基于GIS的城市滑坡灾害易发性评价——以湖北省宜昌市城区为例

目的]对城市滑坡灾害进行易发性分区评价,为城市规划与防灾减灾工程提供理论依据。方法]以湖北省宜昌市城区为研究区,通过GIS平台选取高程、坡度、地层岩性、归一化植被指数(NDVI)、与水系的距离和道路密度等6个评价因子,采用似然比方法分析评价因子和滑坡发育的关系,并以归一化似然比值将评价因子参数分类量化;以量化值作为Logistic回归模型的自变量,抽取样本数据建立滑坡易发性评价回归模型。结果]评价因子结果显著,模型的整体准确率达到79.2%,ROC曲线下面积达0.871;极低易发区和低易发区占全区面积的61.59%,包含滑坡灾害的11.29%;高易发区和极高易发区虽仅占全区面积的17.88%,却发育了68.55%的滑坡灾害,结果与滑坡灾害分布特征相符合。结论]对宜昌市城区的滑坡易发性进行了等级划分。采用GIS和Logistic回归相结合的滑坡易发性评价方法,结果准确可靠。

Landslide Proneness Evaluation Based on GIS Platform in Urban Area of Yichang City, Hubei Province

Objective] In order to provide a theoretical basis for urban planning and disaster prevention and mitigation engineering, the zoning evaluation on urban landslide proneness was conducted.Methods] The evaluation was conducted at urban area of Yichang City, Hubei Province. Evaluation indicators, i.e. elevation, slope gradient, lithology, normalized difference vegetation index(NDVI), distance to watercourse and roading density, were identified by GIS platform; the relations between landslide proneness and evaluation indicators were analyzed based on likelihood ratio method, and the evaluation indicators could be quantified using the generalized likelihood ratio. As the independent variable in Logistic regression model, the regression model of landslide proneness evaluation was established based on sample datum.Results] The significance of evaluation indicators was tested notable. The overall accuracy and the area under the ROC curve of evaluation model reached to 79.2% and 0.871 respectively. The extremely low proneness zone and the low proneness zone covered 61.59% of the total area, where landslide contributed 11.29% of the total landslides. Landslides in the high proneness zone and the extremely high proneness zone accounted for 68.55%, although it covered only 17.88% of the total area. The evaluation outcomes were conincided with the distribution of historical landslides by and large.Conclusion] The landslide proneness of urban area of Yichang City is classified. The result of landslide proneness evaluation based on GIS and Logistic regression model is accurate and reliable.