Study On The Method Of Regional Early Warning Of Landslide In Three Gorges Area Based On Information Model

Accompanied by the construction and operation of the Three Gorges Project, theecological environment of its reservoir area is affected a lot, which caused manygeological disasters of old and new, among them landslide is most. Based on the project“The High-Tech Industrialization Demonstration of Domestic Satellite Data in theThree Gorges Reservoir Area”, this study focused on the research of early warningmethod of landslides in the three gorges reservoir area. Information model is commonlyused in regional early warning of geological disasters. Based on its development, thepaper has given a complete calculation process of the model, and the model was addedwith the impact of four dynamic factors: rainfall, water level changes, vegetation andland-use type. The new method was practiced within Fuling, Chongqing. Themain contents and conclusions have the follow five points:1) Based on the application of information model in the field of geologicaldisasters’ early warning, its developments and the controversial problems have beensummarized. Combined with theoretical basis of information model, this paperintroduced its varied forms, such as the form of single factor superposition. Inaddition, this paper has analyzed the influence and applicable conditions ofthree kinds of computing cells. After all, based on the analysis, a completecalculation process information model was given.2) Based on the explanation of acting mechanisms of influencing factors,this paper has proposed the suitable expression of each internal factor for bothwithin an area or with a particular landslide, such as slope factor, aspect factor,and formation lithology.3) In the aspect of dynamic early warning, use early effective rainfallmodel to consider the effect of rainfall; use the relationship between water levelslope ratio and security coefficient to consider the effect of water level changes;use NDVI to consider the favorable and adverse effect of vegetation; and use theclassification result of land-use type from remote sensing image as the index ofaffected objects’ vulnerability to evaluate the affection of affected object.4) The paper suggests using two ways to evaluate the rationality of the calculation process of information model and the warning method thatconsidered the external factors, which were proposed in this paper.5) Have the improved early warning method practiced within Fuling districtin Chongqing. The result shows that the improvements of information model andthe additional dynamic effect were reasonable.