Font Size: a A A

Space-time Evolution Simulation Of Soil Erosion Based On CA And ANN

Posted on:2011-06-29Degree:MasterType:Thesis
Country:ChinaCandidate:F LiFull Text:PDF
GTID:2143360305463609Subject:Physical geography
Abstract/Summary:PDF Full Text Request
Soil erosion forms serious environmental, ecological and safety problems. The model of soil erosion study this problem scientifically, quantitatively and effectively. The result of simulation will provides informatin and decision for soil and water conservation and management.This research establishs the ANN-CA model of soil erosion based on GIS, CA, ANN, applicationed in the soil erosion area and analyzed rainfall, topography, vegetation, soil, human activity and other factors.This study includs:(1) The establishment of the model:This model is based on Universal Soil Loss Equation (USLE) model and then is improved. We analyze every parameter in the model and try to complete them; Using ANN to find the probability about variety of types in soil erosion and define the rules of CA. Afterwards we establish the soil erosion model of ANN-CA.(2) The integration of the model with GIS:By GIS, the model can read, save, process and output data which the model requires. This model intergrates the function of data processing, spatial analysis, visualization based on GIS and the space-time dynamic simulation function based on CA. It can simulate space-time dynamic evolution of soil erosion.(3) The application of the model in study area:Apply the model on a case study of Lishui River Basin, control the simulation precision of the soil erosion model by validation and improvement, and then realize the quantitative research of spatial distribution and the dynamic simulation of space-time evolution. Afterwards analyze the simulation results and propose control measures.The research gets conclusion after space-time simulation based on the ANN-CA model:(1) Soil erosion has a complex nonlinear characteristic. CA and ANN can effective to simulate the procession of soil erosion. (2) The stucture of the CA model based on ANN has simple code, easy to design, less parameters and high precision. (3) The result of prediction in study area at 2015 reflects three rules:In space, the soil erosion intensity in Lishui Rive Basin has obviously similarity in whole area and clearly difference between distinct areas. The predictions facilitate human prevention with appropriate method. In time, the soil erosion intensity in Lishui Rive Basin has been effectively controlled. The predictions facilitate water and soil conservation, provention of soil erosion and division. Overall, change the sloping fields to terrace (slope less than 25°) and change the arable land (dry land, the slope is greater than 25°) to woodland or lawn.
Keywords/Search Tags:CA, ANN, GIS, Soil Erosion, Lishui River Basin
PDF Full Text Request
Related items