Integration Of Multi-source Data Analysis Rainfall Type Of Shallow Landslide Risk After The Quake And Its Evolution Characteristics

The study area is located in the Longmenshan fault zone, is one of the most serious areas of geological disasters in China. May 12,2008 to the outbreak of the Richter 8.0 earthquake, the epicenter is in located in Wenchuan County, which is part of the study area, choosing landslide-prone areas as the study area after the 5.12 earthquake, evaluate the risk of rainfall-induced shallow landslides in the area, obtaining multi-source data, including relevant statistical data, GIS data, remote sensing data and TRMM data, selecting the Influencing factors of the stability, including slope, vegetation cover, roads, water systems, soil, lithology, fault zones, etc; the causing factors of the risk, including Rainfall;the affecting factors of the vulnerability, including demographic, economic, medical and other, Combing with landslide hazard database of the study area, using analytic hierarchy process model, information model, certainty factor model to assess the stability Respectively, zoning risk using thresholds, evaluating the vulnerability the affecting factors using expert scoring method, at last, we evaluate the risk of rainfall-induced shallow landslides after the 5.12 earthquake according to the risk assessment system proposed by Shi Peijun in 2002.Accuracy assessments of the 3 model are carried out by ROC curve, and the uncertainty coefficient model turns out be the highest accuracy method. Result of the evaluation is divided into five levels of risk from low to high.500m resolution as the evaluation unit, July 2008 to December 2010 for the study period, per month for a time unit, studying the intrinsic link between the factors and landslide risk in the study area, and trying to explore the Characteristics and patterns of the risk in two aspects-the sequence space and time series -in the study area(including 39 counties and 30 time units).In short, the entire study using a multi-source data and a variety of assessment models to a specific region-specific landslide type landslide risk assessment in a particular area, the final analysis of the study area in space and time variation intrinsically linked.