Studies On Analyses And Prediction Of Landslide Monitoring Based On The Technique Of GIS

Landslide is a kind of very severe nature disaster, which can result in great loss to human. It is an import method of the disaster reduction and resistance to monitor, measure, analyse and predict the stability of landslide. In which, enormous monitor and measure data of landslide should be analyzed and treated in space scope. Due to GIS holding predominance in the storage, management, analysis, calculation and picture formation and exhibition of mass spatail data, it is a significant search direction to use GIS technique analyzing and treating spatail data of the landslide mass, and further predict the landslide stability.Based on the GIS, the paper profoundly studies the spatail disaster analysis of landslide and prediction of landslide stability.Digital Elevation Model (DEM) of the landslide mass being established, based on the DEM of the landslide mass, thespatailanalysis and study is carried out on the method of calculating cubic metre of earth and stone of the landslide, the drawing of geologic profile of the internal structure in the landslide, the drawing of terrain section between two-points on the landslide.The paper advanced Earth Calculation Method by Triangulated Irregular Network and designed an arithmetic of ECMTIN .By comparing accuracy of three methods, which accuracy is higher than the section method and the mesh method.The paper designs the thought route using DEM to drawing geologic profile of the internal structure in the landslide: combine organically the geological formation of landslide mass with the DEM to form landslide digital elevation model, though establishing the geological data base of landslide mass and program design, we can rapidly draw the geological profile of the internal structure in the landslide.The paper generalizes and analyses the modeling mechanics and applicable principle of the representative landslide prediction model such as Gray model prediction, exponential smoothing model, nonlinear regress analysis prediction model, trend prediction,Verhulst model prediction and so on. The processing method of the multiple models prediction results was educed.Based on all of above contents, the framework, function settings and developing way of Landslide Monitoring Datum Analyses and Prediction System are put forward, and the system database is also designed.Finally, the partial functions of this system are performed on the ArcView platform in this thesis. Some examples are tested by the system.