Application of an Erosion Model to a Diversified Agricultural Watershed

Abstract

Rural nonpoint sources of water pollution are becoming increasingly important in terms of cleaning up our nation's waters. However, the need for maintaining agricultural production to supply food for growing populations still exists. Agricultural runoff is the major nonpoint pollution source identified in the study area located in Marathon County, Wisconsin. Nutrient and sediment inputs from these lands have contributed to fish kills on the Eau Pleine Reservoir by causing algal blooms which contribute to low dissolved oxygen supplies. A mathematical computer model (ERODE) developed by the USDA-ARS was applied to the Big Eau Pleine watershed. Basis of data collected for model application included land use, rainfall, topography, soils and farm management practices. ERODE uses a modified form of the USLE (Universal Soil Loss Equation) to calculate overland soil movement based on kinetic energy from precipitation and overland flow for individual storms. A spring R Value (USLE) respresenting soil detachment energy of snow melt was derived from simulated snow melt events. The model was modified to handle varying land uses in the test watershed. Results indicate fair agreement between observed and simulated sediment yields. Improved simulation was experienced for the spring season by decreasing the infiltration component for the derivation of the K factor (USLE). The model shows potential as a planning tool for conservation considerations by attempting to fill the gap between soil movement on fields and sediment yield into the stream channel. It also provides a means of evaluating the impact of individual storms on soil loss and for identifying problem areas within the watershed. The model's need for complete information on precipitation and sediment yield will limit its use to areas where this information exists.