基于C5.0决策树-快速聚类模型的万州区库岸段乡镇滑坡易发性区划

Landslide Susceptibility Evaluation for Township Units of Bank Section in Wanzhou District Based on C5.0 Decision Tree and K-Means Cluster Model

以乡镇为评价单元开展区域滑坡易发性评价对用地规划、防灾减灾等方面具有重要意义。以万州区临江段的23个乡镇单元作为研究对象,首先选取地表高程、坡度、坡向、岩性、构造、土地利用类型、地形湿度指数、水系、道路9个指标因子,通过C5.0决策树算法计算该区域发生滑坡的概率,再利用快速聚类算法进行易发性结果分级;基于ArcGIS平台得到各乡镇单元的滑坡易发性分区,结果表明:C5.0决策树-快速聚类模型的易发性评价精度最高,AUC值达到0.950,优于人工神经网络-快速聚类模型的0.826和贝叶斯-快速聚类模型的0.772。利用C5.0决策树-快速聚类模型的计算结果,综合考虑极高(高)易发区面积大小及其占乡镇面积比大小,完成各乡镇单元的滑坡易发性区划。在所有23个乡镇中,滑坡易发性等级高的包括大周镇、万州城区、溪口乡、新田镇等乡镇。通过对比各乡镇滑坡面积占研究区滑坡总面积的比重,发现两者结论基本一致,预测结果可为全区滑坡防灾减灾提供科学依据。

Susceptibility assessment of region landslides with township evaluation units is of great significance for disaster prevention, disaster reduction and land planning. 23 township units in Wanzhou District of the Three Gorges Reservoir region with frequent landslide disasters are the focus of this study. Nine influence factors including surface elevation, gradient, direction, stratum lithology, geological structure, water system, topographic wetness index, roads, and land utilization were chosen to be the evaluation indexes. The C5.0 Decision Tree and K-Means Cluster model was used to build the susceptibility evaluation systems. At the same time, the Artificial Neural Network model and Bayesian model were used to verify the evaluation results. A zoned map of landslide susceptibility for the township units was obtained based on GIS platform. The results indicate that the prediction accuracies of the C5.0 Decision Tree model reach 0.95, better than 0.826 of Artificial Neural Network model and 0.772 of the Bayesian model. The high prone area and the area ratio of the whole units were considered in analysing the susceptibility of landslides in the research area. The townships, including, Dazhou Town, Wanzhou City, Xikou Town, and Xintian Town, are at higher risks. The comparison of the proportion of the landslide area of each township to the total area in the study area illustrates that the situation is basically consistent, which can provide an effective means for risk management in townships.