Multiple Regression Analysis In Spatial Prediction Of Landslide

Landslide is one of the most serious disasters which bring people loss of life and property, so the prediction of landslide is the most effective ways and means to reduce landslide losses, which is also the focus and the difficulty of today's landslide research.This paper is based on the project of detailed investigation of geological disasters on Yaozhou Distract of Tong chuan city, after collecting a wealth of information, studying various factors of landslide disasters, analyzing the relationships between geological hazards and the factors comprehensively, giving exploratory answers to two questions-the landslide area and the intensity of landslide. It draws these conclusions:According to calculating and analysis of the coefficients determination of landslide disaster on Yao city, we can obtain most conducive to research the background conditions of the landslide area.After a sensitivity analysis, we can see slopes, soil properties, roads, as the study area are important factors, based on these conditions, creating landslide hazard zonation model logistic regression.Based on field survey data, quantitative analysis on the sliding distance of the scattered-type landslide on the study area, using multiple regression analysis, inferring mathematical predication (function) model on sliding distance of the scattered-type landslidingSummary of the common slow-acting slow landslide movement patterns on study area, qualitative analysis to get the range of landslide hazards of this type- the front sliding body with stabilizing effect.