|Risse, Mark - UNIV. OF GEORGIA, GA|
|Tiwari, A - UNIV. OF GEORGIA, GA|
Submitted to: International Soil Conservation Organization Conference Proceedings
Publication Type: Abstract Only
Publication Acceptance Date: May 23, 1999
Publication Date: N/A
Technical Abstract: WEPP, the water erosion prediction project, computer model envisages the recent advancements in the process of hydrology, plant sciences, soil physics and erosion mechanics. The model has the capabilities of predicting spatial and temporal distribution of net soil loss or gain for the entire hill slope for any time. Earlier studies depict that WEPP recorded a model efficiency of 0.71 in predicting average annual soil loss and that the model performs equally well when compared with the traditional methods such as USLE, the Universal soil loss equation and RUSLE, Revised universal soil loss equation. The phenomenon of overestimating the low values and under predicting the high values is inherent to all erosion models and WEPP is no exception to it. Research has indicated that errors in soil loss predictions from the soil erosion models are primarily influenced by the land management and topographical factors. The analysis taken up at 20 different sites in USA with 1600 plot years data reveals that model predicts better for the fallow plots and the plots with single crops accounting for model efficiencies of 0.79 and 0.71 respectively. The model efficiency sequentially drops in case of crop rotations, crop + grasses and permanent cover. This error might be attributed to the model over predicting the low values of soil losses in these cases. The model predictions vary slightly for the longer slope lengths, however, model predicts accurately for a wide range of slope and soil conditions. This paper presents an evaluation of the model accuracy and located potential sources of error.