Key Techniques Study Of Remote Sensing And Geographic Information System Based Landslide Risk Assessment

There have been improving emphasizes on the landslide risk assessment and management methodology and key techniques development for the domestic and overseas landslide study in recent years. Based on the illustration of the terminology and framework of landslide risk assessment, this paper reviews the problems and difficulties related to landslide inventory, basic data acquisition and updating, the determination of landslide spatial and temporal probability, the modeling of landslide run-out and characteristic aspect predication, and consequence analysis includes vulnerability quantitative analysis and element at risk quantitative mapping as well. Furthermore, an overview of recent developments in the different approaches to landslide hazard and risk assessment is given. As for remote sensing techniques for landslide study, this paper reviews the newly application and development of for landslide study and states that RS techniques offer a solid technique infrastructure, and exploits its potential development for landslide risk assessment at 5 phases: basic topographical data acquisition and extraction, landslide inventory and mapping, monitoring, landslide diagnostic factors mapping and element at risk mapping. From the role of RS in landslide study, interpretability, causative condition that influence landslide identification, accuracy assessment and RS data source selection, the author states and discuss some questionable and arguable aspects in the application of RS for landslide risk assessment, and as well the main trend of landslide risk assessment methodologies and techniques development is given.On the key techniques study for RS and GIS based landslide risk assessment methodology, this paper states application techniques of RS for landslide identification and spatial data acquisition. And the technique formwork and application scheme of landslide susceptibility, hazard and risk zonation were systematic represented as well. With the help of ArcGIS software spatial analysis function and diagram modeling, software tools for data or information acquisition and data geoprocessing were achieved, which can carry out data and information batch processing and rapid extraction. According to the methodologies and degree of difficulty of expert system, statistic models and etc. for landslide susceptibility mapping, diagram modeling, script and ArcObject software development were used respectively to carry out software tools based on ArcGIS and ERDAS platform for landslide susceptibility mapping.The regional area of Geheyan reservoir on the Qingjiang River in the eastern Hubei province was selected as case study area. Based on the landslide investigation, landslide inventory and RS imagery interpreation, DEM data , ETM+ satellite image and feature dataset, lithological map were used to generate landslide diagnostic factors and the relationship between the landslide spatial distribution and diagnostic factors statistical analysis were carried out the software tools, with different methods the susceptibility zonation of the Geheyan reservoir area of the Qingjiang River were carried out and divided into 4 classes: the high class covers 10.53%, the median class covers 8.54%, the low class covers 18.8% , whereas the extremely low class covers 62.15%. Based on the important section landslide mapping and basic engineering condition investigation at 1:10000 scale, with Quick bird imagery, 5m resolution DEM and basic geological map at 1:10000 scale, weight of evidence model software tools developed by script were used to carry out the landslide susceptibility zonation with natual slope unit as the basic assess unit, which integrated with the geological analysis and topological and gelological profile, landslide susceptibility zonation map were carried out. Landslide hazard were assessed from the following aspects: the dynamic triggering factors, the stability of exsiting landslide, runout and potential affect area, landslide frequence and heavy accident excess probability. And element at risk mapping, vulnerability and risk were qualitative assessed based on landslide inventory as well.