GIS Assisted Modeling Of Soil Erosion And Hydro-geomorphic Processes In The Zuli River Basin

Soil erosion is a hydro-geomorphic process occurring over the earth's surface. At present, Accelerated soil erosion induced by human activities is of global significance, threatening agricultural economies around the world, especially in the semiarid areas. The Chinese Loess Plateau is an extreme example of human-induced soil erosion and the resulted economic sufferings. This research focuses on the erosion problem of ZuLi River basin in the western part of the Chinese Loess Plateau using GIS-assisted soil erosion model. The major findings of this research can be summarized as follow:1) The comparison of USPED model-estimated amounts of soil erosion based on four different scales of DEMs (1:250,000, 1:50,000, 1:10,000, and 1:5,000) indicates that 1:50,000 -scale DEM is almost as good as higher-resolution DEM data in providing needed topographic information for assessing soil erosion, whereas 1:250,000-scale DEM underestimates soil erosion mainly because rill erosion and sheet erosion are greatly underestimated.2) Since precipitation is the most important forcing factor for soil erosion, the accuracy of soil-erosion estimates understandably relies on the adequacy of spatially distributing point-observed precipitation data. Our experiments show that first-order Ordinary Kriging model can provide the best approach to spatially distributing the annual precipitation throughout Zuli River basin, and the spatial estimates are further improved when elevation factor is taken into consideration.3) The observed runoff production and sedimentation rates within Gaoquan catchment (a small watershed within Zuli River basin) show that different land uses have different impact on runoff produciton and soil erosion. Under the same physiographic condition (e.g., precipitation and topography), runoff varies with land uses in the following decreasing order of intensity: wasteland > cropland > fallow > grassland > woodland (low cover) > shrub. However, the dependency of soil erosion on land uses is different from that of runoff Specifically, soil erosion varies with land uses in the following decreasing order of severity: cropland > wasteland > grassland > woodland (low cover) > shrub. It simply implies that soil erosion and runoff can be modulated through land use planning. Furthermore, different croplands (e.g., wheat, pea, potato,etc) generate different responses of runoff and soil erosion. The observed data within the Gaoquan Catchment demonstrate a nonlinear response of runoff production and soil erosion to different degrees of topographic slopes under cropland conditions. The data also demonstrate that concave slope is more pro runoff production and soil erosion than other slope shapes (convex, S, linear) under grassland (alfalfa) conditions. The observed data further confirmed the previously-reported finding that runoff production and soil erosion are highly dependent on vegetation coverage density under grassland conditions and the dependency increases within increasing precipitation intensity.4) Comparison between LISEM simulation and the observed data for 1992/08/29 and 1993/07/20 rainfall events in Gaoquan Catchment shows that LISEM can be adapted for estimating runoff production and soil erosion for the Loess Plateau and the model-estimation is even better for high rainfall intensity conditions. Sensitivity analysis of LISEM indicates that higher initial soil moisture and lower saturate conductivity showed a large sensitivity.5) We have also used SEMMED model, a regional scale soil erosion model, to provide useful information for decision-makers and planners to take appropriate land management measures for soil and water conservation. Comparison between the observed data (1999) from four hydrological stations (Chankou, Huining, Guochengyi and Jingyuan) with SEMMED simulated result within Zuli River basin indicates that SEMMED simulation can be adapted for estimating runoff production and soil erosion of watershed scale after some parameters are adjusted using the observed data. The simulation results for 1999 and 2001 indicate that soil erosion is generally proportional to total precipitation. However, higher vegetation coverage generally resulted from higher total precipitation definitely results in lower runoff production and soil erosion.