Assessment On WEPP Model (Hillslope Version) Applicability To Hilly-Gully Region On The Loess Plateau

Soil erosion prediction model may provide a powerful and scientific guide to soil erosion assessment and soil and water conservation planning. At present, there is little information for assessing WEPP model application on the steep landscape in the world, especially there is a gap for assessing applicability of WEPP model at different crop and grass condition in China. This thesis used field observed data of runoff and soil loss from 15 runoff plots under 62 single rainfall events in 1985-1992 in Ansai Experimental Station to assess the applicability of WEPP model to the hilly-gully region of the Loess Plateau, the research results not only broaden the scope of the WEPP application, but also provided a idea and method for developing Chinese soil erosion prediction model. The main results were as follows:1. Three files of climate, slope gradient and length, and crop management for runing WEPP model were established. Based on analysis of sensitivity between soil parameters and runoff and soil loss, effective hydraulic conductivity, soil critical shear stress, and rill erodibility of three soil parameters were calibrated. For different slope gradients with bare and fallow surface, the calibrated values of effective hydraulic conductivity, soil critical shear stress, and rill erodibility were 19.3mm.h-1, 2.6 Pa, and 0.025 s.m-1, respectively. For different slope lengths with bare and fallow surface, the calibrated values of the above three soil parameters were 22.1mm.h-1, 3.5 Pa, and 0.022 s.m-1, respectively. For different crops and grasses, the average values of the above three soil parameters from different slope gradients and lengths were taken. Testing calibrated results used Nash-Sutcliffe model efficiency (ME) found that calibrated results were reasonable.2. The application of WEPP model at different slope gradients from 5-28°and slope lengths from 10 to 40 m with bare and fallow surface were assessed. The results showed that WEPP model has high simulated accuracy for event runoff and annual runoff, and event soil loss and annual soil loss. ME of the event runoff and annual runoff, and event soil loss and annual soil loss were more than 0.76. Simulated values of average annual runoff and soil loss were accepted, relative errors of average annual runoff were 5%-6%, relative errors of average annual soil loss were 18%-21%. I30, PI30 and∑PI30 had a great effect on simulated results by WEPP model. In addition, comparing to measured values, the changes of simulated runoff with different slope gradients and lengths were not sensitive, but it were sensitive for soil loss simulation.3. The application of WEPP model at different crops of soybean, buckwheat, millet and potato and red clover were assessed. The results showed that ME of event and annual runoff and soil loss at different crops reached to 0.51; related errors of average annual runoff and soil loss were 31.24% and 23.04%, respectively; which demonstrated that WEPP model could be used to simulate runoff and soil loss from steep croplands. ME of event and annual runoff for hillslope with red clover was 0.54-0.74, but ME of event and annual soil loss for hillslope with red clover was only 0.11-0.26; which showed that for the hillslope with red clover cover, WEPP model has high simulated accuracy for event runoff and annual runoff, but has lower simulated accuracy for event soil loss and annual soil loss. Related errors of average annual runoff and soil loss were 4.06% and 21.85%, respectively, which showed simulated results by WEPP model were accepted.4. Compared to observed data, storm duration generated by CLIGEN4.3 and CLIGEN5.22563 were longer for short-duration storm and shorter for long-duration storm. The ability of generating rainfall and relative peak rain intensity generated by CLIGEN5.22563 was better than those by CLIGEN4.3. Runoff and soil loss simulated by WEPP model based on CLIGEN4.3 and CLIGEN5.22563 were over predicted.5. Among three climate models of BPCDG with measured climate file, CLIGEN with measured climate file, and CLIGEN with generated climate file, simulated results by WEPP model based on BPCDG were best; the next one was CLIGEN with measured climate file; and simulated results by CLIGEN with generated climate file were worse. Average annual runoff simulated by WEPP based on above 3 climate files was better than average annual soil loss.6. Soil loss in future 40 years at different landuse under A2, B2 and GGa1 3 emission scenarios based on GCM and WEPP model was predicted, the results showed that compare to current background, precipitation in 2049 under A2, B2 and GGa1 3 emission scenarios could increase by 37%, 22%, and 12%, respectively. For the fallow and bare hillslope, runoff and soil loss in 2049 will increase by 100% and 70% under A2 emission scenario, could reduce by 25% and 30% under B2 emission scenario, and could increase by 7% and 20% under GGa1 emission scenario. For hillslope with crop cover of buckwheat, soybean, millet, and red clover, runoff and soil loss in 2049 could reduce under A2, B2 and GGa1 3 emission scenarios; runoff could reduce by 30%-90%, and soil loss could decrease by 77%-100%. For hillslope with crop cover of potato, runoff and soil loss could reduce by 51%.