Abstract: In an attempt to probe into the distribution of geological disasters in southwest China where geological disasters happen frequently, we built an index system for assessing the eco-geological environment of Neijiang City as a case study based on remotely sensed images and geographical monitoring data. The index system includes factors of natural geography, ecological condition, and fundamental geology. Furthermore, the subjective weight and objective weight of each index were determined by analytic hierarchy process (AHP) and dichotomous Logistic regression model; and the combinatorial weight was synthesized according to minimum relative information entropy principle. The evaluation model of geological hazard was hence established with the support of GIS to evaluate the risk of geological disasters in Neijiang City. The distribution characteristics and laws of geological disaster was acquired by spatial analysis using higher weight coefficients and the corresponding historical geological disaster points. Results show that the low mountainous area in the west of Neijiang, which accounts for 28.93% of the total area, is highly susceptible to geological disasters. Altitude and slope gradient are major influential factors that induce geological disasters; geological disasters are highly prone to occur in high-altitude and steep-slope areas on both sides of highway, whereas in the adjacent of quiet tectonic line the geological formation is relatively stable.

Key words: geological disaster, risk assessment, analytic hierarchy process (AHP), Logistic regression model, minimum information entropy, Neijiang City