基于地貌特征的滑坡堰塞坝形成敏感性研究

Formation sensitivity of landslide dam based on geomorphic characteristics

滑坡堰塞坝是由斜坡失稳堵塞河道而形成的天然坝体,且易溃坝诱发洪水,对沿岸群众生命财产构成巨大的威胁。为提升主动减灾防灾能力,急需构建了一种快速预测与判断滑坡堵江成坝能力的方法。通过文献资料查阅,结合遥感技术,提取了70处典型滑坡的地貌特征参数,其中50处为堵江成坝滑坡。运用K-S检验和M-W U检验方法分析了滑坡地貌特征因子的敏感性,利用Boruta算法确定了因子重要度,筛选了滑坡体积、面积、高差、长度及河宽共5个地貌特征参数。基于此,利用Bayes判别法与逻辑回归方法,分别建立了滑坡堰塞坝形成的预测模型,准确率超过90%。选取高重要度且差异显著的因子,利用比值法建立了滑坡堵江成坝阈值判据,实现了滑坡堰塞坝形成的快速判定。统计不同诱因下滑坡地貌特征,对比V-W r经验公式,确定了滑坡堰塞坝形成与诱因间的关系,为进一步构建不同诱因下滑坡堰塞坝形成预测模型提供了技术支撑。

Landslide dams formed by unstable slope usually blocks the river,and later fail to trigger huge floods threatening the downstream residents′lives and properties.In order to improve the ability of active disaster reduction and prevention,it is urgent to construct a method to quickly predict and judge the ability of landslide to build dams.This paper firstly extracts the geomorphic characteristic parameters of 70 typical landslides based on literature reading and remote sensing,including 50 damming landslides.The five parameters(landslide volume,area,relief,length and river width)are determined by K-S test and M-W U test,and the Boruta algorithm is used to determine the factor importance.Then,the Bayes discriminance method and logistic regression method are used to build a prediction model of the landslide dam,in which the accuracy exceeds 90%.The threshold equation combined with H and L/W r is determined by using factors with high importance and significant differences and the ratio method in order to carry out a rapid judgment of landslide dams formation.The statistic analysis of the landslide geomorphic characteristics of different triggering factors and the prediction result of V-W r empirical formula determine the relationship between landslide dams formation and triggering factors,providing technical support for the further construction of prediction models for the formation of landslide dams with different inducements.