GIS-Based ER-USLE Model to Predict Soil Loss in Cultivated Land

Abstract

In the Dabie Mountains of the lower Yangtze basin, China, soil loss reduction has been deemed as important as flood control after the 1998 flood, which caused great damage to the area and its residents. For soil loss control in the Dabie Mountains, China, the first step is the determination of appropriate strategies. An urgent need exists for soil loss models that can provide sufficient information for developing strategies to control soil loss using field observation data over several years. In this study, field observations of water and sediment discharges were carried out at the micro-plot scale and the USLE (universal soil loss equation)-plot scale in the Shangshe catchment of the Dabie Mountains in China. Through analyses of field observation data in the Shangshe catchment during the period 1999–2007, the USLE model was applied to predict both annual soil loss and that for single events based on GIS in a sub-catchment of cultivated land. Model calibration showed that by using an effective rain erosivity factor (R e) instead of R (rain erosivity) in the RUSLE, and by portraying the interaction among seasonal precipitation, seasonal crop coverage (C s), and practice of discrete-event practices by human beings (P s) in the USLE, an ER-USLE model was developed and model accuracy improved. With these measures, the ER-USLE model can be utilized to predict annual soil loss based on single events from field observation data over a few years. Comparison between predicted results and observed results of the field observation data of soil loss from 2000 to 2007 in the sub-catchment of cultivated land scale with USLE and ER-USLE models shows that the R 2 of ER-USLE (0.82) is much higher than that (0.46) of the USLE. With ER-USLE and GIS, strategies for soil loss control are possible based on seasonal soil loss information with field observation data over a few years.