Study On Evaluation And Analysis Method Of Regional Soil Erosion Intensity

This paper takes Lianhua County Industrial Park in Jiangxi Province as the research area,according to the relevant standards of the Ministry of water resources,this study uses the research ideas of theoretical model establishment,algorithm compilation and field application to study the evaluation process of soil erosion intensity in Lianhua county.The existing DEM generation method is improved to some extent.Combined with 3S technology,using multi-source remote sensing impact data,the monitoring elements of soil erosion are extracted,and the soil erosion intensity in Lianhua county is estimated based on RUSLE model.The effectiveness and objectivity of the theory and algorithm are verified by comparing the field application results with the current methods.The main conclusions are as follows:(1)A seed generation method based on Delaunay grid and distance function is proposed.The main idea of the algorithm is to generate the initial seed in a given area,generate the outer seed through the front edge line,and generate particles to supplement the gap.There are three seed generation methods:fixed mode,random mode and fully random mode.After generating the seed,Delaunay grid is generated for the central coordinates of all particles by distance function.The proposed algorithm is compiled into a running module and embedded in ArcGIS,which can be better applied.The hillside in Lianhua county is selected for field verification and compared with the field measurement results.The error value is within 30%.(2)A method of soil erosion information extraction based on RS technology is proposed.Based on the object-oriented classification method,mrmr feature selection algorithm is combined with SVM forward algorithm to select feature variables,build a decision tree to obtain classification rules and interpret images;For high-resolution images,only four types of end elements,namely vegetation,bare land,impervious surface and water,are selected to extract the vegetation coverage in the study area by using the linear mixed model method.(3)The soil erosion intensity of Lianhua Industrial Park in Jiangxi Province was evaluated.Collect the elevation data of Lianhua Industrial Park,generate DEM by using contour lines and skeleton points,and extract the slope length factor.Using 3S technology,other factors in RUSLE equation are extracted to evaluate the intensity of soil erosion in the industrial park.The data of the industrial park in previous years are collected to generate the soil erosion grade distribution map,analyze the soil erosion in the park,and analyze the relationship between the soil erosion modulus and the slope and vegetation coverage,so as to provide some technical support for the development of water and soil conservation in the park(4)The soil erosion degree of Hualien Industrial Park was estimated by comparing the current method with the seed generation method.The moderate,mild and slight soil erosion areas accounted for 23.7%,26.8%and 33.8%of the total area respectively;The area of moderate and above soil erosion only accounts for 15.2%,of which the area of intense erosion accounts for 12.5%,and the area of extremely intense erosion and intense erosion only accounts for 3.0%and 0.2%respectively.The soil erosion intensity grade calculated based on the seed generation method shows that the Industrial Park accounts for 16.3%,23.5%and 30.2%of the total area with strong,moderate and mild erosion respectively;The area of extremely strong and severe erosion is also large,accounting for 6.8%and 9.9%respectively,accounting for 15.7%of the total area,;The area of micro erosion is 13.3%.