Monday, November 13, 2006 - 1:50 PM

Water Erosion Estimation in Topographically Complex Landscapes: Evaluation Using Field Data Sets.

Carlos A. Bonilla, Pontificia Univ Catolica de Chile, School of Agriculture and Forestry, Av. Vicuņa Mackenna 4860, Santiago, 6904411, Chile, John Norman, UW, Madison, 1525 Observatory Dr, 163 Soils Bldg, Madison, WI 53706-1299, and Christine Molling, CIMSS/SSEC, 1225 W Dayton St, Madison, WI 53706.

Soil erosion can vary significantly over short distances as a function of local soil properties and micro-topography; but regardless of this, many erosion models assume homogeneity in topography and soil characteristics. The objective of this research was to develop and evaluate a method for estimating soil loss from agricultural fields that is faithful to the complex topography and spatial heterogeneity common to managed landscapes. Sediment loss for individual storms was achieved by linking soil detachment/deposition equations adapted from the Water Erosion Prediction Project (WEPP) model to the existing water-flow subroutine in the Precision Agricultural-Landscape Modeling System (PALMS). PALMS works on a grid-cell size of 5-20 m and simulates runoff and soil erosion patterns as affected by soil consolidation, soil texture, anisotropic surface roughness, antecedent moisture, canopy cover, slope gradient, and tillage interactions with topography. To evaluate the overall performance of PALMS under cropped conditions, the runoff and soil loss were collected in three farm fields with complex topographies. A total of 46 events were selected, and 3 cropping scenarios from the three locations were used. The duration of the field measurements was two years. Predictions of runoff and soil loss from PALMS were compared with the field measurements, storm by storm, for an entire cropping season. Sediment loss predictions from PALMS were compared with measurements using two approaches: matching the PALMS runoff predictions to runoff measurements by adjusting the soil water content in the model prior to rain initiation (so that sediment equations could be evaluated without confusion from runoff prediction errors), and using a continuous simulation of runoff and sediment loss with no adjustment for initial conditions. The accuracy in the runoff and soil loss predictions observed in this study show that PALMS is a useful tool for predicting runoff and soil loss for agricultural fields with complex topographies.