Information Extraction Of Landslides Induced By The Wenchuan Earthquake And Susceptibility Assessment Using Remotely Sensed Data

As a kind of secondary disaster caused by strong earthquake, the earthquake-induced landslide has drawn much attention in the world due to the severe hazard. Nowadays, researches on single landslides have been carried out more successfully comparing to the ones on regional landslides. To a certain degree, remote sensing and GIS, as new kinds of tools for investigation and monitor, have been used in many researches on landslide susceptibility and hazard mapping.The Ms 8.0 Wenchuan earthquake, occurred on 12 May 2008 in Sichuan Province, made a great many houses collapsed and hundreds thousand people injured. Meanwhile, damage caused by earthquake-induced landslides, collapse and debris flow became the major part of total losses. In order to remove threat from the secondary disasters effectively, this study used information value method and logistic regression, coupled with techniques of remote sensing and GIS, to generate susceptibility maps, taking the case of Wenchuan County. Main content and production can be summarized as follows:(1) Based on the remote sensing data, this paper introduces the significance of digital image processing technology (e.g. data fusion and ortho-rectification), illustrates the procedures and techniques of information extraction and analyses the characteristics of earthquake-induced landslides in the research area.(2) To perform the potential analysis, it was necessary to establish spatial database of landslides containing landslides'attributes and variables contributing to landslide in the study area. The database following the unified standard of digital surrounding can provide plenty of basic data for advanced process.(3) Seven factors controlling landslide occurrence have been taken account into the susceptibility assessment, including elevation, slop, aspect, lithology, seismic intensity, distance to faults and rivers. According to the probability that predicts the possibility of landslide occurrence calculated applying information value method and logistic regression separately, the study zone was ultimately categorized into five classes, specifically,"extremely low","low","moderate","high"and"very high". These results have been proved to reflect closely the spatial distributions of landslides in the study area.(4) It has been found that the predictive capability of logistic regression model appears to be more accurate comparing to information value method. It is mostly because logistic regression could reduce effectively the subjectivity in selection of evaluation factors and weight assignment.In sum, this paper extracts some information from remote sensing data, analyses the characteristic of earthquake-induced landslides under a GIS environment and accordingly generates the susceptibility maps in Wenchuan County, which undoubtedly contributes to recovered construction, location of habitation and prevention of secondary disasters in the study area.