基于逻辑回归模型的黄土滑坡易发性评价

打开文本图片集

中图分类号：P642.4 文献标志码：A 文章编号：2095-2945（2025）21-0085-05

Abstract：Loesslandslidesareatypical geologicaldisasterwidelydistributedaroundtheworld，andtheirsusceptibility evaluationhasalwaysbeearesearchhotspotinthefieldofengineringgeology.Thispapertakes Xifeng District，Qingyang City，GansuProvince，atypicalareaoftheLoessPlateau，astheresearchobject.Usingthecombinationofgeographical informationsystem（GS）andlogisticregresionmodel，multipleinfluencefactorsicludingprecipitation，slope，soiltype，slope aspect，etc.wereselectedtobuildalandslidesusceptibilityevaluationmodel.Throughmulti-factoranalysis，itisfoundthat precipitationandslopearethemainfactorsafectinglandslidesusceptibility.Theareaswithhighlandslidesusceptiblityin easternGansuaremainlyconcentratedinareaswithlargeslope，highprecipitationandthick loesslayers.Thisstudyprovidesa scientific basis for landslide disaster prevention and risk assessment in Longdong region.

Keywords： deep loess; landslide; logistic regression model； susceptibility evaluation; GIS

滑坡是全球范围内常见且严重的地质灾害，尤其在山区和黄土高原等易滑坡地带，对生命财产安全和社会经济发展构成了巨大的威胁。（剩余6079字）