Transactions of the ASAE Vol. 48(1): 109-120 2005 American Society of Agricultural Engineers ISSN 0001-2351 109 EFFECT OF DEM RESOLUTIONS IN THE RUNOFF AND SOIL LOSS PREDICTIONS OF THE WEPP W ATERSHED MODEL T. A. Cochrane, D. C. Flanagan ABSTRACT. Erosion prediction utilizing digital elevation models (DEMs) is a logical advancement for automating the simula- tion process for models such as the Water Erosion Prediction Project (WEPP). The effects of using different DEM resolutions on watershed simulations and the ability to accurately predict sediment yield and runoff from different rainfall event sizes were studied using three application methods and data from six research watersheds. Simulating watersheds with a range of resolutions can help address the problem of deciding what topographic DEM resolution is ideal for model simulations of the watershed outlet, the end of each hillslope, and along the slope profiles. The three application methods studied here were: (1) Hillslope - Chanleng, (2) Hillslope - Calcleng, and (3) Flowpath. The two Hillslope methods use a representative slope profile to represent each hillslope in the watershed, and the Flowpath method uses all of the individual flowpaths as model input for WEPP simulations. Results show that the Hillslope methods were not significantly influenced by DEM resolutions; however, there was an observable interaction between resolutions and the Flowpath method. Large rainfall events were pre- dicted better than small events, but fine DEM resolutions did not improve predictions of either large or small rainfall events. Using coarse DEM resolutions for the topographic input will not decrease the accuracy of erosion prediction using the WEPP model and the Hillslope methods, unless the coarseness of the DEM compromises the delineation of the watershed or hill- slopes. Keywords. Digital Elevation Models, Geographic information systems, Soil erosion modeling, Water Erosion Prediction Project. n past studies by Cochrane and Flanagan (1999), two general methods have been developed to integrate digi- tal elevation models (DEMs) with the Water Erosion Prediction Project (WEPP) erosion model (Flanagan and Nearing, 1995). These methods have been named the Hillslope methods and the Flowpath method. The two Hill- slope methods (Chanleng and Calcleng) consist of the discre- tization of the watershed into representative hillslopes and channels from a DEM (Cochrane and Flanagan, 2003). In the Hillslope approaches, a channel network is extracted from the DEM using the concept of a critical source area (Gar- brecht and Martz, 1997). Hillslopes are then defined as the areas that drain to the right, left, or top of each of the channel segments. A representative hillslope profile is also created for each hillslope from the DEM. The actual representative hillslope profile created by the Chanleng and Calcleng meth- ods is derived in the same way; however, they differ in how the representative hillslope length is calculated. In the Calc- leng method, a representative hillslope length is calculated by a method of weighting flowpath lengths and flowpath drainage areas (Cochrane and Flanagan, 2003). This same weighting procedure is used for hillslopes draining to the top Article was submitted for review in July 2003; approved for publication by the Soil & Water Division of ASAE in October 2004. The authors are Thomas A. Cochrane, Agricultural Engineer, AGTECA S.A., Santa Cruz, Bolivia; and Dennis C. Flanagan, ASAE Member, Agricultural Engineer, USDA-ARS National Soil Erosion Research Laboratory, West Lafayette, Indiana. Corresponding author: Dennis C. Flanagan, USDA-ARS NSERL, 1196 Soil Bldg., 275 S. Russell St., West Lafayette, IN 47907; phone: 765-494-4478; fax: 765-494-5948; e-mail: flanagan@purdue.edu. of channels in the Chanleng method; however, the lengths of representative hillslopes are calculated differently for hill- slopes draining to the sides of channels. In the Chanleng method, for hillslopes adjacent to a channel, the representa- tive hillslope width is set to equal the channel length and the hillslope length is calculated by dividing the total area of the hillslope by its set width. WEPP is then applied to the hill- slopes and channel structure in a watershed model simula- tion. The Flowpath method consists of applying WEPP to all possible flowpaths in the watershed. Flowpaths are defined in terms of DEM grid cells starting at a point where no other cell in the grid flows into it, and then following through a path defined by individual grid flow vectors, and ending when it reaches a channel. Interactions with other flowpaths are frequent, and soil loss results from the application of WEPP to each flowpath are weighted with the interactions of other flowpaths. However, since there are many flowpaths draining at distinct points along the channels, the WEPP channel routines cannot be used. The current WEPP channel routines are limited in the number of channel segments that can be simulated and limited conceptually in the way runoff and sediment entering from hillslopes are distributed along the channel. Consequently, WEPP simulation results for the Flowpath method can only show sediment loss and runoff from each hillslope or the spatial distribution of soil loss and deposition across all flowpaths in a watershed. The two Hillslope methods and the Flowpath method predicted runoff and sediment yield comparable to measured data for the six watersheds studied, using DEMs of the finest resolution available (Cochrane and Flanagan, 1999, 2003). It is now important to determine if these methods are accept- I