Evaluation Of Geological Hazard Zonation Based On ArcGIS And Logistic Regression

Geological hazard happened frequently annually all over the world, which bring great damage to human beings. The study of geological hazard zonation could help the local government to manage and eliminate the hazard. This kind of studies began from 1950’s; it has already got some achievement in relevant areas. However researches most focused on the specific individual hazard such as landslide and mud-rock flow, the one on reprehensive zonal geological hazard is still developing. The hazard zonation would not only help manage and eliminate the geological hazard in one site, but also would improve the way people access the hazard zonation and investigate the hazard; make it both economical and cost efficient.This paper based on the particular investigation of Suide County, combined with the current relevant research methods and the specific geological condition of that county, chooses the influence factor system. Quantize the factors on ArcGIS, and then manage them through SPSS using Logistic Regression. And get the logistic parameter and individual weight.The author then combined the influence graph and their individual weight gets the probability graph on ArcGIS. The author chooses the slope unit to be the zonation basement and get the geological hazard zonation graph of Suide County. This paper divides the whole research region to four hazard level, which are: almost stable region, low risk region, medium risk region, and high risk region. Their acreage are respectively: 413km2、771.9km2、473km2、217.8km2. The result of this research is mainly consistent with the County’s site situation, which proves that logistic regression is a proper method in geological hazard zonation.