Utilizing geodetectors to identify conditioning factors for gully erosion risk in the black soil region of northeast China  

作　　者：Donghao Huang Xinrui Zhao Zhe Yin Wei Qin 

机构地区：[1]College of Water Conservancy,Shenyang Agricultural University,Shenyang,110866,China [2]Key Laboratory of Soil Erosion Control and Ecological Restoration in Liaoning Province,Shenyang,110866,China [3]State Key Laboratory of Simulation and Regulation of Water Cycle in River Basin,China Institute of Water Resources and Hydropower Research,Beijing,100048,China [4]Research Center on Soil and Water Conservation of the Ministry of Water Resources,Beijing,100048,PR China

出　　处：《International Soil and Water Conservation Research》2024年第4期808-827,共20页国际水土保持研究(英文)

基　　金：supported by the MWR Major Scientific&Technological Project[Grant number SKS-2022047,SKR-2022053];the IWHR Research&Development Support Program[Grant number SE0145B032021];the National Natural Science Foundation of China[Grant number 42007050];the Basic Scientific Research Projects of Liaoning Province[Grant number LJKMZ20221049].

摘　　要：In the black soil region of Northeast China,the issue of gully erosion persists as a significant threat,resulting in extensive damage to farmland,severe degradation of the black soil,and decreased productivity.It is therefore of utmost importance to accurately identify areas that are susceptible to gully erosion to effectively prevent and control its negative impact.This study tried to utilize geographical detectors(geodetectors)as a means to identify the factors that contribute to the distribution of gullies and assess the risk of gully erosion(GER)in five catchments within the region,with areas ranging from approximately 80 km^(2) -200 km^(2).By employing the geodetectors method,fourteen geo-environmental factors were analyzed,including topographic attributes(such as aspect,catchment area,convergence index,elevation,plan curvature,profile curvature,slope length,slope,stream power index,and topographic wetness index),channel network distance,vegetation index(NDVI and EVI),as well as land use/land cover(LULC).The modeling of GER was conducted using the random forest algorithm(RFA).Out of the fourteen examined geo-environmental factors,only a subset,comprising less than or equal to 50%,demonstrated a significant(p<0.05)influence on the spatial distribution of gullies.These selected factors were sufficient in assessing GER,with LULC(mean q-value=0.270)and elevation(mean qvalue-0.113)identified as the two most important factors.Furthermore,the RFA exhibited satisfactory performance across all catchments,achieving AUC values ranging from 0.712 to 0.933(mean=0.863)in predicting GER.Overall,the catchment areas were classified into high,moderate,low,and very low-risk levels,representing 9.67%-15.95%,19.28%-26.08%,24.59%-30.55%,and 30.54%-39.08%of the total area,respectively.Importantly,a significant positive linear relationship(r^(2)=0.722,p<0.05)was observed between the proportion of cropland area and the occurrence of high-level GER.Although the primary risk levels were categorized as low and very low,the proportion o

关 键 词：Gully erosion Geodetectors Random forest Gully erosion risk Black soil region 

分 类 号：R73[医药卫生—肿瘤]