基于Maxent的四川省滑坡易发性评价

Assessment on Landslide Susceptibility in Sichuan Based on Maxent Model

滑坡易发性建模有利于识别滑坡危险区域,对滑坡灾害防治具有重要意义。基于Maxent模型分析影响四川省滑坡易发性的环境变量并讨论滑坡与相关变量的响应关系,进一步依据滑坡发生概率进行四川省滑坡易发性分区评价。模型AUC值为0.863,表明模型性能较好。距河流距离、土地利用类型、距道路距离、高程、降水和岩性是影响四川省滑坡发生的重要变量。根据模型预测,滑坡极易发生区域主要出现在成都平原与西部山地和高原的过渡地带。

Landslide susceptibility modeling is helpful to identify dangerous areas and is of great significance to landslide disaster prevention. Based on Maxent model, this study analyzes the environmental variables affecting the landslide susceptibility in Sichuan Province and discusses the variable-response relationships between landslides and related variables, and further evaluates the landslide susceptibility zone in Sichuan Province according to the probability of landslide occurrence. The AUC value of the model is 0. 863, indicating that the model has good performance. Distance from river, land use type, distance from road, elevation,precipitation and lithology are the important variables affecting the occurrence of landslides in Sichuan Province. The landslide prone area by model prediction is mainly in the transition zone between Chengdu plain and western mountain and plateau.