Font Size: a A A

Construction Of Soil Erosion Intensity Assessment Model In Lalin River Watershed Of Heilongjiang Province

Posted on:2015-02-18Degree:MasterType:Thesis
Country:ChinaCandidate:N ZhouFull Text:PDF
GTID:2283330434951116Subject:Soil and Water Conservation and Desertification Control
Abstract/Summary:PDF Full Text Request
Calculated soil erosion modulus of Lalin river watershed of Heilongjiang province by USLE,and established the spatial raster database based on GIS.Used soil erosion modulus as discriminant conditions,validated each applicability of soil erosion prediction model built based on logistic regression and RBF neural network, and then built and validated the improved model-soil erosion prediction model based on LOG-RBF neural network. The results showed that:Logistic regression model had obvious advantages of determineing the land whether the occurrence of soil erosion, predictive accuracy of not occurring and occurring was77.4%and97.9%respectively, the total predictive accuracy was94.9%.RBF neural network model had the stronger ability to estimate soil erosion modulus, relative error and error sum of squares of the simulation results was0.612%and13.292respectively,and R was0.57.Relative error and error sum of squares of the simulation results was lower0.157%and2.601respectively based on LOG-RBF neural network model than RBF neural network model,and R2was0.82, so LOG-RBF neural network model had a better fitting degree,with soil erosion modulus increased misjudge phenomenon demonstrated a downtrend. Determined by ROC curve,the value of AUC based on LOG-RBF neural network model was larger0.063than RBF neural network model,and was more better. This model could be used to estimate soil erosion modulus, and can be combined with GIS,predicted the spatial distribution situation of soil erosion.Analysed the spatial characteristics of soil erosion modulus by geostatistical method. Contrasted two kinds of current classification standard of soil erosion intensity,and analysised to determine which was more applicable. Analysed the spatial characteristics of soil erosion intensity by the space and trend analysis method. The results showed that:Distribution of soil erosion modulus in the study area had strong spatial correlation, and the direction was from west to east, was from north to south had a linear increase trend.The classification standard of soil erosion intensity in the 《Techniques standard for comprehensive control of soil erosion in the black soil region》 was more applicable. Soil erosion area had amounted to94%of the study area, strong, mild and moderate hydraulic erosion was the major type of erosion intensity. Strong hydraulic erosion was mainly distributed in the western which was the diluvial platform, mild hydraulic erosion was mainly distributed in the eastern which was the hills and wide valley, moderate hydraulic erosion was mainly distributed in the central which was transition zone of landforms, extremely strong and severe hydraulic erosion showed a scattered distribution in the hills and gully of eastern,and had a sporadic distribution in the central and western. With the decrease of vegetation coverage, classification of soil erosion intensity showed a trend of increase from east to west.
Keywords/Search Tags:Lalin River basin, soil erosion intensity, Logistic regression, Radical basis functionneural network, ULSE
PDF Full Text Request
Related items