Assessment of gully erosion susceptibility using different DEM-derived topographic factors in the black soil region of Northeast China

作     者：Donghao Huang Lin Su Lili Zhou Yulu Tian Haoming Fan Donghao Huang;Lin Su;Lili Zhou;Yulu Tian;Haoming Fan

作者机构：College of Water ConservancyShenyang Agricultural UniversityShenyang110866China College of ForestryShenyang Agricultural UniversityShenyang110866China Key Laboratory of Soil Erosion Control and Ecological Restoration inLiaoning ProvinceShenyang110866China College of Urban and Environmental SciencesNorthwest UniversityXi'an710127China 

出 版 物：《International Soil and Water Conservation Research》 (国际水土保持研究（英文）)

年 卷 期：2023年第11卷第1期

页      面：97-111页

核心收录：

学科分类：0830[工学-环境科学与工程（可授工学、理学、农学学位）] 082802[工学-农业水土工程] 090707[农学-水土保持与荒漠化防治] 0907[农学-林学] 08[工学] 0828[工学-农业工程] 09[农学] 0903[农学-农业资源与环境] 0815[工学-水利工程] 

基　　金：supported by the National Key R&D Program of China[Grant number 2021YFD1500700] the National Natural Science Foundation of China[Grant number 42007050] Scientific Research Foundation of Liaoning Province[Grant number LSNPT202002] 

主　　题：Gully erosion Machine learningmethods Topographicattribute Pixel size Northeast China 

摘      要：As a primary sediment source,gully erosion leads to severe land degradation and poses a threat to food and ecological ***,identification of susceptible areas is critical to the prevention and control of gully *** study aimed to identify areas prone to gully erosion using four machine learning methods with derived topographic *** topographic attributes(elevation,slope aspect,slope degree,catchment area,plan curvature,profile curvature,stream power index,and topo-graphic wetness index)were derived as feature variables controlling gully occurrence from digital elevation models with four different pixel sizes(5.0 m,12.5 m,20.0 m,and 30.0 m).A gully inventory map of a small agricultural catchment in Heilongjiang,China,was prepared through a combination of field surveys and satellite *** topographic attribute dataset was randomly divided into two portions of 70%and 30%for calibrating and validating four machine learning methods,namely random forest(RF),support vector machines(SVM),artificial neural network(ANN),and generalized linear models(GLM).Accuracy(ACC),area under the receiver operating characteristic curve(AUC),root mean square error(RMSE),and mean absolute error(MAE)were calculated to assess the performance of the four machine learning methods in predicting spatial distribution of gully erosion susceptibility(GES).The results suggested that the selected topographic attributes were capable of predicting GES in the study catchment area.A pixel size of 20.0 m was optimal for all four machine learning *** RF method described the spatial relationship between the feature variables and gully occurrence with the greatest accuracy,as it returned the highest values of ACC(0.917)and AUC(0.905)at a 20.0 m *** RF was also the least sensitive to resolutions,followed by SVM(ACC=0.781-0.891,AUC=0.724-0.861)and ANN(ACC=0.744-0.808,AUC=0.649-0.847).GLM performed poorly in this study(ACC=0.693-0.757,AUC=0.608-0.703).Ba