Field Evaluation of Selected Soil Erosion Models for Catchment Management in Indonesia

Abstract

Three models, viz., areal non-point source watershed environment response simulation (ANSWERS), universal soil loss equation (USLE) and adapted universal soil loss equation (AUSLE) are evaluated for their performance under the field conditions of the Riam Kanan catchment in South Kalimantan province of Indonesia. While ANSWERS is evaluated for its accuracy to predict both runoff and soil loss, USLE and AUSLE are evaluated for soil loss only. The study was carried out in the context of sedimentation concerns for the Muhammad Nur Reservoir—an important source of drinking and irrigation water supply for the catchment. The models are evaluated using field data collected under four different land uses and during 2 years of field experiments. The land uses considered are cropland with minimum tillage, cropland with conventional tillage, grassland and areas reforested with rubber trees. The ANSWERS model in general has a tendency to overpredict runoff values. The ANSWERS model also was relatively better for predicting soil loss followed by the AUSLE and USLE models. Overall, the ANSWERS model proved superior for predicting soil loss in the Riam Kanan catchment. However, given that the AUSLE model produced sufficiently reliable results and is relatively easy to use, the AUSLE model would also appear to be a useful tool for predicting soil erosion in the catchment.