川西高山峡谷地区震后滑坡演化趋势研究

Study on the Evolution Trend of Post-earthquake Landslide in Alpine Valley Area of Western Sichuan

汶川地震引发的同震滑坡给川西高山峡谷地区人民带来了巨大威胁。震后在强降雨的影响下,新生滑坡及古滑坡活动加剧,分析震后滑坡的演化趋势,对重点地区及时开展监测预警显得尤为重要。以汶川县东北部为研究区,基于GIS平台选取岩性、距断层距离、PGA、高程、坡度、坡向、剖面曲率、地形起伏度、年最大24 h降雨等9个影响因子,利用滑坡频率密度、面积密度和数量密度,统计分析2009年、2011年、2015年、2021年4期滑坡时空演化特征;同时结合证据权重(WOE)、随机森林(RF)、证据权重-随机森林(WOE-RF)模型,开展研究区滑坡易发性演化趋势分析。结果表明:随时间推移,滑坡数量、面积和规模均大幅减小,已处于较低水平;滑坡时空演化以2015年为转折由南向北发展;经ROC验证,WOE-RF得到的滑坡易发性精度最高,且三种模型均显示滑坡极高易发区演化也呈远离震中的由南向北之势。研究结果为川西高山峡谷地区滑坡灾害早期识别与监测预警提供理论依据。

The same earthquake landslide caused by the Wenchuan M S8.0 poses a massive threat to the people in the alpine valley area of western Sichuan.Under the influence of heavy rainfall after the earthquake,the new landslides and ancient sliding slopes have intensified.The evolution of the post-earthquake landslide is significant for conducting monitoring and early warning in critical areas on time.we take the northeast of Wenchuan County as the research area.The impact factors are selected for the lithology,distance from the fault,PGA,elevation,slope,slope direction,profile curvature,topographic relief,and annual maximum 24H rainfall.Based on the frequency density,area density,and quantity density of landslides,landslides’temporal and spatial evolution characteristics in 2009,2011,2015,and 2021 are statistically analyzed.Meanwhile,combined with WOE,stochastic forest(RF),and evidence weighting-stochastic forest(WOE-RF)models,the evolution trend of landslide susceptibility in the study area is analyzed.The results show that:with the passing of time,the number of landslides,the area,and the scale of the landslide has been dramatically reduced,which is at a low level;the evolution of landslides developed from south to north in 2015;after the ROC verified the landslide obtained by WOE-RF is easy to develop.The highest sexual accuracy and the three models show that the evolution of the high-tech area of the landslide is also far from south to north.The research results provide the theoretical basis for early identification,monitoring,and warning of landslides in the alpine valley area of western Sichuan.