hydrology, 697-732, Water Resources Pub., Highlands Ranch, CO.
Smith, R.E., and J.Y. Parlange. 1978. A parameter-efficient hydrologic infiltration model. Water Resources Research, 14:553-
    538.
Woolhiser, D.A., R.E. Smith, and D.C. Goodrich. 1990. KINEROS-A kinematic runoff and erosion model: Documentation and
    user manual. USDA-ARS, ARS-77.
Zevenbergen, L.W., and M.R. Peterson. 1988. Evaluation and testing of storm event hydrologic models. Proc. ASCE Nat. Conf.
    On Hydraulic Engr, Colorado Springs, CO, Aug. 6-12., p. 467472.
Ziegler, A.D., T.W. Giambelluca, and R.A. Sutherland. 2001. Erosion prediction on unpaved mountain roads in northern
    Thailand: Validation of dynamic credibility modeling using KINEROS2. Hydrological Processes, 15:337-358.


MIKE-11:  MIKE-11 is a software  tool for the simulation of hydrology, hydraulics, water quality  and sediment transport in
estuaries, rivers, irrigation systems and other inland waters. It is based on an integrated modular structure with a variety of basic
modules and add-on modules, each simulating certain phenomena in river systems. Each module  can be operated separately and
data transfer between modules is automatic.  Coupling of physical processes (e.g. river morphology, sediment  re-suspension, and
water quality) are facilitated. MIKE-11 includes basic modules for:
    •   Rainfall-runoff (RR): This  module contains  three different models that can be used to estimate  catchment runoff: i)
        NAM is a lumped, conceptual rainfall-runoff model simulating overland flow, interflow and baseflow as a function of
        the moisture content in four mutually interrelated storages:  snow storage, surface storage, root  zone storage,  and
        groundwater storage. In addition NAM allows treatment of man-made interventions in  the hydrological cycle such as
        irrigation and groundwater pumping, ii) The present UHM module simulates the runoff from single storm events by the
        use of the unit hydrograph technique  and constitutes an alternative to the NAM model for flood simulation in areas
        where no stream flow records are available or where unit hydrograph techniques have already been well established. The
        module calculates simultaneously  the  runoff from several catchments and includes  facilities for presentation  and
        extraction of the  results. The output from the module can be used  as lateral inflow to the advanced hydrodynamic
        module in MIKE-11, iii) SMAP: A monthly soil moisture accounting model. The RR module can either be applied
        independently or used to represent one or more contributing catchments that generate lateral inflows  to a river network.
        In this manner it is possible to treat a single catchment or a large river basin containing numerous catchments and a
        complex network of rivers and channels within the same modeling framework.  An auto-calibration tool is available for
        the NAM module which uses a global optimization routine called the Shuffled Complex Evolution (SCE) algorithm.
    •   Hydrodynamics (HD): The HD module 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 non-linear 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: i)
        high-order, fully dynamic, ii) diffusive wave, iii) kinematic wave, and iv) quasi-steady state. Within the standard HD
        module advanced computational formulations enable flow over a variety of structures  to be simulated: broad-crested
        weirs, culverts, regulating  structures,  control structures, dam-break structures, user-defined structures, and tabulated
        structures.
    •   Advection-dispersion and cohesive sediments  (AD):  The AD module is based on the one-dimensional equation of
        conservation of mass of a dissolved or  suspended material (e.g.,  salt or cohesive  sediments).  The behavior of
                                                         55

-------
        conservative materials which decay linearly can be simulated. The module requires output from the hydrodynamic
        module, in space and time, of discharge and water level, cross-sectional area and hydraulic radius. The module includes
        a description of the erosion and deposition of cohesive sediment. Erosion and deposition are modeled as source/sink
        terms in the advection-dispersion equation. Whereas the erosion rate depends on the local hydraulic conditions, the
        deposition rate  depends also on the concentration of suspended sediment. It is also possible to  simulate non-cohesive
        sediments with  the AD module. Here the transport of the suspended sediment is described with the advection-dispersion
        equation, and the erosion and deposition terms are described by conventional sediment transport formulations.
    •   Water quality (WQ): WQ is coupled to the advection-dispersion (AD) module and simulates the reaction processes of
        multi-compound systems including the degradation of organic matter, the photosynthesis and respiration of plants,
        nitrification  and the exchange of oxygen with  the atmosphere. The mass balance for the parameters involved are
        calculated for all grid points at all time steps using a rational extrapolation method in an integrated two-step procedure
        with the  AD module. A number of modules have been developed describing BOD-DO relationships, nitrification, the
        influence of bed vegetation on water quality, sedimentation and re-suspension, and oxygen consumption from reduced
        chemicals. Two add-on modules are available for the WQ-module: Water Quality Heavy Metals module (WQHM), and
        the Eutrophication module (EU).
    •   Non-cohesive sediment transport: The non-cohesive sediment transport module (ST) can be used to study the sediment
        transport and morphological conditions in rivers. The features include:  i) five models for the calculation of sediment
        transport  capacity:  Engelund-Hansen, Ackers-White,  Engelund-Fredsee, van Rijn and  Smart Jeaggi,  ii)  sediment
        description by an average particle size and standard deviation of the grain size distribution, iii) explicit (no feedback with
        HD) or morphological (with feedback via sediment continuity  and bed resistance) models, and  iv) output of sediment
        transport rates, bed level changes, resistance numbers and dune dimensions.
An Arc View interface of the model is available which facilitates input data preparation and output visualization.  The US Federal
Emergency Management Agency (FEMA) has recently  approved and  included MIKE-11  on their list of hydraulic models
accepted for use in the National Flood Insurance Programme (NFIP).
URL: http://www.dhisoftware.com/mikel 1

Application and Model References:
Please visit http://www.dhi.dk/ContactUs/Library for all DHI compendium of technical papers and publications.


MIKE-SHE: MIKE-SHE is a distributed, physically based, dynamic modeling tool that can simulate the entire land phase of the
hydrologic cycle. It has the capability of handling both single  events and  continuous simulations. Watershed is divided into
square grid cells.  Overland flow routing is based on 2-D  diffusive wave equations whereas options vary  for channel flow from
simple Muskingum routing to the Higher Order Dynamic Wave formulation of the Saint-Venant equations. Ground water flow is
solved with 3-D full Richards' equation. Stream-ground water interactions are considered. In general, depending on the size of the
watershed, simulations can be computationally very intensive. Typical MIKE-SHE applications are:
    •   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
Arc View interface is available. Most of the applications find in literature belong to the model developers.
BMPs: Agricultural practices, wetlands, nutrient and pesticide management.
URL: http://www.dhisoftware.com/mikeshe

Application and Model References:
Abbott, M.., J. Bathurst,  P.Cunge, P. O'Connel, and J. Rasmussen. 1986.  An introduction to the European Hydrologic Systern-
    Systeme-Hydroloque European, SHE,  1: History and Philosophy of a physically-based distributed modeling system. Journal
    of Hydrology (87):45-59.
                                                         56

-------
Abbott, M.., J. Bathurst, P.Cunge, P. O'Connel, and J. Rasmussen. 1986. An introduction to the European Hydrologic System-
    Systeme-Hydroloque European, SHE, 2: Structure of a physically-based distributed modeling system. Journal of Hydrology
    87: 61-77.
Abbot, M., and Refsgaard (eds.). 1996. Distributed Hydrologic Modeling. Kluwer Academic Publishers, Dodrecht.
Gustafsson, L.G., S. Winberg, A. Refsgaard.  1997. Towards a distributed physically based model description of the urban aquatic
    environment. Water Science & Technology, 36:8-9.
Jayatilaka, C., B. Storm, andL. Mudgway. 1998. Simulation of water flow on irrigation bay scale with MIKE-SHE. J. Hydrology,
    208(1-2): 108-130.
Refsgaard, J.C.. 1997. Paramterization, calibration and validation of distributed hydrological models, Journal of Hydroloigy,
    198(l-4):69-97.
Refsgaard, J.C., and J. Knudsen. 1996. Operational validation and intercomparison of different types of hydrologic models, Water
    Resources Research, 32(7):2189-2202.
Refsgaard, J., andB. Storm. 1995. Mike  She. Chap 23, Computer Models of Watershed Hydrology, V. Singh, Ed., 809-846.
    Highland Ranch, CO, Water Resources Publications.
Xevi, E, K. Christiaens, A. Espino, W. Sewnandan, D. Mallants, H. Sorensen, J. Feyen. 1997. Calibration, validation and
    sensitivity analysis of the MIKE-SHE model using the Neuenkirchen catchment as case study. Water Resources
    Management, ll(3):219-242.

Also visit http://www.dhi.dk/ContactUs/Library for additional all DHI compendium of technical papers and publications.


OPUS: Opus is a continuous field-scale (unit area) root-zone model, developed as a  research and management tool to assist in
agricultural nonpoint source pollution control. Hydrology, erosion, nutrient, pesticide,  and crop growth components are included.
Runoff/infiltration is partitioned using either a daily hydrology option (curve number) or infiltration equation using break-point
rainfall. Unsaturated flow is modeled  with Richards'  equation. Evapotranspiration  is  computed from air temperature, solar
radiation,  soil-water, and crop stage.  The crop growth component  considers  radiation, nutrients,  temperature,  and water
availability. Carbon, nitrogen, and phosphorus processes are represented in the soil-water-plant dynamics. Pesticides are modeled
assuming  equilibrium or kinetic adsorption, first-order decay, and advective transport. If daily  runoff option is used erosion is
estimated based on the Modified Universal Loss Equation  (MUSLE). A more detailed,  spatially and temporally distributed
approach that considers particle size classes  is used with the infiltration equation. OPUS  considers variation in vertical direction
(soil column), but assumes uniform soil, crop and climate characteristics. Fields with divided flow, and features such as terraces,
contours, furrows, grassed buffer-strips  or waterway, and farm ponds can be simulated. Model documentation is published, and
the model is distributed free. Model and the manual is available through the National Technical Information Service, 5285 Port
Royal Road, Springfield, VA 22161
BMPS: terraces, contours, furrows, grassed buffer-strips or waterway, and farm ponds.

Application and Model References:
Arenstein, D.J., S.R. Workman, and S.E. Nokes.  1995. Calibration and validation of the Opus model at the Ohio Management
    Systems Evaluation Area. Paper 95-2403. St. Joseph, Mich.: ASAE.
Ferreira, V. A., and R.E. Smith. 1992. Opus, an integrated simulation model for transport of nonpoint source pollutants at the field
    scale: Volume II, User Manual. ARS-98. Washington: USDA Agricultural Research Service. 200 pp.
Heatwole, C.D., S. Zacharias, and N. Persaud. 1997. Comparison of Opus and GLEAMS in simulating spatial variability of
    pesticide movement in a field soil. In: Application of GIS, Remote Sensing, Geostatistics, and Solute Transport Modeling.
    Washington: Amer. Geophysical Union.
Ma, Q.L., R.D. Wauchope, J.E. Hook, A.W. Johnson, C.C. Truman, C.C. Dowler, G.J.  Gascho, J.G. Davis, H.R. Summer, and
    L.D. Chandler. 1998. GLEAMS, Opus, and PRZM-2 model predicted versus measured runoff from a coastal plain loamy
    sand. Transactions of the ASAE 41(l):77-88.
Pierson, F.B., G.N. Flerchinger, and J.R. Wight.  1992. Simulating near-surface soil temperature and water on sagebrush
    rangelands: A comparison of models. Transactions of the  ASAE 35(5):1449-1455.
Ramanarayanan, T.S., G.J. Sabbagh, M.R. Reyes, R.L. Bengston, D.E. Storm, and J.L.  Fouss. 1994. Performance of transport
    models in predicting nitrate runoff from high water table areas. Paper 94-2152.  St.  Joseph, Mich.: ASAE. 13 pp.
Santos, D. V., R.E. Smith, P.L. Sousa, and L. S. Pereira. 1996. Calibration and validation of model Opus for water and nitrate
    simulation. In: R.Ragab, D.E. El-Quosy, B. vanDenBoek, andL.S. Pereira, eds, Crop Water Environment Models,
    proceedings of the Cairo Workshop, ICID, Cairo, pp. 17-28.
Santos, D.V., P.L. Sousa, and R.E. Smith. 1997. Model simulation of water and nitrate  movement in a level-basin under
    fertigation treatments. Agricultural Water Management 32:293-306.
                                                         57

-------
Smith, R.E. 1992. Opus, an integrated simulation model for transport of nonpoint source pollutants at the field scale: Volume I,
    Documentation. ARS-98. Washington: USDA Agricultural Research Service. 120 pp.
Smith, R.E. 1993a. Simulation of crop water balance with Opus. In: L.S. Pereira, BJ. van denBroek, P.Kabat, and R.G. Allen,
    editors, Crop-Water-Simulation Models in Practice. Selected papers, 15th ICID Congress, The Hague, pp.215-227.
Smith, R.E. 1993b. Simulation experiments  on the role of soil hydraulic characteristics in Agro-Ecosystems. Modeling of Geo-
    Biosphere Processes 2(1/4): 1-14.
Smith, R.E. 1995. Opus simulation of a wheat/sugarbeet plot near Neuenkirchen, Germany. Ecological Modeling 81:121-132.
Smith, R.E., and B. Diekkruger. 1992. Field-scale soil water flow in heterogeneous soils, I, Modeling statistical soil variation and
    large-scale constituent relations. Modeling of Geo-Biosphere Processes 1:205-227.
Smith, R.E., and V.A. Ferreira.  1989. Comparative evaluation of unsaturated flow methods in selected USDA simulation models.
    In: H.J. Morel-Seytoux (ed), Unsaturated Flow in Hydrologic Modeling Theory and Practice. Kluwer Academic Publishers.
    pp.391-412.
Zacharias, S. and C.D. Heatwole. 1993. Predicting tillage treatment effects on pesticide transport: A validation study. Paper 93-
    2592. St. Joseph, Mich.:  ASAE.
Zacharias, S., and C.D. Heatwole. 1996. A stochastic framework for incorporating spatial variability in NPS models. Paper 96-
    2028. St. Joseph, Mich.:  ASAE.
Zacharias, S., and C.D. Heatwole. 1997. Stochastic simulation of root zone water and solute movement in an agricultural field.
    Paper 97-2001. St. Joseph,  Mich.: ASAE.


PRMS (Precipitation-Runoff Modeling System): PRMS is  a distributed watershed model that simulates  precipitation- and
snowmelt-driven movement of water through the basin via overland flow, interflow, and baseflow. Watershed response can be
simulated at a daily time step or more frequently over the course of a storm. Kinematic routing of the unidirectional flow and the
transport of sediments through  a receiving network of well-mixed channel reaches can be simulated when the model is in "storm
mode". Simulation of the energy balance in the snowpack and the water balance is based on many theoretically- and empirically-
developed relations. The resulting model is comprehensive and flexible, but also very complex and requires a large  number of
parameters. The model contains procedures for parameter optimization and sensitivity analyses. A Unix-based GUI is available
through the modeling framework MMS. Watershed is divided into subunits based on such basin characteristics as slope, aspect,
elevation, vegetation type, soil type, land use, and precipitation distribution. Two levels of partitioning are  available. The first
divides the basin into homogeneous response units (HRU) based on the basin characteristics. Water and energy balances are
computed daily for each HRU.  The sum of the responses of all HRU's, weighted on a unit-area basis, produces the daily system
response and  streamflow for a basin. A second level of partitioning is available for storm hydrograph simulation. The watershed
is conceptualized as a series of interconnected flow  planes and channel segments. Surface runoff is routed over the flow planes
into the  channel  segments;  channel flow is routed  through the watershed channel system. An HRU can be  considered the
equivalent of a flow  plane or it can be delineated into a number of flow planes. The source of code of RPMS is available to
public. It is written in Fortran 77, and therefore can be considered platform independent.
URL: http://smig.usgs.gov/cgi-bin/SMIC/model  home_pages/modelhome?selection=prms
    http://water.usgs.gov/software/prms.html

Application and Model References:
Carey, W.P., and A. Simon. 1984. Physical basis and potential estimation techniques for soil erosion parameters in the
    Precipitation-Runoff Modeling System  (PRMS). U.S. Geological Survey Water-Resources Investigations Report 82-4218, 32
    P-
Gary, L.E. 1984. Application of the U.S. Geological Survey's Precipitation-Runoff Modeling System to the Prairie Dog Creek
    basin, Southeastern Montana. U.S. Geological Survey Water-Resources Investigations Report 84-4178, 98 p.
Kidd, R.E., and C.R. Bossong. 1987. Application of the precipitation-runoff model in the Warrior Coal Field, Alabama. U.S.
    Geological Survey Water-Supply Paper 2036, 42 p.
Kuhn, G. 1989. Application of the U.S. Geological Survey's Precipitation-Runoff Modeling System to Williams Draw and Bush
    Draw basins,  Jackson County, Colorado. U.S. Geological Survey Water-Resources Investigations Report 88-4013,  38 p.
Norris, J.M., and R.S. Parker. 1985. Calibration procedure for a daily flow model of small watersheds with snowmelt runoff in
    the Green River coal region of Colorado. U.S. Geological Survey Water-Resources Investigations Report 83-4263, 32 p.
Parker, R.S., and J.M. Norris. 1989. Simulation of streamflow in small drainage basins in the southern Yampa River Basin,
    Colorado. U.S. Geological Survey Water-Resources Investigations Report 88-4071, 47 p.
Puente, C., and J.T. Atkins. 1989. Simulation of rainfall-runoff response in mined and unmined watersheds in coal areas of West
    Virginia. U.S. Geological Survey Water-Supply Paper 2298, 48 p.
                                                         58

-------
Scott, A.G. 1984. Analysis of characteristics of simulated flows from small surface-mined and undisturbed Appalachian
    watersheds in the Tug Fork basin of Kentucky, Virginia, and West Virginia. U.S. Geological Survey Water-Resources
    Investigations Report 84-4151, 169 p.


'REMM (Riparian Ecosytem Management Model):
REMM is a tool for estimating the nonpoint source pollution control by field-scale riparian ecosystems. It can be used to simulate
hydrology, nutrient dynamics and plant growth for land areas between the edge of fields and a waterbody. Management options
such as vegetation type, size of the buffer zone, and biomass harvesting can also be simulated. A riparian buffer system is divided
into three zones i) Zone 1 is permanent woody vegetation immediately  adjacent to the stream bank., ii) Zone 2 is managed forest
occupying a strip upslope from zone 1, iii) Zone 3 is an herbaceous strip upslope from zone 2. The primary purposes of zone 3 are
to remove sediment from surface runoff and to convert channelized flow to sheet flow. The primary function of zone 2 is to block
transport of sediment and chemicals from upland areas into the adjacent wetland or aquatic system. The purpose of Zone 1 is to
maintain the integrity of the stream bank and a favorable habitat for aquatic organisms. Movement and storage of water within
riparian buffer systems is simulated by a process-based, two-dimensional water balance operating on a daily time step. Surface
runoff is assumed to be generated by infiltration excess and saturation excess. Infiltration is estimated using an explicit form of
modified Green-Ampt equation. A very simple surface runoff routing scheme is used which is based on the time of concentration
concept. Only incoming runoff is routed. Runoff generated within the riparian area by infiltration excess and saturation excess is
not subject to routing. Upward flux from a shallow water table is computed using Dary-Buckingam equation. Sediment transport
is simulated both in channels and overland flow areas, but channel erosion or detachment is not simulated. Channel shapes are
assumed triangular. Lateral  subsurface movement is modeled with Darcy's equation. Because  of the roughness of the riparian
buffers, it is assumed that sediment transport is primarily of suspended particles. Upland loadings are assumed to be provided as
input to the REMM. Overland flow erosion is based on the USLE equation. Five classes of sediment are considered: sand,  large
aggregate,  small aggregate, silt and  clay. Sediment load  computations  are performed for  each of these classes. Steady state
continuity equation is used to compute the sediment at the downslope edge.
BMPs: Agricultural practices, riparian buffers
URL: http://sacs.cpes.peachnet.edu/remmwww

Application and Model References:
Altier, L.S., R.G. Williams, R. Lowrance, and S.P. Inamdar. 1998. The Riparian Ecosystem Management Model: Plant growth
    component. Proceedings of the First Federal Interagency Hydrologic Modeling Conference, Las Vegas, NV,  April 1998, Pgs:
    1.33-1.40.
Bosch, D.D., R.G. Williams, S.P. Inamdar, J.M. Sheridan, and R. Lowrance. 1998. Erosion and sediment transport through
    riparian forest buffers. Proceedings of the First Federal Interagency Hydrologic Modeling Conference, Las Vegas, NV, April
    1998, Pgs: 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. Proceedings of the First Federal Interagency Hydrologic Modeling Conference, Las Vegas, NV, April
    1998, Pgs: 1.73-1.80.
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. Proceedings of the First Federal Interagency Hydrologic
    Modeling Conference, Las Vegas, NV, April  1998, Pgs: 7.17-7.24.
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. Proceedings of the First
    Federal Interagency Hydrologic Modeling Conference, Las Vegas, NV, April 1998, Pgs:  1.81-1.88.
Williams, R.G., R. Lowrance, L.S. Altier, and S.P. Inamdar. 1998. The Riparian Ecosystem Management Model: A
    demonstration. Proceedings of the First Federal Interagency Hydrologic Modeling Conference, Las Vegas, NV, April 1998,
    Pgs: 8.133-8.138.


SWAT (Soil Water Assessment Tool): SWAT is a conceptual, continuous time model and is more suitable for large river basins.
The SWAT model emerged from the models SWRBB, CREAMS, GLEAMS, EPIC and ROTO. It operates on daily time step.
The watershed is divided into sub-basins and each sub-basin is further partitioned into Hydrologic Response Units (HRU) having
uniform topographic, soil and land use properties. Input information for each subbasin is grouped or organized into the following
categories: weather or climate; unique areas of land cover, soil, and management within the  subbasin (hydrologic response units
or HRUs); ponds/reservoirs; groundwater; and the main channel, or reach, draining the subbasin. In SWAT water balance is the
driving force behind everything that happens in the watershed. Simulated hydrologic processes are surface runoff with SCS curve
number or Green-Ampt infiltration, lateral subsurface flow, groundwater flow, evapotranspiration, snowmelt, transmission losses
                                                        59

-------
from streams and water storage and losses from ponds. Flow is routed through the channel using a variable storage coefficient
method. Sediment yield is computed from MUSLE for each sub-basin. The transport of sediment in the channel is controlled by
the simultaneous operation of two processes, deposition and degradation. Deposition in the channel is based on sediment particle
fall velocity calculated  with  Stake's Law.  Stream power is used to predict degradation in the routing reaches. An Arc View
interface is available which enables extraction of input parameters easily, and visualization of results. SWAT is integrated into the
USEPA's BASINS  and  USDA's AGWA systems. It is also linked to the river and stream water quality  model QUAL2E. Some
applications   of   SWAT    and   projects    in   which   the   model   has   been   used   are    summarized   on
http://www.brc.tamus.edu/swat/swatapp.html
BMPs: Agricultural practices, ponds, tile drains.
URL: http://www.brc.tamus.edu/swat

Application and Model References:
Arnold, J.G. and P.M. Allen. 1992. A Comprehensive surface-groundwater flow model. J. Hydrol. 142:47-69.
Arnold, J.G. and P.M. Allen. 1999. Automated Methods for Estimating Baseflow and Groundwater Recharge from Streamflow
    Records. JAWRA, 34(2):411-424.
Arnold, J.G., R. Srinivasin, R.S. Muttiah, and J. R. Williams. 1998. Large Area Hydrologic Modeling and Assessment: Part I.
    Model Development. JAWRA 34(l):73-89.
Arnold, J.G., R. Srinivasan, R.S. Muttiah, and P.M. Allen. 1999. Continental Scale Simulation of the Hydrologic Balance.
    JAWRA 35(5):1037-1051.
Arnold, J.G., Williams, J.R., and Maidment D.A. 1992. Continuous-Time Water and Sediment-Routing Model for Large Basins.
    Journal of Hydraulic Engineering, Vol 121. No. 2., February, 1995, ASCE. Pgs.  171-183.
Srinivasan, R. and J.G. Arnold. 1994. Integration of a Basin-Scale Water Quality Model with GIS. Water Resources Bulletin.
    Vol. 30, No. 3., June 1994. Pgs. 453-462.
Srinivasan, R., J.G. Arnold, R.S. Muttiah, and P.T. Dyke. 1995. Plant and Hydrologic Simulation for the Conterminous U.S.
    Using SWAT and GIS. Hyd Sci &Tech, Vol.  11, No 1-4, Amer. Inst of Hyd., Pg 160-168.


SWMM (Storm Water Management Model):  SWMM is a comprehensive computer model for analysis of quantity and quality
problems  associated with urban runoff.  Both  single-event and continuous simulation can be performed on catchments having
storm sewers, or combined sewers and natural  drainage, for prediction of flows,  stages and pollutant  concentrations.  It is
structured in the  form of blocks. The principal computational blocks  include the  Runoff Block for generation of runoff and
quality constituents from rainfall (plus simple  routing of flow and quality), the  Transport Block for kinematic wave routing and
for additional dry-weather flow and quality routing, the Storage/Treatment Block for reservoir routing and simulation of treatment
and storage quality processes, and the Extended  Transport or Extran Block for hydraulic  routing of flow (no  quality routing)
using the  complete Saint-Venant  equations. Using SWMM, the modeler can simulate  all aspects  of the urban hydrologic and
quality cycles, including rainfall,  snowmelt, surface and subsurface runoff, flow routing through the drainage network,  storage
and treatment. The Rain Block is used for processing of hourly and 15-minute precipitation time series for input to continuous
simulation. Although the historical basis of the model was for analysis of urban runoff quality problems, the model often is used
just for hydrologic and hydraulic analysis. The  model is designed for use by engineers  and scientists experienced in urban
hydrological and water quality processes. An engineering background is necessary to appreciate most methods being used and to
verify that the model results are reasonable. SWMM Version 4 is microcomputer based (DOS-compatible), although the Fortran
code may be compiled on  any machine. For hydrologic simulation  in  the  Runoff Block, data requirements include  area,
imperviousness, slope, roughness, width (a shape  factor), depression storage, and infiltration parameters  for either the Horton or
Green-Ampt  equations  for up  to  100  subcatchments.  (Number  of subcatchments,  pipes, etc. is variable depending on the
compilation). Flow  routing can be performed in the Runoff, Transport and Extran Blocks, in increasing order of sophistication.
Extran can also simulate dynamic boundary conditions, e.g., tides. Quality processes are initiated in the Runoff Block and include
options for constant concentration, regression of load vs. flow, and buildup washoff,  with the latter requiring the most  data.
Additional options include street cleaning, erosion, and quality contributions from precipitation, catchbasins, adsorption, and base
flow. EPA Nationwide Urban Runoff Program data are often used as starting values for quality computations. SWMM interfacing
requirements are clearly defined. E.g., output  may be directed to the EPA WASP receiving water model. Basic SWMM output
consists of hydrographs and pollutographs (concentration  vs. time) at any desired location in the drainage system. Depths and
velocities  are also  available  as are  summary statistics on surcharging,  volumes, continuity and other quantity parameters.
Additional quality output includes loads, source  identification, continuity, residuals  (e.g., sludge), and other parameters. GIS
linkage is available. The model performs best in urbanized areas  with impervious drainage, although it has been widely  used
elsewhere. Technical limitations include lack of subsurface quality routing (a constant concentration is  used), no interaction of
quality processes  (apart  from adsorption), difficulty in simulation of wetlands quality processes (except as can be represented as
storage processes), and a weak  scour deposition routine in the Transport Block. The  biggest impediment to model usage is the
                                                         60

-------
user interface, with its lack of menus and graphical output. The model is still run in a batch mode (the user constructs an input file
with an editor), unless third-party software is used for pre- and post-processing. It has been used in scores of U.S. cities as well as
extensively in Canada, Europe, Australia and elsewhere.  Source  code, executable version and the models manuals can be
downloaded freely from
URL:  http://www.cee.odu.edu/model/swmm.php
BMPs: Detention basins, street cleaning.

Application and Model References:
Curtis, T.G., and W.C. Huber. 1993. SWMM AML - An ARC/INFO Processor for the Storm Water Management Model
    (SWMM). Proc. 1993 Runoff Quantity and Quality Modeling Conference, Reno, NV, (NTIS, in press), U.S. EPA, Athens,
    GA, 30605.
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. EPA,  Athens, GA, 30605.
Huber, W.C. 1986. Deterministic Modeling of Urban Runoff Quality. In: H.C.Torno et. al. (eds.) Urban Runoff Pollution,
    Proceedings of the NATO Advanced Research Workshop on Urban Runoff Pollution, Montpellier, France. Springer-Verlag,
    New York, Series G: Ecological Sciences, 10:167-242.
Huber, W.C. 1992. Experience with the U.S. EPA SWMM Model for Analysis and Solution of Urban Drainage Problems.
    Proceedings, Inundaciones Y Redes De Drenaje Urbano, J. Dolz, M. Gomez, and J.P. Martin, eds., Colegio de Ingenieros de
    Caminos, Canales Y Puertos, Universitat Politecnica de  Catalunya, Barcelona, Spain, p. 199-220.
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. EPA, Athens, GA, 30605.
Huber, W.C., Heaney, J.P. and B. A. Cunningham. 1985. Storm Water Management Model (SWMM) Bibliography. EPA/600/3-
    85/077 (NTIS PB86-136041/AS), U.S. EPA, Athens, GA, September 1985.
Huber, W.C., Zollo, A.F., Tarbox, T.W. and J.P. Heaney. 1991. Integration of the SWMM Runoff Block with ARC/INFO and
    AutoCAD: A Case Study. Final Report to Foster-Wheeler Enviresponse, Inc. and U.S. EPA, Edison, NJ, Contract VN1-320-
    420000, fromDept. of Environmental Engineering Sciences, University of Florida, Gainesville.
Martin, J.L. 1993. Modification of the Storm Water Management Model's (SWMM's) Transport Submodel for Creation of a
    Hydrodynamic Linkage to the Water Analysis Simulation Program (WASP). Report to Camp, Dresser and McKee, Inc. by
    AScI Corp., Athens, GA, 30605.
Roesner, L.A., Aldrich, J.A. and R.E. Dickinson.  1988. Storm Water Management Model, Version 4, User's Manual: Extran
    Addendum. EPA/600/3-88/00Ib (NTIS PB88-236658/AS), U.S. EPA, Athens, GA, 30605.


 VFSMOD (Vegetative Filter Strips hydrology and sediment transport MODel): VFSMOD is a field scale, mechanistic, storm-
based model designed to route  the  incoming hydrograph and sedimentograph from an adjacent field through a vegetative  filter
strip (VFS)  and  to calculate the outflow,  infiltration and sediment trapping efficiency. The  model handles  time dependent
hyetographs, space distributed filter parameters (vegetation  roughness or density, slope, infiltration characteristics) and different
particle size of the incoming  sediment.  Any combination of unsteady storm and incoming hydrograph types can be used.
VFSMOD consists of a series of modules simulating the behavior of water and sediment in the surface of the VFS: i) Green-Ampt
infiltration module: a module for calculating the water balance in the soil surface; ii) kinematic wave overland flow module: a 1-
D module for calculating flow depth and rates on the infiltrating soil surface; iii) sediment filtration module: a module for
simulating transport and deposition of the incoming sediment along the VFS. The model can be  used to describe transport at the
field scale (or field edge) if flow  and transport is mainly  in the form of sheet flow (Hortonian) and  the 1-D path represents
average conditions (field effective  values) across the  VFS.  A windows version of the model called VFSMOD-W has recently
been developed. The model is provided free of charge as an educational and research tool. The model and documentation can be
downloaded from the internet. No  formal training is available.  Limited support is available from the authors. Through the web
site, the user can send feedback and questions to the authors.
URL: http://www3.bae.ncsu.edu/vfsmod/
BMPs: Vegetative filter strips.

Application and Model References:
Munoz-Carpena, R., J. E. Parsons and J. W. Gilliam. 1993b.  Numerical approach to the overland flow process in vegetative filter
    strips. Transactions of ASAE. 36(3):761-770.
Munoz-Carpena, R., J. E. Parsons and J. W. Gilliam. 1999. Modeling hydrology and sediment transport in vegetative filter strips
    and riparian areas. J. of Hydrology 214(1-4):111-129.
                                                        61

-------
Munoz-Carpena, R. and J. E. Parsons. 1999. Evaluation of VFSMOD, a vegetative filter strip hydrology and sediment filtration
    model. 1999 ASAE/CSAE-SCGR Annual International Meeting, Toronto, Ontario, Canada. July 18-22, 1999. ASAE Paper
    No. 992152.
Parsons, J.E. and R. Munoz-Carpena. 2001. Impact of uncertainty on the design of vegetative filter strips. ASAE Annual
    International Meeting, Sacramento, California. July 29-Aug. 1, 2001. ASAE Paper No. 012214.


WEPP (Water Erosion Prediction Project):  The Water Erosion Prediction Project (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) is applicable to hillslope erosion processes (sheet and rill erosion), as well as simulation
of the hydrologic  and erosion processes on small watersheds (<640 Acres). Processes considered in  hillslope profile model
applications include  rill and interrill erosion, sediment transport and deposition, infiltration,  soil  consolidation, residue and
canopy effects on soil detachment and infiltration, surface sealing, rill hydraulics, surface runoff, plant growth, residue
decomposition, percolation, evaporation, transpiration, snow melt, frozen soil effects on infiltration and credibility,  climate,
tillage effects on soil properties, effects of soil random roughness, and contour effects including potential overtopping of contour
ridges. The model accommodates the spatial and temporal variability in topography, surface roughness, soil properties, crops, and
land use conditions on hillslopes. In watershed  applications,  the model  allows linkage of hillslope profiles to  channels and
impoundments. Water and sediment from one or more hillslopes can be routed through a small field scale watershed. Almost all
of the parameter updating for hillslopes is duplicated for channels. The model simulates channel detachment, sediment transport
and deposition. Impoundments such as farm ponds, terraces, culverts, filter fences and check dams can be simulated to remove
sediment from the flow. The procedures do not consider classical gully erosion. Also, model application is limited to areas where
the hydrology is dominated by Hortonian overland flow. The infiltration component of the hillslope model is based on a modified
Green-Ampt equation. Overland flow routing procedures include both an analytical solution to the kinematic wave equations and
an approximate method. Soil erosion is represented in two ways for WEPP overland flow profile applications: i) soil particle
detachment by raindrop impact and transport by sheet  flow on interrill areas  (interrill delivery  rate), and ii)  soil particle
detachment, transport and deposition by concentrated flow in rill areas (rill erosion). Flow depth and hydraulic shear stress along
the channel are computed by regression equations based on a numerical  solution of the steady-state spatially-varied  flow
equation. Detachment, transport, and deposition of sediment are calculated by a steady-state solution to the sediment continuity
equation. Impoundment component outputs include: i) peak outflow rate and volume leaving the impoundment;  ii) peak sediment
concentration and the total sediment yield leaving the impoundment for the five particle size classes; and iii) the median particle
size diameter  of the  sediment  leaving the  impoundment for the five particle size classes.  WEPP has a weather generator
(CLIGEN) which generates mean daily precipitation, daily maximum and minimum temperature, mean daily solar radiation, and
mean daily wind direction and speed using two-sate Markov Chain model.
BMPs: Agricultural practices, ponds, terraces, culverts, filter fences, check dams.
URL: http://topsoil.nserl.purdue.edu/nserlweb/weppmain/wepp.html

Application and Model References:
Cochrane, T.A. and D.C. Flanagan. 1999. Assessing water erosion in small watersheds using WEPP with GIS and digital
    elevation models: J. Soil and Wat. Conserv., 54(4):678-685.
Elliot, W.J., W. Qiong and A.V. Elliot. 1993. Application of the WEPP model to surface mine reclamation. Paper presented at
    Challenge of Integrating Diverse Perspectives in Reclamation, 10th National Meeting. Spokane, WA: Am. Soc. Surface Mine
    Reclam.
Flanagan, D.C and S.J. Livingston (eds.).  1995. USDA-Water Erosion Prediction Project: WEPP User Summary. NSERL Report
    No. 11. USDA-ARS National Soil Erosion Research Laboratory, West Lafayette, Indiana.
Flanagan, D.C. and M.A. Nearing (eds.). 1995.  USDA-Water Erosion Prediction Project: Hillslope Profile and Watershed Model
    Documentation. NSERL Report No. 10, USDA-ARS National Soil Erosion Research Laboratory, West Lafayette, Indiana.
Flanagan, D.C. and M.A. Nearing. 2000. Sediment particle sorting on hillslope profiles in the WEPP model: Trans. Am. Soc.
    Agric. Eng., 43(3):573-583.
Nearing, M.A., L.A. Deer-Ascough, and J.M. Laflen. 1990. Sensitivity analysis of the WEPP hillslope profile erosion model.
    Trans. Am. Soc. Agric. Eng. 33(3):839-849.
Nearing, M.A., G.R. Foster, L.J. Lane, and S.C. Finkner. 1989. A process-based soil erosion model for USD A-Water Erosion
    Prediction Project (WEPP) technology. Trans. Am. Soc. of Agric. Eng. 32(5):1587-1593.
Nearing, M.A. and A.D. Nicks. 1998. Evaluation of the Water Erosion Prediction Project (WEPP) model for hillslopes: in
    Modelling Soil Erosion by Water (J. Boardman and D.T. Favis-Mortlock, eds.), Springer-Verlag NATO-ASI Series 1-55,
    Berlin: 45-56.
Savabi, M.R., D.C. Flanagan, B. Hebel, B.A. Engel. 1995. Application of WEPP and CIS-GRASS to a Small Watershed in
    Indiana. Journal of Soil and Water Conservation 50(5):477-483.
                                                         62

-------
WMS (Watershed Modeling System): The WMS software provides a comprehensive environment for hydrologic analysis of
watershed systems. Developed in cooperation with the Waterways Experiment Station (WES), WMS provides graphical tools for
use in the delineation of watersheds and flood plains. Hydrologic models may be set up and viewed in a user-friendly graphical
environment. The WMS software package is divided logically into six well-integrated, task-oriented modules. These modules are:
Triangulated Irregular Networks (TINs), DEMs, Tree, Grid, Scatter Point, and Map (GIS). The WMS software package provides
an interface to HEC-1, TR-20, Rational Method, National Flood Frequency (NFF), GSSHA, and HSPF. The interface to last two
models is still a beta version. WMS can be operated under UNIX or WINDOWS operating systems.
URL\ http://chl.wes.army.mil/software/wms. http://www.ems-i.com/WMS/wms.html
8.2    Receiving Water Models:


CE-QUAL-ICM & CE-QUAL-ICM/TOXI: 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. Subsequent to employment in the Bay study,
the model code was generalized and minor corrections and improvements were installed. 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 model computes  and reports concentrations, mass
transport, kinetics transformations, and  mass balances. Features to aid debugging include the ability to activate or deactivate
model features, diagnostic output, and volumetric and mass balances. Computations can be restarted following interruption due to
computer failure  or similar circumstances. CE-QUAL-ICM is coded in ANSI Standard FORTRAN F77. The model operates on a
variety of platforms including 486 PC,  Silicon Graphics, and Hewlett Packard workstations. A multi-processor version is
available but not generally released. The user must provide processors that prepare input files and process  output for presentation.
The model does  not compute hydrodynamics. Flows, diffusion coefficients, and volumes must be specified externally and read
into the model. For simple configurations, flows may be entered through an ASCII input file. For more advanced applications,
hydrodynamics are  usually obtained from a hydrodynamics model such as the  CH3D-WES model. The  unstructured,  finite
volume structure of the model was selected to facilitate linkage to a variety of hydrodynamic models. There are two distinctly
different development pathways to ICM: a eutrophication model (ICM), and an organic chemical model  (ICM/TOXI). 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 sub-model.
The sub-model interactively predicts sediment-water oxygen and nutrient fluxes. Alternatively,  these fluxes may be specified
based on observations. The 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
ICM/TOXI  model resulted from  incorporating  the toxic chemical routines from EPA's WASP (Water 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;  and  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. WASP toxic  chemical model upon  which ICM/TOXI is based has been applied to  a wide
variety of sites. CE-QUAL-ICM also has been linked to EFDC hydrodynamic model.
URL: http://www.wes.army.mil/el/elmodels

Application and Model 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.


                                                         63

-------
DiToro, D.M., and J.F. Fitzpatrick. 1993. Chesapeake Bay sediment flux model. Prepare by Hydroqual, Inc. for US. EPA
    Chesapeake Bay Program, US. Army Engineers District, Baltimore, MD, and US. Army Engineer Waterways Exp. Station.
    Contact Report EL-93-2, 200 pp.
Mark, D., B. Bunch, and N. Scheffner. 1992. Combined hydrodynamic and water quality modeling of Lower Green Bay. Water
    Quality '92: Proceedings of the 9th Seminar, pp 226-233. Miscellaneous Paper W-92-3, Environmental Laboratory, Army
    Engineers Waterways Experiment Station, Vicksburg, MS.


CE-QUAL-R1: CE-QUAL-R1 is spatially one dimensional and horizontally averaged reservoir water quality model. Temperature
and concentration gradients are computed only in the vertical  direction. The reservoir is conceptualized as a vertical sequence of
horizontal layers where thermal energy and materials are uniformly distributed in each layer. The mathematical structure of the
model is based on horizontal layers whose thicknesses depend on the balance of inflowing and outflowing waters. Variable layer
thicknesses permit accurate mass balancing during periods of inflow and outflow. The distribution of inflowing waters among the
horizontal layers is based on density differences. Simulations of surface flows, interflows, and underflows are possible. Similarly,
outflowing waters are withdrawn from layers after considering layer densities, discharge rates, and outlet configuration. Reservoir
outflows may take place according to a specified schedule of  port releases. Alternately, specification of total release and desired
release temperatures can be  made. In this case, the  model will select port flows. In addition, both continuous (normal) and
scheduled operations can be simulated. Continuous operation refers to normally uninterrupted port and weir outflows. Scheduled
operation refers to fluctuating generation outflows or pumpback  inflows. Vertical transport of thermal energy  and materials
occurs through entrainment and turbulent diffusion. Entrainment is a transport process that sharpens gradients and determines the
depth of the upper mixed region and the onset of stratification. It is calculated from the turbulent kinetic energy influx generated
by wind shear and convective mixing. Turbulent diffusion is a transport process that reduces gradients and is calculated using a
turbulent diffusion coefficient that is dependent on wind speed, inflow and outflow magnitudes, and density stratification. The
interaction of numerous biological and chemical factors is a major  attribute of CE-QUAL-R1. The model simulates interactions
of physical factors (such as  flow and temperature),  chemical factors (such as nutrients), and biological assemblages in both
aerobic and anaerobic environments. It can perform stochastic simulations using Monte Carlo methods. Statistical data describing
biological and chemical coefficients are used to provide probabilistic estimates of key output variables. The thermal analysis
portion  of CE-QUAL-R1  is provided as an independent model (CE-THERM-R1) to simplify simulation of water budgets and
temperature profiles. CE-THERM-R1 includes  the  variables of  temperature, suspended  solids,  and  total dissolved solids.
Algorithms representing physical processes are the same as in CE-QUAL-R1. A number of utilities are also provided with CE-
QUAL-R1. These include preprocessors, which are  aids in  assembling a usable data set, two graphic utilities, statistics for
comparing measured and predicted data, and a flux model. The  flux model  calculates and lists  the rates  of  change for all
biological processes, which should aid the users of CE-QUAL-R1 to correctly predict variable concentrations.  An  interactive
windows package (WESWIN) is available which enables the execution of CE-QUAL-R1 and the utilities associated with it. This
interface also  has a plotting program which makes model calibration  easier by letting the user view the  model results
immediately.
URL: http://www.wes.army.mil/el/elmodels

Application and Model References:
Chen, R. L., Brannon, J. M., and Gunnison, D. 1984. Anaerobic and aerobic rate coefficients for use in CE-QUAL-R1.
    Miscellaneous Paper E-84-5, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS. NTIS No. AD A145 499.
Collins, C. D., and Wlosinski, J. H. 1983. Coefficients for use in the U.S. Army Corps of Engineers reservoir model, CE-QUAL-
    Rl. Technical Report E-83-15, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS. NTIS No. AD A135
    733.
Environmental Laboratory. 1986. CE-QUAL-R1: A Numerical One- Dimensional Model of Reservoir Water Quality; User's
    Manual. Instruction Report E-82-1 (Revised Edition), US Army Engineer Waterways Experiment Station, Vicksburg, Miss.
U.S. Army Engineer Waterways Experiment Station. 1982. CE-QUAL-R1: A numerical one-dimensional model of reservoir
    water quality; User's manual.  Instruction Report E-82-1, Vicksburg, MS. NTIS No. AD A116 538.
U.S. Army Engineer Waterways Experiment Station. 1995. CE-QUAL-R1: A numerical one-dimensional model of reservoir
    water quality; User's manual.  Instruction Report E-82-1, Vicksburg, MS.
Wlosinski, J. H. 1984. Evaluation techniques for CE-QUAL-R1: A one-dimensional reservoir quality model. Miscellaneous Paper
    E-84-1, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS. NTIS No. AD A140 766.
Wlosinski, J. H., and Collins, C. D. 1985. Confirmation of the water quality model CE-QUAL-R1 using Data from Eau Galle
    Reservoir, Wisconsin. Technical Report E-85-11, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
    NTIS No.  AD A164 226.
                                                        64

-------
CE-QUAL-RIV1: CE-QUAL-RIV1  is a one-dimensional hydrodynamic  and water quality model, meaning that the model
resolves longitudinal variations in hydraulic and quality characteristics and is applicable where lateral and vertical variations are
small.  CE-QUAL-RIV1 consists of two parts, a hydrodynamic code  (RIV1H)  and a water  quality code (RIV1Q). The
hydrodynamic code is applied first to predict water transport and its results are written to a file, which is then read by the quality
model. It can be used to predict one-dimensional hydraulic and water quality variations in streams and rivers with highly unsteady
flows, although it can also be used for prediction under steady flow conditions. RIV1H predicts flows, depths, velocities, water
surface elevations, and other hydraulic characteristics. The hydrodynamic model solves the St. Venant equations as the governing
flow equations using the widely accepted four-point implicit finite difference numerical scheme. 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.  Numerical accuracy for the
advection of sharp gradients is preserved in the water quality code through the use of the explicit two-point, fourth-order accurate,
Holly-Preissman scheme.
URL: http://www.wes.army.mil/el/elmodels

Application and Model References:
Curtis, L. T., J.M. Nestler, and J.L. Martin. 1987. Comparative effects on trout habitat of hydropower modification with and
    without reregulation in the Cumberland River below Wolf Creek Dam, Kentucky. Miscellaneous Paper EL-87-2, U.S. Army
    Engineer Waterways Experiment Station, Vicksburg, MS. NTIS No. AD A179 787.
Environmental Laboratory. 1985. CE-QUAL-RIV1: A Dynamic, One-Dimensional (Longitudinal) Water Quality Model for
    Streams. User's Manual," Instruction Report EL-95-2, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
Martin, J. L. 1986. Water quality study of proposed reregulation dam downstream of Wolf Creek Dam, Cumberland River,
    Kentucky. Miscellaneous Paper EL-86-4, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS. NTIS No. AD
    A167 132.
Martin, J. L., T. Curtis, and J.M. Nestler. 1986. Effects of flow alterations on trout habitat in the Cumberland River below Wolf
    Creek Dam, Kentucky. Miscellaneous Paper EL-86-11, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
    NTIS No. ADA176481.
Martin, J.L., and L. Owoputi. 1997. Water Quality Model Application to Youghiogheny Lake and River, Pennsylvania. Presented
    at the 17th International Symposium on Lake and Reservoir Management, Decemer 2-5, Houston, Texas.
Nestler, J. M., J.A. Gore, L.T. Curtis, and J.L. Martin. 1988. Predicted effects of hydropower uprate  on trout habitat in the
    Cumberland River, downstream of Wolf Creek Dam, Kentucky. Miscellaneous Paper EL-88-10, U.S. Army Engineer
    Waterways Experiment Station, Vicksburg, MS. NTIS No. AD A200 562.
Owoputi, L., and J.L. Martin. 1998. Water Quality Model Application to Stonewall Jackson Lake and River, Pennsylvania.
    Presented at the 18th International Symposium on Lake and Reservoir Management, November 10-13, Banff, Alberta.
Schreiner, S. 1997. A Temperature Simulation Model of the Youghiogheny River From Deep Creek Station To Sang Run. Report
    PPRP-DC1, Maryland Power Plant Research Program, Annapolis, Maryland.
Zimmerman, M. J., and Dortch, M. S. 1988. Water quality modeling study of proposed reregulation dam downstream from
    Buford Dam, Chattahoochee River, Georgia. Technical Report EL-88-14, U.S. Army Engineer Waterways Experiment
    Station, Vicksburg, MS. NTIS No. AD A200 039.


CE-QUAL-W2: CE-QUAL-W2 is a two-dimensional, longitudinal/vertical, hydrodynamic and water quality model developed by
the Waterways Experiment Station (WES). Because the model assumes lateral homogeneity, it is best suited for relatively long
and narrow waterbodies exhibiting longitudinal and vertical water quality gradients.  The  model has been applied to rivers, lakes,
reservoirs, and estuaries. Application of CE-QUAL-W2 is complicated and very time consuming. The WES website offers "A
word of caution to the first time user". The model predicts water surface elevations,  velocities, and temperatures. Temperature is
included in the hydrodynamic calculations because of its effect on water density. Water quality. The water quality algorithms
incorporate 21  constituents in addition to temperature  including nutrient/phytoplankton/dissolved  oxygen (DO) interactions
during anoxic conditions. Any combination of constituents  can be simulated.  The  effects  of  salinity or total dissolved
solids/salinity on density and thus 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 additional 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). Storage
                                                        65

-------
requirements for  hydrodynamic output to drive the water quality model are prohibitive for anything except very small grids.
Additionally, reduction in computer time is minimal when hydrodynamic data used to drive water quality are input every time
step. The WERF 2001 reports over 200 applications of CE-QUAL-W2 to rivers, lakes, reservoirs, and estuaries in the U.S. and
throughout the world.
URL: http://www.wes.army.mil/el/elmodels

Application and Model References:
Adams, W., E. Thackston,and R. Speece. 1997. Modeling CSO impacts from Nashville using EPA's demonstration approach. J.
    Environ. Engr, 123 (2), pp. 126-133.
Cole, T.M. 1994. The future role of sophisticated models in reservoir management. Lake and Reservoir Management, 9 (2):64.
Cole, T.M., and Buchak, E.M.  1995. CE-QUAL-W2: A two-dimensional, laterally averaged, hydrodynamic and water quality
    model, version 2.0. Instruction Report EL- 95-1, US Army Engineer Waterways Experiment Station, Vicksburg, MS.
Easley, E., L. Barness-Walz, P. Neichter, and J. Bohannon. 1994. Evaluation of water quality in Taylorsville Lake, Kentucky,
    using the CE-QUAL-W2 model. Lake and Reservoir Management, 9(2):71-72.
Guenduez, O, S. Soyupak, and C. Yurteri. 1998. Development of water quality management strategies for the proposed Isikli
    Reservoir. Reservoir Management and Water Supply - An Integrated System (P. Delojs, J. Edzwald, C. O'Melia, and G.
    Oskam,  eds), Water Science & Technology, 37(2).
Harrison, J.,  and K. Anderson.  1997. Brownlee Reservoir water quality model response to nutrient and algae inflow
    concentration. Draft report to Idaho Power. HDR/CH2M-Hill Project Team.
Hayes, B., G. Hauser, and M. Eiffe. 1994. Two-dimensional water quality modeling of Douglas Reservoir. Lake and Reservoir
    Management, 9(2):80.
Kingery, D., and J. Harrison. 1997. Brownlee Reservoir: water quality model development. Draft report to Idaho Power.
    HDR/CH2M-Hill Project Team.
Shiao, M., P. Craig, B. Hayes,  and J. Parsly.  1994. Learning reservoir water quality dynamics with computer animation. Lake and
    Reservoir Management, 9(2): 114.


CH3D-SED & CH3-WES: CH3D-SED is the newly developed mobile bed version of CH3D-WES which is a three dimensional
hydrodynamic  model developed for the Chesapeake Bay Program. It is applicable to rivers, streams, estuaries and coastal zones.
The physical processes modeled are tides, wind, density effects (salinity and temperature) freshwater inflows, turbulence and the
effect of the earth's rotation. A boundary fitted, non-orthogonal, 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.  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-s
closure for  vertical  momentum diffusion.  Sedimentation computations  are based  on  a two dimensional solution  of the
conservation of mass 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 will have negligible
effects on sediment mobility characteristics. Since each grain size specification is independently tracked, however, tracking  of
zones of contaminated bed material is possible. Model requires substantial expertise for efficient usage. It is publicly available but
not well documented.
URL: http://chl.wes.army.mil/software/ch3d

Application and Model References:
Cerco, C.F. andT. Cole. 1993. Three-Dimensional Eutrophication Model of Chesapeake Bay. Journal of Environmental
    Engineering. 119(6): 1006-1025.
Chapman, R.S., B.H. Johnson, and S.R. Vemulakonda. 1996. User's Guide for the Sigma Stretched Version of CH3D-WES.
    Technical Report HL-96-21, US Army Engineer Waterways Experiment Station, Vicksburg, MS.
Engel, J. J., R.H. Hotchkiss, and B.R. Hall. 1995. Three Dimensional Sediment Transport Modeling Using CH3D Computer
    Model. 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, 628-632.
                                                        66

-------
Hall, B.R. 1996. Quantifying Sedimentation Using a Three Dimensional Sedimentation Model. Water Quality '96, Proceedings of
    the llth Seminar, Corps of Engineers Committee on Water Quality, Seattle, WA, 88-93.
Johnson, B.H., R.E. Heath, B.B. Hsieh, K.W. Kim, and H.L. Butler. 1991. User's Guide for a Three-Dimensional Numerical
    Hydrodynamic, Salinity, and Temperature Model of Chesapeake Bay. Department of the Army, Waterways Experiment
    Station, Corps of Engineers, Vicksburg, MS.
Johnson, B.H., K.W. Kim, R.E., 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.
Spasojevic, M., and P.M. Holly. 1994. Three-Dimensional Numerical Simulation of Mobile-Bed Hydrodynamics. Contract
    Report HL-94-2, US Army Engineer Waterways Experiment Station, Vicksburg, MS.


DELFT3D: DelftSD  is a 2D/3D integrated modeling environment for hydrodynamics, waves, sediment transport, morphology,
water quality, particle tracking for water quality,  and ecology. The FLOW module of DelftSD is a multi-dimensional calculates
non-steady flow and transport phenomena resulting from tidal and meteorological forcing on a curvilinear, boundary fitted grid.
The areas of applications are: 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. The sediment module (SED) of
DelftSD can be applied to  model the transport of cohesive and non-cohesive sediments,  e.g. spreading of dredged materials, to
study sediment/erosion patterns.  Sedimentation takes place when the bottom shear stress drops below a critical value. The model
treats each of the paniculate fractions independently (i.e. sand and silt). 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. Sediment can be transferred downward from one sediment
layer to an underlying layer in a process known as 'burial'. Sediment can be transferred  upward to one sediment layer from an
underlying layer in a process known as 'digging'.  The water quality (WAQ) module can include any combination of constituents
and is not limited to  the number and complexity of the processes. For many water quality problems, the process formulations
have been standardized in the form of a library. The water quality processes may be described by linear or non-linear functions of
the selected state variables and model parameters. Typical applications of WAQ are biochemical reactions like the decay of BOD
and nitrification, growth of algae (primary production) and nutrient cycling, 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,  mortality of bacteria, and predation (e.g. zooplankton  on phytoplankton). The PART module
of DELFT3D simulates transport processes and simple chemical reactions by means of a particle tracking method using the flow
data from the FLOW module.  The tracks are followed in three  dimensions over  time,  whereby a dynamic concentration
distribution is obtained through averaging of separate particle tracks. DELFT3D requires huge amount of resources. According to
the model web site the minimal and recommended resources are as follows:
                                                   Minimal        Preferred
                                   Processor       Pentium         Pentium 4
                                                   166 MHz        1 GHz or more

                                   Internal         64MB          512MB or
                                   memory                         more

                                   Free disk        2 GB            10 GB
                                   space


URL: http://www.wldelft.nl/soft/d3d/index.html

Application and Mode I References:
Bent E.J., L. Postma, A. Roelfzema, and R.J.H. Stive. 1991. Hydrodynamic and dispersion modeling of Swansey Bay, IK,
    Enivronmental Hydraulics, 1:865-870.
Gerritsen, H., A.C. Baart, and J.G. Boon. 1997. NOMADS: North Sea Model Advection Dispersion Study : experiments:
    instantaneous releases : intercomparison of 2D and 3D model results. WL, research Z2084, January 1997.
Salden, R.M., J.M. de Kok, J.G. Boon, and H. Gerritsen. 1996. NOMADS: NOrth sea Model Advection-Dispersion Study : the
    Dutch contribution to the simulations. WL report Z 0854/Z 0995/T 1643. (Rijkswaterstaat, RIKZ, report 96.010).
                                                        67

-------
Vatvani, D., and M. Montazeri. 1989. Performance of some high accurate semi Langrangian numerical schemes for the scalar
    advection equation, Dt. Hydrogr. Zeitung, 42, H.3-6, Semi-Langrangian numerical Schemes, pp 279-305.
Vos, R.J., A.G. Dekker, S.W.M. Peters, G.A. van Rossum, and LJ. Hooijkaas. 1998. RESTWAQ 2, part II: comparison of
    remote sensing data, model results and in-situ data for the southern Frisian lakes (1998). BCRS report no. 98-08b), i.s.m.
    WL, IvM-VU, NIOZ, KNMI, K&M en waterschap Friesland.  - ISBN 90-5411-255-7.
Vos, R.J. 1995. Restwaq : applications  of remote sensing to water quality modeling : data assessment and development of
    methodology. WL report T1083/T1479, maart 1995, i.o.v. Rijkswaterstaat, Meetkundige Dienst.
Vos, R.J., and M. Schuttelaar. 1995. RESTWAQ : data assessment, data-model integration and application to the Southern North
    Sea. BCRS report no. 95-19, December 1995. ISBN 90-5411-168-2
Vos, R.J., E. J. de Goede, and R.E. Uittenbogaard. 1999. Validation of a 3D temperature model for the North Sea with in-situ data
    and remote sensing data. WL report Z 2506, February 1999. I.o.v. Rijkswaterstaat, RIKZ.


DYNHYD5: The DYNHYD5  model is a USEPA supported simple hydrodynamic model  that simulates variable tidal cycles,
wind, and unsteady inflows. It produces  an output file that can be linked with WASP5 to supply the flows and volumes to the
water quality model.  It can simulate velocity, volume, and water depth in rivers and streams, estuaries and costal waters, and
reservoirs  and lakes.  The WASP hydrodynamics model DYNHYD is an  enhancement of  the Potomac Estuary hydrodynamic
model which was a component of  the Dynamic Estuary Model. DYNHYD solves the one-dimensional equations of continuity
and momentum for a branching or  channel-junction (link-node), computational network. Driven by variable upstream flows and
downstream heads, simulations typically proceed at one- to five-minute intervals. The  resulting unsteady hydrodynamics are
averaged over larger time intervals and stored for later use by the water quality program. The hydrodynamic model  solves one-
dimensional equations  describing  the propagation of a long  wave through a shallow water system while conserving both
momentum (energy)  and volume (mass). The equation of motion, based on the conservation  of  momentum,  predicts water
velocities  and flows. The  equation of continuity, based on the conservation of volume,  predicts water heights (heads) and
volumes. This approach assumes that flow is predominantly  one-dimensional, Coriolis  and other accelerations normal to the
direction of flow are negligible, channels can be adequately represented by a constant top width with a variable hydraulic depth,
i.e., rectangular,  the wave  length is significantly greater than the depth,  and bottom slopes are moderate. Although no strict
criteria are available for the  latter two  assumptions, most natural flow conditions in large rivers and estuaries would be
acceptable. Dam-break situations could not be simulated with DYNHYD nor could small mountain streams. Both DOS and
Windows versions are available.
URL: http://www.epa.gov/ceampubl/swater/wasp/index.htm. http://www.cee.odu.edu/model/wasp.php.
    http://www.scisoftware.com/products/wasp overview/wasp overview.html

Application and Model References:
Ambrose, R., T.A. Wool, and J.L. Martin. 1993. The Dynamic Estuary Model Hydrodynamics Program, DYNHYD5: Model
    Documentation and User Manual. Environmental Research Laboratory. USEPA. Athens, GA.
Cusimano, R.F. 1995. Snohomish River Estuary Dry Season, TMDL Study - Phase I: Water Quality Model Calibration.
    Washington State Department of Ecology, Watershed Assessment Section, Olympia, Washington. 56 p.
Cusimano, R.F. 1997. Snohomish River Estuary Dry Season, TMDL Study - Phase II: Water Quality Model Confirmation and
    Pollutant Loading Capacity Recommendations.
Roesch, S.E., LJ. Clark, and M.M.  Bray.  1979. User's Manual for the Dynamic (Potomac) Estuary Model. U.S. Environmental
    Protection Agency, Annapolis,  MD. EPA-903/97-001.
Warwick, J.J., and KJ. Heim. 1995. Hydrodynamic Modeling of the Carson River and Lohontan Reservoir, Nevada. Water
    Resources Bulletin 3l(l):67-77.


EFDC (Environmental Fluid Dynamics Code): EFDC is a three dimensional hydrodynamic and transport model, but it can be
used for two, even one-dimensional problems, though not recommended.  It is applicable to estuaries, costal ocean, lakes, and
reservoirs. Momentum and  conservation equations form the basis of governing hydrodynamic equations. A Mellor-Yamada level
2.5 turbulence closure  scheme is employed  to compute vertical mixing coefficients. The model is based on the curvilinear-
orthogonal horizontal grid with a sigma stretched (or topography following) vertical coordinate system. Effects of wind waves on
bottom stresses can be  simulated. Vegetation resistance can be simulated  in submerged  and emergent vegetated environments.
Wetting and drying computational cells can be simulated allowing modeling of wetlands and estuaries with shallow marshes. The
sediment routine used in EFDC is relatively unsophisticated. Both cohesive and non-cohesive sediments can be simulated. User is
given the option to select number of sediment size classes. The model does not consider the effect of armoring which is  shown to
be a very important process in estuarine waterbodies. A simplistic rather obsolete heat exchange budget model is utilized. EFDC
has the  internal  capability  to  simulate the transport  and transformation of an arbitrary number of dissolved and suspended
                                                        68

-------
constituents. Transformation kinetics are specified by a user-specified subroutine. The model is written is Fortran-77 meaning
that it can be used on any platform after proper calibration. However its usage requires very high level of expertise. Indirect
linkages between EFDC and WASPS and CE-QUAL-ICM water quality models are possible, as EFDC has the ability to generate
outputs files already in the format for input to these water quality models. Works is going on to include EFDC to the USEPA's
BASINS system. There is no web site dedicated to EFDC for providing information. Model source code and manual can be
obtained by contacting:
Contacts:
        John M. Hamrick                                     Virginia Institute of Marine Science
        Tetra Tech, Inc.                                       School of Marine Science
        10306 Eaton Place, Suite 340                            The College of William and Mary
        Fairfax, VA 22030                                    Gloucester Point, VA 23 502
        (703) 385-6000                                        (804) 642-7000
        ham@visi.net

Application and Model 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
    Estuarine and Coastal Modeling, Proceedings of the 2nd International Conference, ed. M.L. Spaulding, et al., pp. 292-303.
    American Society of Civil Engineers, New York.
Hamrick, J.M. 1996. A User's Manual for the Environmental Fluid Dynamics Computer Code (EFDC). The College of William
    and Mary, Virginia Institute of Marine  Science, Special Report 331, 234 pp.
Hamrick, J.M., and T.S. Wu. 1996.  Computational design and optimization of the EFDC/HEM3D surface water hydrodynamic
    and eutrophication models. In Computational Methods of Next Generation Environmental Models, ed. G. Delich, Society of
    Industrial and Applied Mathematics, Philadelphia.
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. The College of William and Mary, Virginia Institute of
    Marine Science, Gloucester Point, VA. Special Report 327, 113 pp.
Tetra Tech. 1994. User's guide for the three-dimensional EFDC hydrodynamic and salinity model of Indian River Lagoon and
    Turkey Creek. Final Report. Tetra Tech, Inc., Fairfax, VA.


HSPF: See loading models.

MIKE-11: See loading models.


MIKE-21: MIKE-21 is supported and distributed by the DHI Software. It contains a comprehensive modeling system for 2D free-
surface flows and is applicable to the simulation of hydraulic and related phenomena in lakes, estuaries, bays, coastal areas and
seas where stratification can be neglected.  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 compiled as a true  32-bit application implying that it  can only be executed under Windows 95/98 or
Windows NT. MIKE-21 is constructed in a modular manner around the four main application areas:
    •   Coastal hydraulics and oceanography: Includes two modules: the Hydrodynamic Module  (HD)and  the Nested  Grid
        Hydrodynamic Module  (NHD). The HD  Module  (MIKE-21 HD)  is the  basic module in the MIKE-21  package. It
        provides the  hydrodynamic basis for the computations  performed  in the modules for Sediment  Processes and
        Environmental Hydraulics. The HD Module simulates the  water level variations and flows in response to a variety of
        forcing functions in lakes, estuaries and coastal areas. The water levels and flows are resolved on a rectangular grid
        covering the area of interest when provided with the bathymetry, bed resistance coefficients, wind field, hydrographic
        boundary conditions, etc. The system solves the full time-dependent non-linear equations of continuity  and conservation
        of momentum. The solution is obtained using an implicit ADI finite difference scheme of second-order accuracy. The
        outcome of a simulation is the water level and fluxes (velocities) in the computational domain.
    •   Environmental hydraulics: The group of environmental modules include Advection-Dispersion Module (AD) plus three
        process modules: Water Quality Module  (WQ), Eutrophication Module (EU), Heavy Metal Module (ME) and  Spill
        Analysis Module (SA). All these environmental modules are also available as nested grid versions: NAD, NWQ, NEU,
        NME, and NSA.  All modules use output from the HD (or NHD) Module,  and the AD  (or NAD) Module is used
        automatically by  the three process modules. The AD Module simulates the spreading of dissolved substances subject to
                                                        69

-------
        advection and dispersion processes, eg: salt, heat, conform bacteria, xenobiotic compounds etc. Linear decay and heat
        dissipation  to the atmosphere are included. The WQ Module used for advanced water quality studies considers the
        following determinants: dissolved oxygen (DO), organic matter (BOD), ammonia, nitrate, and phosphorus. EU Module
        simulates carbon and nutrient  cycling,  growth  of  phytoplankton and zooplankton,  oxygen balance, and benthic
        vegetation.  The state variables included  in the ME modules are dissolved metal in water, adsorbed metal in water,
        suspended sediment, dissolved metal in the bed porewater, and metal adsorbed on sediment in the bed sediment layer
        thickness. The Spill Analysis Module of MIKE-21 simulates the spreading and weathering of suspended substance in an
        aquatic environment under the influence of the fluid transport and the associated dispersion processes.
    •   Sediment processes: MIKE-21 comprises three types of sediment transport models.  Sand Transport Module  (ST), Mud
        Transport Module (MT), and Particle Module (PA). ST is used to determine the sediment transport rates due to the effect
        of current only, or a combination of current and waves in areas with a sandy bottom. MT describes the erosion, transport
        and deposition of cohesive sediments  (mud, silt or clay) under the action of waves and currents. The model also  takes
        into account the  consolidation of the  bed. The model can be used to determine  the siltation of cohesive materials in
        harbors,  lagoons or coastal  areas and to  determine the fate of dredged spoils. PA  describes the transport and fate of
        solutes or suspended matter. The model  can be used to determine the fate of suspended matter that is discharged or
        accidentally spilled in lakes, estuaries,  coastal areas or the open sea. Settling and decay processes are included.
    •   Waves: A range of wave modules are included in MIKE-21, each with their particular area of application. The models
        can be divided basically into two groups:  models based on wave action concept (OSW and NSW), and models based on
        the momentum concept (BW, EMS and PMS). Interested reader's can find details of this module at the URL below.
The US  Federal  Emergency Management Agency (FEMA) has approved three modules  of MIKE-21 for National  Flood
Insurance Program (NFIP) usage. The three modules, which are hydrodynamic module (HD/NHD), near-shore spectral wind-
wave module (NSW) and offshore spectral wind-wave module (OSW), have been accepted for coastal storm surge, coastal  wave
height, and coastal wave effect usage.
URL: http://www.dhisoftware.com/mike21

Application and Model References:
Gierlevsen, T., M. Hebsgaard, and J. Kirkegaard. 2001. Wave Disturbance Modeling in the Port of Sines, Portugal- with special
    emphasis on long period oscillations. Presented at the International Conference on Port  and Maritime R&D and Technology,
    Singapore, 29-31 October 2001.
Hansen, H.K., P. Sloth, O.R. Serensen, and J. Fuchs.  2000. Combined numerical and physical modelling of seiching in exposed
    new marina. In Proceedings of 27th International  Coastal Engineering Conference, 16-21 July 2000, Sydney, Australia.
Johnson, H.K., C.M. Appending M. Soldati, B.  Elfrink, P. Serensen. 2001. Numerical modeling of morphological changes due to
    shoreface nourishment. In: Proc of the 4th Conference on Coastal Dynamics, ASCE, pp.878-887. Lund, Sweden, June 2001.
McCowan, A.D., E.B. Rasmussen, and P. Berg. 2001. Improving the Performance of a Two-dimensional Hydraulic Model for
    Floodplain Applications. Presented at the 6th Conference on Hydraulics in Civil Engineering, I.E. Aust, 28-30 November
    2001, Hobart.

Also visit http://www.dhi.dk/ContactUs/Library for additional all DHI compendium of technical papers and publications.


MIKE-3:  Yet another DHI product, MIKE-3, is applicable  for simulations  of hydrodynamics, water quality and sediment
transport in all waterbodies where 3D effects are important.  MIKE-3 is compatible with MIKE-21 and  other DHI Software
products.  MIKE-3 simulates unsteady flow taking  into account density variations, bathymetry and external forcing such as
meteorology, tidal elevations, currents and other hydrographic conditions. MIKE-3 is designed in a modular structure with the
three main components:
    •   Estuarine and coastal hydraulics  and oceanography:  The hydrodynamic module  (HD) is the  core of the MIKE-3
        modeling system. It provides the hydrodynamic basis for computations performed in other modules (water quality,
        eutrophication etc). MIKE-3 HD solves  the time-dependent conservation equations of mass and momentum in  three
        dimensions, the  so-called Reynolds-averaged Navier-Stokes  equations, where  the flow  is decomposed  into  mean
        quantities and turbulent fluctuations.  The flow field and pressure variation are computed in response to a variety of
        forcing functions, when provided with the  bathymetry, bed resistance, wind field, hydrographic boundary  conditions,
        etc. The closure problem is solved in the  turbulence module through the Boussinesq eddy viscosity concept relating the
        Reynold  stresses to the mean velocity  field. To handle density variations, the equations for conservation of salinity and
        temperature are included and solved in the transport equation module.  An equation of state (the UNESCO formulation)
        constitutes the relation between the density  and the variations in salinity and temperature. Thus, the turbulence module
        and the transport equation module are integrated components of the hydrodynamic module, and the suite of  those three
        constitutes the HD module. The hydrodynamic phenomena included in the equations are  tidal propagation, effects of
                                                         70

-------
        stratification, turbulent (shear) diffusion and dispersion, Coriolis forces, barometric pressure gradients, wind stress,
        variable bathymetry and bed resistance,  flooding and drying of intertidal areas, hydrodynamic effects of rivers  and
        outfalls, sources and sinks (both mass and momentum), and heat exchange with the atmosphere including evaporation
        and precipitation.
    •   Environmental hydraulics: The group of environmental modules includes the advection-dispersion module  (AD),  and
        two process modules: the water quality module (WQ) and the eutrophication module (EU). All environmental modules
        are similar to those used  in the MIKE-11 and MIKE-21 packages. The WQ Module used for advanced water quality
        studies considers, dissolved oxygen (DO), organic matter  (BOD), ammonia, nitrate, and phosphorus. The simulated
        physical, chemical and biological processes include  carbon and nutrient  cycling, growth of phytoplankton  and
        zooplankton, oxygen balance, and benthic vegetation.
    •   Sediment processes: MIKE-3 includes two types of sediment transport modules: the mud transport module (MT) and the
        particle module (PA). The modules for sediment processes are also similar to those used in MIKE-11 and MIKE-21.
All facilities necessary for data preparation and analysis are contained in MIKE-3 or under the common MIKE Zero shell.  The
compatibility between MIKE-3 and MIKE-21 implies that many of the facilities are common in the two model packages. All
input to MIKE-3 is handled through a dialogue-based user interface. The output from MIKE-3 can be either time series of points,
lines, 2D  maps or full 3D matrices. This output may be further  processed, analyzed, printed and presented graphically as
appropriate.
URL: http://www.dhisoftware.com/mike3

Application and Model References:
Reference Manual and Scientific Documentation are provided for each module within the MIKE-3 package along with an on-line
    help system. The URL http://www.dhi.dk/ContactUs/Library lists all DHI compendium of technical papers and publications.


QUAL2E:  The Enhanced  Stream Water Quality  Model (QUAL2E) is in public domain and is supported and distributed by
USEPA. It is included in the EPA's BASINS system. QUAL2E is applicable to well mixed dendritic streams. It is basically one-
dimensional and operates  as a steady  state model. It can simulate up to  15 water constituents including dissolved  oxygen,
biochemical oxygen demand, temperature, algae, organic nitrogen, ammonia, nitrite, nitrate, organic Phosphorous, and dissolved
phosphorous. Advection, dispersion, dilution, constituent reactions  and interactions, and sources and sinks are all considered
within the model. Analyzing the impact of waste loads on the stream quality, effects of diurnal variations in meteorological data
on water quality (mainly dissolved oxygen and temperature) and diurnal oxygen variations due to algal growth are some potential
areas of use of QUAL2E. QUAL2E does not have a hydrodynamic component, therefore data pertinent to flow must be provided
by the user. QUAL2E has been one of the most heavily used water quality models in the United States.  Most of its applications
were addressing dissolved oxygen problems. QUAL2EU is an enhancement  to  QUAL2E  which allows  users to perform
uncertainty analysis. It offers three uncertainty options to the user: sensitivity analysis, first order error analysis, and Monte Carlo
simulations. The windows version of QUAL2E greatly facilitates the input preparation. It provides screens to prepare input, run
the model and visualize the model results. It also offers a help screen. The windows version comes with three examples with data
sets included to demonstrate the usage of the model. This version including model manual can be downloaded from
http://www.epa.gov/waterscience/QUAL2E WINDOWS/index.html.
The DOS version can be downloaded from
http://www.epa.gov/ceampubl/swater/qual2eu/index.htm

Application and Model References:
Brown,  L.C. and T.O. Barnwell, Jr., 1987. The Enhanced Stream Water Quality Models QUAL2E and QUAL2E-UNCAS:
    Documentation and User's Manual. (EPA 600/3-87-007). NTIS Accession Number:PB87 202 156.
Cubilo,  F., B. Rodriguez, and T.O. Barnwell, Jr.. 1992. A system for control of river water quality for the community of Madrid
    using Qual2E. Water Science and Technology 26(7/8): 1867-1873.
Johnson, C.R., and G. Mercer.  1994. Modeling the water quality processes of the Chicago waterway. In Proceedings of the
    National Symposium on Water Quality, American Water Resources Association, Chicago, IL, November 6-10, 1994, p. 315.
Little, K.W. and R.E. Williams. 1992. Least squares calibration of QUAL2E. Water Environment Research 64(2):79-185.
Macaitis, B. and C. Johnson. 1993. Water quality model of the Chicago waterway. Proceedings of the 20th Anniversary
    Conference on Water Management in the '90s, ASCE, pp 189-192.
Melching, C.  and T. Chang. 1996. Simulation of water quality for Salt Creek in northeastern Illinois. USGS Open-File Report:
    96-318.
Paschal, J.E., Jr., and O.K. Mueller. 1991. Simulation of water quality and the effects of wastewater effluent on the South Platte
    River from Chatfield Reservoir through Denver,  Colorado. Water-Resources Investigations Report 91-4016. U.S. Geological
    Survey, Denver,  CO.
                                                        71

-------
Pelletier, G. 1997. Colville River Water Quality: Pollutant Loading Capacity and Recommendations for Total Maximum Daily
    Loads. Report Number: 96-349, Washington State Department of Ecology. (Also available at:
    http://www.ecy.wa.gov/pubs/96349.pdf)
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.


WASP6 (Water quality Analysis Simulation Program): WASP6 is an enhanced Windows version of the USEPA Water Quality
Analysis Simulation Program (WASP). WASP6 has been developed to aid modelers in the implementation of WASP.  WASP6
has features  including a pre-processor, a rapid data processor, and a graphical post-processor that enable the modeler to run
WASP more quickly and easily and evaluate model results both numerically and graphically. With WASP6, model execution can
be performed up to ten times faster than the previous USEPA DOS version of WASP. Nonetheless, WASP6 uses the same
algorithms to solve water quality problems as those used in the DOS version of WASP. The WASP6 modeling system, supported
and distributed by EPA's CEAM,  is a generalized modeling framework for contaminant fate and transport in surface waters.
Based on flexible compartment modeling,  WASP6 can be applied in one, two, or three dimensions. Problems  that have been
studied  using  WASP6 include biochemical  oxygen demand, dissolved oxygen dynamics, nutrients/eutrophication, bacterial
contamination, and toxic chemical movement.  The WASP6 system consists of two stand-alone computer programs, DYNHYD5
and WASP6 that can be run in conjunction or separately. WASP6 is supplied with two kinetic submodels to simulate two of the
major classes  of water quality problems:  conventional  pollution (involving  dissolved oxygen,  biochemical oxygen demand,
nutrients and eutrophication) and toxic pollution (involving organic chemicals,  metals, and sediment). The linkage of either
submodel with the WASP6 program gives the  models EUTRO and TOXI, respectively. The hydrodynamic data can be supplied
in three different ways to WASP: i) user can provide steady state flow data in a file, ii) DYNHYD5 output can be used or iii)
another hydrodynamic model can be linked. The Eutrophication Model  (EUTRO) combines a kinetic structure adapted from the
Potomac Eutrophication Model with  the  WASP6 transport structure.  This model predicts dissolved oxygen, carbonaceous
biochemical oxygen demand, phytoplankton, carbon, chlorophyll-a, ammonia, nitrate, organic nitrogen, and orthophosphate in
bed and overlying waters. The Toxic Chemical Model (TOXI) combines a kinetic structure adapted from the Exposure Analysis
Modeling System  (EXAMS) with the WASP6 transport structure and simple  sediment balance algorithms.  TOXI  predicts
dissolved and  sorbed chemical concentrations in the bed and overlying waters. Sediment modeling is based on simple mass
balance. The WASP6 package also includes three  other programs: PREDYN, W5DSPLY and PLOT. PREDYN is an interactive
preprocessor program for DYNHYD5. W5DSPLY is a tabular post processor program for TOXI,  EUTRO and DYNHYD5.
PLOT is a graphical post processor for TOXI, EUTRO and DYNHYD. WASP6  is one of the well-established  models  and
numerous applications are available.  There are  several other  hydrodynamic models that have been linked  with WASP6:
DYNHYD5, RIVMOD, EFDC and SWMM's transport module.
URL: http://www.epa.gov/ceampubl/swater/wasp/index.htm. http://www.cee.odu.edu/model/wasp.php.
    http://www.scisoftware.com/products/wasp overview/wasp overview.html

Application and Model References:
Cheng, C. J.E.  Atkinson, and J.V. DePinto.  1994. A coupled GIS-water quality modeling study. In Proceedings of the 1994
    Hydraulic  Engineering Conference, ASCE, Buffalo, NY, 1994, pp. 247251.
Cockrum, O.K., and JJ. Warwick. 1994. Assessing the impact of agricultural activities on water quality in a periphytondominated
    stream using the Water Quality Analysis Program (WASP). In Proceedings of the Symposium on the Effects of Human
    Induced Changes onHydrologic Systems, AWRA, Jackson Hole, WY, June 26-29, 1994, p. 1157.
Hajda, P., and V. Novotny. 1996. Modelling Impact of Urban and Upstream Nonpoint Sources on Eutrophication of the
    Milwaukee River.  Wat. Sci. Tech. 44:153-158.
Minei, V., and W. Dawydiak.  1995. Controlling nitrogen inputs into the Peconic Estuary system. 2nd Annual Marine & Estuarine
    Shallow Water Science and Management Conf., USEPA, p. 32.
Lang, G.A., and T.D. Fontaine. 1990. Modeling the fate and transport of organic contaminants in Lake St. Clair. Journal of Great
    Lakes Research 16(2):216-232.
Lu, Z., G.C. April, D.C. Raney, and W.W. Schroeder. 1994. DO, BOD, and organic nitrogen transport in Weeks Bay, Alabama.
    In Proceedings of the National  Symposium on Water Quality, American Water Resources Association, Chicago, IL,
    November 6-10, 1994, pp.
Lung, W., and  C.E. Larson. 1995. Water quality modeling of the upper Mississippi River and Lake Pepin. Journal of
    Environmental Engineering 121(10):691-699.
Tetra Tech. 1995. Hydrodynamic and water quality mathematical modeling study of Norwalk harbor, Connecticut. Final report.
    Tetra Tech, Inc., Fairfax, VA.
                                                       72

-------
Thomann, R.V., and J.J. Fitzpatrick. 1982. Calibration and Verification of a Mathematical Model of the Eutrophication of the
    Patomac estuary. Prepared for Department of Environmental Services, Government of the District of Columbia, Washington,
    D.C.
Zhou, I, 1998. Water quality modeling of reservoir using WASP. Proceedings of the 1998 International Water Resources
    Engineering Conference, Part 2, ASCE, pp 1458-1463.
                                                        73

-------