Effect of Digital Elevation Model (DEM) Resolution on the Hydrological and Water Quality Modeling

Abstract

The resolution of digital elevation model (DEM) used for watershed delineation and data extracted from the delineated watershed relevant to hydrography plays an important role in the calibration of hydrological models. The effect of DEM resolution on the drainage length, watershed area and average slope are studied in many earlier works. DEM resolution can have impact on hydrologic modeling parameters, water quality simulation results. The impacts of DEM resolution on results of the watershed response and water quality modeling are studied using HSPF (Hydrologic Simulation Program Fortran) model applied to a watershed in Kentucky in this study. Three major streams in this watershed are impaired due to pathogen pollution according to the current state of Kentucky water quality standard set for recreation. The effects of DEM resolution on water quality modeling and calibration for fecal coliform bacteria are studied using a HSPF model. Results from this study indicate that the DEM resolution affects hydrography parameters of the delineated watershed, influence the calibration process, and the non-point and point source loadings to the streams in the watershed. Introduction Digital elevation models (DEM) along with spatial analysis tools are commonly used for automatic delineation of watersheds as they are superior to manual delineation of watershed boundaries. The automatic derivation of topographic watershed data from DEMs is faster, less subjective, and provides more reproducible measurements than traditional manual techniques applied to topographic maps (Tribe, 1992). Vieux (2001) has shown using several examples that the resolution of DEM will affect the average slope of the watershed, drainage length and direction and streams. Maidment Copyright ASCE 2006 World Environmental and Water Resources Congress 2006 and Djokic (2000) provide a comprehensive review of digital elevation models and their use for hydrologic and hydraulic modeling. However, few studies (e.g., Islam, 2004) indicate that the accuracy of manual delineation using topographic maps in not lower than that of automatic delineation methods that use DEM. The main objective of this study is to assess the effect of DEM resolution on hydrological and water quality modeling. Digital Elevation Model (DEM) Resolution The resolution of digital elevation model (DEM) used for watershed delineation and data extracted from the delineated watershed relevant to hydrography plays an important role in the calibration of hydrological models. Furthermore, the water quality modeling of streams depends on the hydrologic features of the watershed that are obtained using DEMs. DEM resolution can affect hydrologic modeling parameters, water quality simulation results. The effect of resolution on the latter will influence water quality management strategies (e.g. total maximum daily load development) which quantify pollutant load reductions linked to non-point pollution sources within a watershed. The watershed area considered based on a specific DEM resolution will alter the land use composition and thereby ultimately affecting the non-point source loadings in the watersheds. The DEM resolution is known to have effect on simulation results related to hydrograph response and the fate and transport of water quality constituents in the streams. The impacts of DEM resolution on results of the watershed response and water quality modeling are studied using HSPF (Hydrologic Simulation Program Fortran) (Bicknell, et al., 1997) applied to a watershed in Kentucky. The watershed identified in the current study is South Elkhorn Creek watershed in Kentucky. Three major streams in this watershed are impaired due to pathogen pollution according to the current state of Kentucky water quality standard set for recreation. In order to study the effects of DEM resolution on water quality modeling and calibration, fecal coliform bacteria is modeled using HSPF model. The case study watershed provides an ideal hydrologic test unit to assess the influence of DEM resolution on water quality modeling. Hydrologic Modeling A lumped watershed simulation model, HSPF (Bicknell, et al., 1997) is used to simulate observed flows at a point of interest in the 8 digit USGS HUC (05100205) watershed in the state of Kentucky. The modeling steps adopted in the current study are shown in the Figure 1. Digital elevation models of resolutions 10 m and 90 m are obtained from BASINS web site (USEPA, 2004) and through Kentucky Geological Survey. Watershed delineation is carried out using two different DEM resolutions with the help of automatic delineation tool available under BASINS (USEPA, 2004) environment. The tool uses the conventional 8-point pour model (Peucker, 1978) to determine the flow paths and finally delineate the watershed. The hydrological Copyright ASCE 2006 World Environmental and Water Resources Congress 2006 parameters of are fine tuned and the HSPF model is calibrated using a trial and error approach. Figure 1. Modeling steps used for hydrologic calibration The hydrologic calibration of HSPF model involved initial estimates and subsequent adjustment of the appropriate model parameters (such as infiltration index capacity , lower zone evapotranspiration, lower zone soil moisture storage, fraction of ground water flow to deep recharge etc.) to reproduce the observed streamflow at the USGS Station. The initial estimates of the parameters are obtained from a report (USEPA, 2000). Rainfall data for use in the model was developed using hourly rainfall data obtained from regional NOAA weather station in Lexington, Kentucky which is nearest to the watershed of interest. Based on the available land use pattern details, four different land use patterns namely urban area, crop land (agriculture land and pasture land), forest land and barren land were used for the watershed modeling. Observed and simulated streamflow hydrographs were compared in each simulation and the essential parameters were revised in different trials until acceptable performance is achieved. Water Quality Modeling The Kentucky 2004 303(d) Report identifies a 17.6 mile segment of South Elkhorn Creek as not supporting the designated warm water uses of primary contact recreation (swimming) and aquatic life due to pathogen impairment. A stream sampling system was developed for the watershed and representative tributaries within the watershed area. Sampling occurred between May 2002 and October 2002. Sampling was done at various sites within the South Elkhorn Creek watershed during the monitoring Digital Elevation Model (10 m or 90 m resolution) BASINS Modeling Environment (Automatic Delineation Tool) Delineated Watershed Hydrologic Features HSPF Model Calibration and Validation Copyright ASCE 2006 World Environmental and Water Resources Congress 2006 period. In an effort to account for the potential impact of point and non-point pathogen sources, water quality samples were collected during wet and dry events. Preliminary data suggests that more than 90% of the time, pathogen values in South Elkhorn (and its tributaries) exceed the 30-day geometric mean limit set forth by the EPA, (i.e. 200cfu/100ml) for full contact recreation. In order to assess the sources and associated pathogen loadings in the South Elkhorn Creek watershed, the calibrated HSPF simulation model is the used to model the fecal coliform bacteria. The model is used to determine the daily fecal coliform concentrations by incorporating point and non-point source loads. The loads were sub-divided into distributed loads associated with the two main land uses in the watershed (i.e. agricultural and urban) and concentrated loads (associated with failing septic systems, straight pipes, and instream loads associated with livestock). Details of the point and non-point sources, sampled data relevant to the watershed are provided elsewhere (KWRRI, 2004). The loading intensities for different land use types are obtained from USEPA (2001) and using Bacterial Indicator Tool (BIT). Details of Case Study Area South Elkhorn Creek, a fourth order stream, originates in northwest Jessamine County and southwest Fayette County. The creek flows northwest to merge with the North Elkhorn Creek forming the Elkhorn Creek in Franklin County. South Elkhorn Creek contributes to the Kentucky River Watershed, HUC # 05100205. The location of streamflow gaging station and case study area are shown in the Figure 2. The USGS streamflow observations from gage # 03289000 on South Elkhorn creek at Fort Springs is used for calibration of the HSPF model. The location of the gaging station is used as an outlet for delineation of the watershed that drains into this point. Figure 2. Map showing the location of streamflow gaging station and the 8-digit USGS HUC Copyright ASCE 2006 World Environmental and Water Resources Congress 2006 Results and Analysis The location of USGS gauging station is used as an outlet for the delineation of watersheds using 10 m and 90 m DEMs. The delineated watersheds using two different resolutions are shown in the Figure 3. It is evident from Figure 3 that the boundaries of the delineated watersheds are different based on 10m and 90m digital elevation model. Major watershed parameters derived using the DEMs are given in the Table 1. The four major land use types in the watershed and their corresponding areas are provided in the Table 2. The total watershed areas differ by 6% and a major difference is seen in the agricultural areas of the watersheds.