LARGE AREA HYDROLOGIC MODELING AND ASSESSMENT PART I: MODEL DEVELOPMENT1

ABSTRACT: A conceptual, continuous time model called SWAT (Soil and Water Assessment Tool) was developed to assist water resource managers in assessing the impact of management on water supplies and nonpoint source pollution in watersheds and large river basins. The model is currently being utilized in several large area projects by EPA, NOAA, NRCS and others to estimate the off-site impacts of climate and management on water use, non-point source loadings, and pesticide contamination. Model development, operation, limitations, and assumptions are discussed and components of the model are described. In Part II, a GIS input/output interface is presented along with model validation on three basins within the Upper Trinity basin in Texas.     

WATERSHED AND LAKE WATER QUALITY ASSESSMENT: AN INTEGRATED MODELING APPROACH1

Watershed planning groups and action agencies seek to understand how lake water quality responds to changes in watershed management. This study developed and demonstrated the applicability of an integrated modeling approach for providing this information. An integrated model linking watershed conditions to water-quality of the receiving lake incorporated the following components: (1) an event-based AGNPS model to estimate watershed pollutant losses; (2) annualization of AGNPS results to produce annual lake pollutant loadings; (3) a base flow separation package, SAM, to estimate base flow; (4) estimates of nutrients in base flow and point sources; and (5) linkage of watershed loadings directly to EUTROMOD lake water quality algorithms. Results are presented for Melvern Lake, a 28-km² multipurpose reservoir with a 900-km² agricultural watershed in east central Kansas. Reasonable estimates of current lake quality were attained using an average phosphorus availability factor of 31 percent to calibrate model results to measured in-lake phosphorus. Comparison of a range of possible scenarios, including all cropland changed to no-till (best case) and all CRP and good-condition grasslands changed to cropland (worst case), indicated only a (4 percent change for in-lake phosphorus and a (2 percent change for chlorophyll a. These results indicated that this watershed is not sensitive to projected changes in land use and management.     

HYDROLOGIC CALIBRATION OF THE SWAT MODEL IN A WATERSHED CONTAINING FRAGIPAN SOILS1

ABSTRACT: The Soil and Water Assessment Tool (SWAT) model, designed for use on rural ungaged basins and incorporating a GRASS GIS interface, was used to model the hydrologic response of the Ariel Creek watershed of northeastern Pennsylvania. Model evaluation of daily flow prior to calibration revealed a deviation of runoff volumes (D_v) of 68.3 percent and a Nash-Sutcliffe coefficient of-0.03. Model performance was affected by unusually large observed snowmelt events and the inability of the model to accurately simulate baseflow, which was influenced by the presence of fragipans. Seventy-five percent of the soils in the watershed contain fragipans. Model calibration yielded a D_v of 39.9 percent and a Nash-Sutcliffe coefficient of 0.04, when compared on a daily basis. Monthly comparisons yielded a Nash-Sutcliffe coefficient of 0.14. Snowmelt events in the springs of 1993 and 1994, which were unusually severe, were not adequately simulated. Neglecting these severe events, which produced the largest and third largest measured flows for the period of record, a D_v of 4.1 percent and Nash-Sutcliffe coefficient of 0.20 were calculated on a daily comparison, while on a monthly basis the Nash-Sutciffe coefficient was 0.55. These results suggest that the SWAT model is better suited to longer period simulations of hydrologic yields. Baseflow volumes were accurately simulated after calibration (D_v= -0.2 percent). Refinements made to the algorithms controlling subsurface hydrology and snowmelt, to better represent the presence of fragipans and snowmelt events, would likely improve model performance.     

HYDROLOGIC MODELING OF A BIOINFILTRATION BEST MANAGEMENT PRACTICE1

ABSTRACT: The goal of this research was to develop a methodology for modeling a bioinfiltration best management practice (BMP) built in a dormitory area on the campus of Villanova University in Pennsylvania. The objectives were to quantify the behavior of the BMP through the different seasons and rainfall events; better understand the physical processes governing the system's behavior; and develop design criteria. The BMP was constructed in 2001 by excavating within an existing traffic island, backfilling with a sand/soil mixture, and planting with salt tolerant grasses and shrubs native to the Atlantic shore. It receives runoff from the asphalt (0.26 hectare) and turf (0.27 hectare) surfaces of the watershed. Monitoring supported by the hydrologic model shows that the facility infiltrates a significant fraction of the annual precipitation, substantially reducing the delivery of nonpoint source pollution and erosive surges downstream. A hydrologic model was developed using HEC‐HMS to represent the site and the BMP using Green‐Ampt and kinematic wave methods. Instruments allow comparison of the modeled and measured water budget parameters. The model, incorporating seasonally variable parameters, predicts the volumes infiltrated and bypassed by the BMP, confirming the applicability of the selected methods for the analysis of bioinfiltration BMPs.     

HYDROLOGICAL MODELING OF THE IROQUOIS RIVER WATERSHED USING HSPF AND SWAT1

ABSTRACT: The performance of two popular watershed scale simulation models — HSPF and SWAT — were evaluated for simulating the hydrology of the 5,568 km² Iroquois River watershed in Illinois and Indiana. This large, tile drained agricultural watershed provides distinctly different conditions for model comparison in contrast to previous studies. Both models were calibrated for a nine‐year period (1987 through 1995) and verified using an independent 15‐year period (1972 through 1986) by comparing simulated and observed daily, monthly, and annual streamflow. The characteristics of simulated flows from both models are mostly similar to each other and to observed flows, particularly for the calibration results. SWAT predicts flows slightly better than HSPF for the verification period, with the primary advantage being better simulation of low flows. A noticeable difference in the models' hydrologic simulation relates to the estimation of potential evapotranspiration (PET). Comparatively low PET values provided as input to HSPF from the BASINS 3.0 database may be a factor in HSPF's overestimation of low flows. Another factor affecting baseflow simulation is the presence of tile drains in the watershed. HSPF parameters can be adjusted to indirectly account for the faster subsurface flow associated with tile drains, but there is no specific tile drainage component in HSPF as there is in SWAT. Continued comparative studies such as this, under a variety of hydrologic conditions and watershed scales, provide needed guidance to potential users in model selection and application.     

DETERMINATION OF BEST TIMING FOR POULTRY WASTE DISPOSAL: A MODELING APPROACH1

ABSTRACT: Confined production of poultry results in significant volumes of waste material which are typically disposed of by land application. Concerns over the potential environmental impacts of poultry waste disposal have resulted in ongoing efforts to develop management practices which maintain high quality of water downstream of disposal areas. The timing of application to minimize waste constituent losses is a management practice with the potential to ensure high quality of streams, rivers, and lakes downstream of receiving areas. This paper describes the development and application of a method to identify which time of year is best, from the standpoint of surface water quality, for land application of poultry waste. The procedure consists of using a mathematical simulation model to estimate average nitrogen and phosphorus losses resulting from different application timings, and then identifying the timings which minimize losses of these nutrients. The procedure was applied to three locations in Arkansas, and three different criteria for optimality of application timing were investigated. One criterion was oriented strictly to water quality, one was oriented only to crop production, and the last was a combination. The criteria resulted in different windows of time being identified as optimal. Optimal windows also varied with location of the receiving area. The results indicate that it is possible to land-apply poultry waste at times which both minimize nutrient losses and maximize crop yield.     

EMPIRICAL MODELING OF HYDROLOGIC AND NFS POLLUTANT FLUX IN AN URBANIZING BASIN1

ABSTRACT: Long term effects of precipitation and land use/land cover on basin outflow and nonpoint source (NFS) pollutant flux are presented for up to 24 years for a rapidly developing headwater basin and three adjacent headwater basins on the urban fringe of Washington, D.C. Regression models are developed to describe the annual and seasonal responses of basin outflow and IMPS pollutant flux to precipitation, mean impervious surface (IS), and land use. To quantify annual change in mean IS, a variable called delta IS is created as a temporal indicator of urban soil disturbance. Hydrologic models indicate that total annual surface outflow is significantly associated with precipitation and mean IS (r²= 0.65). Seasonal hydrologic models reveal that basin outflow is positively associated with IS during the summer and fall growing season (June to November). NPS pollutant flux models indicate that total and storm total suspended solids (TSS) flux are significantly associated with precipitation and urban soil disturbance in all seasons. Annual NPS total nitrogen flux is significantly associated with both urban and agricultural soil disturbance (r²= 0.51). Seasonal models of phosphorus flux indicate a significant association of total phosphorus flux with urban soil disturbance during the growing season. Total soluble phosphorus (TSP) flux is significantly associated with IS (r²= 0.34) and urban and agricultural soil disturbance (r²= 0.58). In urbanizing Cub Run basin, annual TSP concentrations are significantly associated with IS and cultivated agriculture (r²= 0.51).     

MODELING THE BUILDUP AND WASHOFF OF POLLUTANTS ON URBAN WATERSHEDS1

ABSTRACT: A model for urban stormwater quality was developed in this study. The basis for the model is the process by which pollutants build up on the watershed surface. For the wet climate of the study site, it was assumed that there exists an interval of time over which the pollutant buildup equals the pollutant washoff (no accumulation of pollutant). The buildup model was represented by a linear function of the antecedent dry time. The buildup function was then linked with a pollutant washoff model represented by a power function of the storm runoff volume. Various time intervals for no net accumulation were tested to calibrate the model. The model was calibrated to observed data for two small urban basins in Baton Rouge, Louisiana, and model results were used to analyze the behavior of phosphorus concentrations in storm runoff from these basins over a long period of time.     

CHALLENGES IN MODELING HYDROLOGIC AND WATER QUALITY PROCESSES IN RIPARIAN ZONES1

This paper presents key challenges in modeling water quality processes of riparian ecosystems: How can the spatial and temporal extent of water and solute mixing in the riparian zone be modeled? What level of model complexity is justified? How can processes at the riparian scale be quantified? How can the impact of riparian ecosystems be determined at the watershed scale? Flexible models need to be introduced that can simulate varying levels of hillslope‐riparian mixing dictated by topography, upland and riparian depths, and moisture conditions. Model simulations need to account for storm event peak flow conditions when upland solute loadings may either bypass or overwhelm the riparian zone. Model complexity should be dictated by the level of detail in measured data. Model algorithms need to be developed using new macro‐scale and meso‐scale experiments that capture process dynamics at the hillslope or landscape scales. Monte Carlo simulations should be an integral part of model simulations and rigorous tests that go beyond simple time series, and point‐output comparisons need to be introduced. The impact of riparian zones on watershed‐scale water quality can be assessed by performing simulations for representative hillsloperiparian scenarios.     

MODELING THE HYDROCHEMISTRY OF THE CANNONSVILLE WATERSHED WITH GENERALIZED WATERSHED LOADING FUNCTIONS (GWLF)1

ABSTRACT: A previous modeling study used the Generalized Watershed Loading Functions (GWLF) model to simulate stream‐flow, and nutrient and sediment loads to Cannonsville Reservoir from the West Branch Delaware River (WBDR). We made several model revisions, calibrated key parameters, and tested the original GWLF model and a revised GWLF model using more recent data. Model revisions included: addition of unsaturated leakage between unsaturated and saturated subsurface reservoirs; revised timing of sediment export; inclusion of urban sediments and dissolved nutrients; tracking of particulate nutrients from point sources; and revised timing of septic system loads. The revision of sediment yield timing resulted in significant improvements in monthly sediment and particulate phosphorus predictions as compared to the original model. Addition of unsaturated leakage improved hydrologic predictions during low flow months. The other model changes improve realism without adding significant model complexity or data requirements. Goodness of fit of revised model predictions versus stream measurements, as measured by the Nash‐Sutcliff coefficient of model efficiency, exceeded 0.8 for streamflow‐0.7 for sediment yield and dissolved nitrogen (N) and 0.6 for particulate and dissolved phosphorus (P). The revised GWLF model, with limited calibration, provides reasonable estimates of monthly streamflow, and nutrient and sediment loads in the Cannonsville watershed.     

INTEGRATING LANDSCAPE ASSESSMENT AND HYDROLOGIC MODELING FOR LAND COVER CHANGE ANALYSIS1

ABSTRACT: Significant land cover changes have occurred in the watersheds that contribute runoff to the upper San Pedro River in Sonora, Mexico, and southeast Arizona. These changes, observed using a series of remotely sensed images taken in the 1970s, 1980s, and 1990s, have been implicated in the alteration of the basin hydrologic response. The Cannonsville subwatershed, located in the Catskill/Delaware watershed complex that delivers water to New York City, provides a contrast in land cover change. In this region, the Cannonsville watershed condition has improved over a comparable time period. A landscape assessment tool using a geographic information system (GIS) has been developed that automates the parameterization of the Soil and Water Assessment Tool (SWAT) and KINEmatic Runoff and EROSion (KINEROS) hydrologic models. The Automated Geospatial Watershed Assessment (AGWA) tool was used to prepare parameter input files for the Upper San Pedro Basin, a subwatershed within the San Pedro undergoing significant changes, and the Cannonsville watershed using historical land cover data. Runoff and sediment yield were simulated using these models. In the Cannonsville watershed, land cover change had a beneficial impact on modeled watershed response due to the transition from agriculture to forest land cover. Simulation results for the San Pedro indicate that increasing urban and agricultural areas and the simultaneous invasion of woody plants and decline of grasslands resulted in increased annual and event runoff volumes, flashier flood response, and decreased water quality due to sediment loading. These results demonstrate the usefulness of integrating remote sensing and distributed hydrologic models through the use of GIS for assessing watershed condition and the relative impacts of land cover transitions on hydrologic response.     

A MODEL TO ENHANCE WETLAND DESIGN AND OPTIMIZE NONPOINT SOURCE POLLUTION CONTROL1

ABSTRACT: A dynamic, compartmental, simulation model (WETLAND) was developed for the design and evaluation of constructed wetlands to optimize nonpoint source (NPS) pollution control. The model simulates the hydrologic, nitrogen, carbon, dissolved oxygen (DO), bacteria, vegetative, phosphorous, and sediment cycles of a wetland system. Written in Fortran 77, the WETLAND models both free‐water surface (FWS) and subsurface flow (SSF) wetlands, and is designed in a modular manner that gives the user the flexibility to decide which cycles and processes to model. WETLAND differs from many existing wetland models in that the interactions between the different nutrient cycles are modeled, minimizing the number of assumptions concerning wetland processes. It also directly links microbial growth and death to the consumption and transformations of nutrients in the wetland system. The WETLAND model is intended to be utilized with an existing NPS hydro‐logic simulation model, such as ANSWERS or BASINS, but also may be used in situations where measured input data to the wetland are available. The model was calibrated and validated using limited data from a FWS wetland located at Benton, Kentucky. The WETLAND predictions were not statistically different from measured values for of five‐day biochemical oxygen demand (BOD₅), suspended sediment, nitrogen, and phosphorous. Effluent DO predictions were not always consistent with measured concentrations. A sensitivity analysis indicated the most significant input parameters to the model were those that directly affected bacterial growth and DO uptake and movement. The model was used to design a hypothetical constructed wetland in a subwatershed of the Nomini Creek watershed, located in Virginia. Two‐year simulations were completed for five separate wetland designs. Predicted percent reductions in BOD₅ (4 to 45 percent), total suspended solids (85 to 100 percent), total nitrogen (42 to 56 percent), and total phosphorous (38 to 57 percent) were similar to levels reported by previous research.     

IDENTIFICATION OF CRITICAL NONPOINT POLLUTION SOURCE AREAS USING GEOGRAPHIC INFORMATION SYSTEMS AND WATER QUALITY MODELING1

n integrated approach coupling water quality computer simulation modeling with a geographic information system (GIS) was used to delineate critical areas of nonpoint source (NPS) pollution at the watershed level. Two simplified pollutant export models were integrated with the Virginia Geographic Information System (VirGIS) to estimate soil erosion, sediment yield, and phosphorus (P) loading from the Nomini Creek watershed located in Westmoreland County, Virginia. On the basis of selected criteria for soil erosion rate, sediment yield, and P loading, model outputs were used to identily watershed areas which exhibit three categories (low, medium, high) of non-point source pollution potentials. The percentage of the watershed area in each category, and the land area with critical pollution problems were also identified. For the 1505-ha Nomini Creek watershed, about 15, 16, and 21 percent of the watershed area were delineated as sources of critical soil erosion, sediment, and phosphorus pollution problems, respectively. In general, the study demonstrated the usefulness of integrating GIS with simulation modeling for nonpoint source pollution control and planning. Such techniques can facilitate making priorities and targeting nonpoint source pollution control programs.     

INTEGRATED RIVER BASIN OPTIMIZATION: MODELING ECONOMIC AND HYDROLOGIC INTERDEPENDENCE1

ABSTRACT: This paper presents an integrated optimal control model that optimizes economic performance of reservoir management in watersheds in which there are significant economic and hydrologic interdependencies. The model is solved using the General Algebraic Modeling System (GAMS). Results show that application of this model to New Mexico's Rio Chama basin can increase total system benefits over historical benefits by exploiting complementarities between hydroelectricity production, instream recreation, and downstream lake recreation.     

A WATERSHED-LEVEL ECOLOGICAL RISK ASSESSMENT METHODOLOGY1

ABSTRACT: We present an ecological risk assessment methodology at the watershed level for freshwater ecosystems. The major component is a pollutant transport and fate model (a modified EUTROMOD) with an integrated uncertainty analysis utilizing a two-phase Monte Carlo procedure. The uncertainty analysis methodology distinguishes between knowledge uncertainty and stochastic variability. The model assesses the ecological risk of lentic (lake) ecosystems in response to the stress of excess phosphorus resulting in eutrophication. The methodology and model were tested on the Wister Lake watershed in Oklahoma with the lake and its trophic state as the endpoint for ecological risk assessment. A geographic information system was used to store, manage, and manipulate spatially referenced data for model input.     

A WINDOWS‐BASED GIS‐AGNPS INTERFACE1

ABSTRACT: ArcView Nonpoint Source Pollution Modeling (AVNPSM), an interface between ArcView GIS and AGNPS (Agricultural Nonpoint Source Pollution Model) is developed in support of agricultural watershed analysis and nonpoint source pollution management. The interface is PC‐based and operates in a Windows environment. It consists of seven modules: AGNPS utility, parameter generator, input file processor, model executor, output visualizer, statistical analyzer, and land use simulator. Basic input data to the interface include: soil, digital elevation model, land use/cover, water features, climate, and information on management practices. Application of the AVNPSM to a sample watershed indicates that it is user friendly, flexible, and robust, and it significantly improves the efficiency of the nonpoint source pollution modeling process.     

DEVELOPMENT AND APPLICATION OF A SPATIAL HYDROLOGY MODEL OF OKEFENOKEE SWAMP,GEORGIA1

ABSTRACT: :The model described herein was used to assess effects of the Suwannee River sill (a low earthen dam constructed to impound the Suwannee River within the Okefenokee National Wildlife Refuge to eliminate wildfires) on the hydrologic environment of Okefenokee Swamp, Georgia. Developed with Arc/Info Macro Language routines in the GRID environment, the model distributes water in the swamp landscape using precipitation, inflow, evapotranspiration, outflow, and standing water. Water movement direction and rate are determined by the neighborhood topographic gradient, determined using survey grade Global Positioning Systems technology. Model data include flow rates from USGS monitored gauges, precipitation volumes and water levels measured within the swamp, and estimated evapotranspiration volumes spatially modified by vegetation type. Model output in semi‐monthly time steps includes water depth, water surface elevation above mean sea level, and movement direction and volume. Model simulations indicate the sill impoundment affects 18 percent of the swamp during high water conditions when wildfires are scarce and has minimal spatial effect (increasing hydroperiods in less than 5 percent of the swamp) during low water and drought conditions when fire occurrence is high but precipitation and inflow volumes are limited.     

Integrated Modular Modeling of Water and Nutrients From Point and Nonpoint Sources in the Patuxent River Watershed1

Abstract: We present a simple modular landscape simulation model that is based on a watershed modeling framework in which different sets of processes occurring in a watershed can be simulated separately with different models. The model consists of three loosely coupled submodels: a rainfall‐runoff model (TOPMODEL) for runoff generation in a subwatershed, a nutrient model for estimation of nutrients from nonpoint sources in a subwatershed, and a stream network model for integration of point and nonpoint sources in the routing process. The model performance was evaluated using monitoring data in the watershed of the Patuxent River, a tributary to the Chesapeake Bay in Maryland, from July 1997 through August 1999. Despite its simplicity, the landscape model predictions of streamflow, and sediment and nutrient loads were as good as or better than those of the Hydrological Simulation Program‐Fortran model, one of the most widely used comprehensive watershed models. The landscape model was applied to predict discharges of water, sediment, silicate, organic carbon, nitrate, ammonium, organic nitrogen, total nitrogen, organic phosphorus, phosphate, and total phosphorus from the Patuxent watershed to its estuary. The predicted annual water discharge to the estuary was very close to the measured annual total in terms of percent errors for both years of the study period (≤2%). The model predictions for loads of nutrients were also good (20‐30%) or very good (<20%) with exceptions of sediment (40%), phosphate (36%), and organic carbon (53%) for Year 1.     

GEOGRAPHIC INFORMATION SYSTEM BASED NONPOINT POLLUTION MODELING1

ABSTRACT: The reauthorization of the Clean Water Act reemphasizes the need for regional scale monitoring and management of nonpoint pollution loads. The magnitude of the task will require that local governments and their consultants integrate information systems and modeling if they are to manage the massive data sets and conduct the array of simulations that will be needed to support the decision making processes. Interfacing geographic information systems (GIS) and nonpoint pollution modeling is a logical approach. The objective of the present study was to use the 37,000-acre area defined by the Kensington Quadrangle sheet in Montgomery County, Maryland, to show that GIS-supported nonpoint pollution modeling is practical and economically attractive. The purpose of the GIS is to estimate the spatial distribution of nonpoint nitrogen, phosphorous, zinc, lead, BOD, and sediment using a model developed by the Northern Virginia Planning District Commission. The system allows the user to change land uses in subareas to simulate the consequences of additional development or alternate management strategies. The tests show that in-house development of this type of special purpose GIS is a practical alternative to vendor supplied systems and that the required databases can be developed quite reasonably.     

ESTIMATING NONPOINT SOURCE POLLUTION LOADS WITH A GIS SCREENING MODEL1

ABSTRACT: The St. Johns River Water Management District (SJR-WMD) is using a Geographic Information System (GIS) screening model to estimate annual nonpoint source pollution loads to surface waters and determine nonpoint source pollution problem areas within the SJRWMD. The model is a significant improvement over current practice because it is contained entirely within the district's GIS software, resulting in greater flexibility and efficiency, and useful visualization capabilities. Model inputs consist of five spatial data layers, runoff coefficients, mean runoff concentrations, and stormwater treatment efficiencies. The spatial data layers are: existing land use, future land use, soils, rainfall, and hydrologic boundaries. These data layers are processed using the analytical capabilities of a cell-based GIS. Model output consists of seven spatial data layers: runoff, total nitrogen, total phosphorous, suspended solids, biochemical oxygen demand, lead, and zinc. Model output can be examined visually or summarized numerically by drainage basin. Results are reported for only one of the SJRWMD's ten major drainage basins, the lower St. Johns River basin. The model was created to serve a major planning effort at the SJRWMD; results are being actively used to address nonpoint source pollution problems.