Submitted to: Transactions of the ASAE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 7/28/1997
Publication Date: N/A
Citation: Interpretive Summary: Simulating the spatial variability of a watershed requires the selection of many subdivisions, each describing a homogeneous area. This study was performed to evaluate the influence the number of subdivisions within a watershed has on the quality of runoff and sediment yield, and to evaluate techniques to determine an appropriate subdivision level. A topographic analysis and subwatershed generation program was used in conjunction with a sediment yield model to simulate a watershed containing mixed landuse conditions. Runoff volume was not affected by the subdivision level, but sediment yield was highly dependent on the subdivision level. Sediment yield increased with increasing subdivisions mainly because the description of cultivated landuse and land slope also increased. Cultivation produces the most erosion among landuse practices defined for the watershed and erosion increases with increasing slope. There is a level where increasing the number of subwatersheds reduces the rate of increase of sediment yield with succeeding subdivision levels. This point may provide an optimal level to reduce the computational and input parameter requirements of the model. This information is important for efficient use of models used for the selection of best management practices to control erosion.
Technical Abstract: The objective of this study was to evaluate the capability of the Soil and Water Analysis Tool (SWAT) erosion model to adequately simulate runoff and sediment from Goodwin Creek Watershed using a few to several hundred watershed subdivisions. Runoff volume was not affected by the number of subwatersheds. However, a minimum number of subwatersheds was required to adequately simulate sediment yield. Increasing the number of subwatersheds beyond a threshold of about 227 did not substantially affect the computed sediment yield. The proper identification of this threshold number of subwatersheds for other watersheds can reduce input parameter and time requirements to run the model and analyze the results. An evaluation of the actual stream network, digitized from USGS topographic maps, with the generated stream network demonstrated that comparing the two for a determination of an adequate subdivision level does not provide the information needed, but can be used as a starting point. This study emphasizes that a proper selection of the number of subwatersheds is needed to evaluate best management practices used for erosion control.