基于I+LR方法及GIS平台的麻城市地质灾害建模

Geological Hazard Modeling in Macheng City Based on I+LR Method and GIS

随着城镇化建设的推进,频发的山地地质灾害不仅制约了山区集镇的发展,还可能造成重大的人员伤亡和财产损失,因此进行地质灾害易发性研究对区域地质灾害的预测、防治工作具有重要意义。为提高山区集镇地质灾害易发性建模的精度及合理性,基于ArcGIS数字高程模型栅格数据,选取地形地貌、岩性、距断层距离、归一化植被指数(normalized difference vegetation index,NDVI)等11个因子,采用信息量、信息量-逻辑回归耦合模型对麻城市滑坡、崩塌灾害进行地质灾害易发性建模,最终通过栅格最大值提取合成综合地质灾害易发性评价模型。通过对单一模型与耦合模型的结果进行ROC(receiver operating characteristic curve)检验和合理性验证,以期表明信息量-逻辑回归(I+LR)耦合模型较传统信息量(I)模型评价结果更加合理、精度更高,耦合模型能够较为准确地评价山区地质灾害易发性。

With the advancement of urbanization,frequent geological disasters not only restrict the development of town in mountain areas,but also may cause major casualties and property losses.Consequently,the suspension of geological disasters is of great significance to predict and prevent regional geological disasters.In order to improve the accuracy and rationality of geological hazard susceptibility modeling,11 key factors of geological hazards,such as topographic features,lithology,distance from fault and NDVI(normalized difference vegetation index),were selected by digital elevation model grid data,and the geological hazard susceptibility model of landslide and collapse in Macheng City was built by using information,information value-logistic regression coupling analysis,finally the comprehensive geological hazard susceptibility assessment model was extracted by the maximum value of grid.The ROC(receiver operating characteristic)curve and rationality verification of the results of the single model and the coupling model show that the information value-logistic regression coupling model is more effective and more accurate than the traditional information value model,and the susceptibility of geological disasters in mountain areas can be evaluated more accurately by the coupling model.