Study On The Plateform Of Rainfall-induced Landslide Prediction System

Landslide caused the most serious problems in natural disaster. It brings numerous tragedies to human being: it take lives, make wealth loss, and hold up the sustainable development of regional economy. The reason that causes disaster is landslide is unstable and sudden. Therefore the forecasting of landslide is very important.Though there many domestic and overseas studies of prediction methods on landslide, few can succeed. The prediction on rainfall-induced landslide is rather practical. This thesis researched the current of rainfall-induced landslide prediction system, and drew conclusions: Some systems are just weather forecast, have no relations with the topographty, physiognomy and other landslide factors. Some systems are monitoring and forecasting of single landslide and can not fit for local large-scale landslides. Some systems are just databases and lack of electronic-map display and simple & convenient operations. This thesis uses impersonal weights for landslide zonation and makes it as space-prediction, the rain forecast as time-prediction. As most current prediction systems are for science research, and do not fit for disaster-reduction management and disaster-prevention, this thesis aimed at a GIS based prediction platform for disaster-reduction department which directed by software-engineering theories.The target research area is located in Wanzhou, Chongqing, China, where the landslide hazards is most severe in the Three Gorges area. According to the statistic data, 90% landslides happened during July to September and rain is its principal causing factor. On the base of already-finished credible big-scale landslide zonation maps and the research of rain-landslide relation, overlay these two together and the prediction model was achieved (with the help of GIS, database, and software developing technology).After detailed user demands investigate, the target population and system target were precisely located. Contrast with similar GIS systems, the total structure and detail function were confirmed. With interface design, the user-kind interface were made and every function with carried out by different modular. The developing process and quality was ensured. The rain data is provided by the Chongqing Meteorological Bureau. This research provides a effective prediction platform and a easy-use landslide disaster management tool for Wanzhou government. This prediction system provides electronic maps, data tables, figures, and other visual and graphical means to express prediction process and the probability of landslide of a local area to none-professional officials. The study consists of three features as follows:As a big-scale rainfall-induced landslide prediction system, it meet the requirements of government.Most domestic GIS apply system lack of scientific process, and just concentrated on is coding. This system have a rather formal developing process which is directed by software engineering, and based on user-demand investigate, it emphasized on system design and system test and maintenance.The system provided many view modes (special data, attribute data, pictures, figures) to make it easy to understand the process, landslide prediction, and corresponding treatments.