Submitted to: Proceedings of the American Society of Agricultural and Biological Engineers International (ASABE)
Publication Type: Proceedings
Publication Acceptance Date: July 15, 2011
Publication Date: September 18, 2011
Citation: Dun, S., Wu, J.Q., Elliot, W.J., Frankenberger, J.R., Flanagan, D.C., McCool, D.K. 2011. Applying online WEPP to assess forest watershed hydrology [abstract]. In: D.C. Flanagan, J.C. Ascough II, and J.L. Nieber, editors. Proceedings of the International Symposium on Erosion and Landscape Evolution (ISELE). ISELE Paper No. 11085. American Society of Agricultural and Biological Engineers, September 18-21, 2011, Anchorage, AK. 711P0311cd. Technical Abstract: The U.S. Army Corps of Engineers (USACE) and the Great Lakes Commission are developing technologies and predictive tools to aid in watershed management with an ultimate goal of improving and preserving the water quality in the Great Lakes Basin. A new version of the online Water Erosion Prediction Project (WEPP) GIS interface has been developed to assist in evaluating sediment sources associated with forests and forest management within the Basin. The online WEPP GIS interface uses the OpenLayers and MapServer GIS software with base image data from Google. WEPP inputs for watershed applications, including digital elevation data (USGS 30-m National Elevation Data), land cover maps (USGS land cover), and soil maps (NRCS SSURGO soils), are automatically retrieved from web servers. DEM data are used to generate WEPP watershed structure and topographical inputs for each watershed element. Landuse and management files in the WEPP database are linked to polygons in the land cover map, and WEPP soil input files are generated on the server for the area of interest by querying the NRCS soil database. For areas where NRCS SSURGO data are not available, soil files in the WEPP database are linked to the polygons of the land cover map. Surface cover and soil properties of the WEPP management file and soil file can be customized to represent site-specific conditions, and functions to substitute the default inputs with user-specified landuse or soil files for a sub-catchment are provided. Daily climate inputs are generated from the long-term climate parameters of the nearest weather station using CLIGEN. Monthly climate parameters can be adjusted by using the gridded data of PRISM to account for locations distant from CLIGEN stations. No instrumented forested watersheds with long-term observation data within the Great Lakes Basin were available for testing the online WEPP GIS interface. However, within the Fernow Experimental Forest, West Virginia, there are seven watersheds with more than 50 years of observed climate and streamflow data available online. Among them, WS1 and WS4 have been completely covered with mature forest for the last two decades and were therefore selected for applying the online WEPP GIS interface. The most recent 20 years (1987–2007) of available observation data for WS1 and WS4 were acquired and used for WEPP simulation. The Penman-Monteith method in the WEPP model was used for ET simulation and a crop coefficient for growth season in the FAO Penman-Monteith equation was set to 0.71 to reproduce the 640 mm of the average annual ET observed at the Fernow Experimental Forest. A site-specific WEPP soil file was created by combining the automatically generated WEPP soil file from SSURGO database with the soil file for mature forest in the WEPP database and adding a restrictive layer to represent the bedrock beneath the soil profile. The saturated hydraulic conductivity of the restrictive layer was fitted through desktop WEPP simulation using observed precipitation and air temperature as climate inputs to reproduce observed average annual runoff. Online WEPP simulations were carried out and the model performance was examined by comparing simulated and observed streamflow data and sediment yield derived from the observed stream water chemistry data. The online WEPP model reasonably simulated average annual runoff and the annual maximum runoff series for both watersheds. Sediment yield was slightly under-predicted for annual average and slightly overpredicted for the annual maximum series. In addition, the simulation results adequately reflected the differences between WS1 and WS4 in their hydrological characteristics. Future studies assessing the performance of the online WEPP GIS interface in simulating different hydrologic, landuse, and management conditions are needed.