Modeling Soil Erosion in Chinese Loess Plateau using Cellular Automata

Abstract

Chinese Loess Plateau has the most extensive soil erosion in the world, which results from the fragile geophysical environment and human activities. The soil erosion has shown more and more severe trends due to the increasing intensity of human disturbances. Soil erosion is a potential cellular automata (CA) application field, because it can be considered as a system evolving exclusively by means of local interactions. The developed CA model in this study involves a larger number of states in comparison to the previous models, including altitude, water depth, vegetation density, infiltration, erosion volume, sediment transport and deposition. The model was applied to a small catchment in Ansai county, Chinese Loess Plateau. Landsat ETM data was used to obtain the vegetation and land use information, DEM and altitude can be created and obtained from the digital topographical maps, respectively, soil types from soil map, and the other necessary data by means of field work. The results indicated that the soil erosion can be well modeled by CA. From the modeling process, the driving factors of soil erosion can be analyzed.