一种适用于黄土斜坡地震稳定性快速评估的综合判别方法——以宁夏西吉县为例

A comprehensive method for rapid assessment ofseismic stability of loess slopes: a case studyof Xiji County in Ningxia

摘要:

选取宁夏西吉县为目标区,对区内1920年海原8.5级特大地震所诱发的347处黄土地震滑坡进行野外调查,获取了详实的基础数据。在此基础上结合遥感影像解译,总结了研究区内黄土地震滑坡的分布特征,并依此提出了宏观、定性的黄土斜坡地震稳定性快速判别方法。基于分布特征选取了坡高、坡角、坡向及地震动强度作为基本参数,应用MLP神经网络模型提出了黄土斜坡地震稳定性的定量快速判定模型。提出了一种定性、定量快速判定的综合判别方法,对研究区内震前规划选址及震后应急救援中黄土斜坡地震稳定性的快速评估均有着十分重要的指导意义。

Abstract:

In this paper, a detailed field investigation was carried out on 347 loess seismic landslides in Xiji County, Ningxia induced by the Haiyuan MS8.5 earthquake in 1920. Based on the detailed basic data and the interpretation of remote sensing images, the distribution characteristics of the loess seismic landslides in the study area were summarized. Then, a macroscopic and qualitative rapid discrimination method for the seismic stability of loess slopes was proposed. Based on the distribution characteristics, the slope height, slope angle, slope direction, and ground motion intensity were selected as basic parameters. A quantitative rapid discrimination model for the seismic stability of loess slopes was proposed by using the MLP neural network model. The comprehensive discrimination method proposed in this paper, which integrates the qualitative and quantitative fast judgment methods, is of great significance to the site selection before earthquake, as well as the rapid evaluation of seismic stability of loess slopes in the emergency rescue after earthquake.