Faculty of Land Resources Engineering, Kunming University of Science and Technology, Kunming 650093, China.
Key Laboratory of Geospatial Information Integration Innovation for Smart Mines, Kunming 650093, China.
Sensors (Basel). 2023 Feb 24;23(5):2549. doi: 10.3390/s23052549.
A landslide is one of the most destructive natural disasters in the world. The accurate modeling and prediction of landslide hazards have been used as some of the vital tools for landslide disaster prevention and control. The purpose of this study was to explore the application of coupling models in landslide susceptibility evaluation. This paper used Weixin County as the research object. First, according to the landslide catalog database constructed, there were 345 landslides in the study area. Twelve environmental factors were selected, including terrain (elevation, slope, slope direction, plane curvature, and profile curvature), geological structure (stratigraphic lithology and distance from fault zone), meteorological hydrology (average annual rainfall and distance to rivers), and land cover (NDVI, land use, and distance to roads). Then, a single model (logistic regression, support vector machine, and random forest) and a coupled model (IV-LR, IV-SVM, IV-RF, FR-LR, FR-SVM, and FR-RF) based on information volume and frequency ratio were constructed, and the accuracy and reliability of the models were compared and analyzed. Finally, the influence of environmental factors on landslide susceptibility under the optimal model was discussed. The results showed that the prediction accuracy of the nine models ranged from 75.2% (LR model) to 94.9% (FR-RF model), and the coupling accuracy was generally higher than that of the single model. Therefore, the coupling model could improve the prediction accuracy of the model to a certain extent. The FR-RF coupling model had the highest accuracy. Under the optimal model FR-RF, distance from the road, NDVI, and land use were the three most important environmental factors, ac-counting for 20.15%, 13.37%, and 9.69%, respectively. Therefore, it was necessary for Weixin County to strengthen the monitoring of mountains near roads and areas with sparse vegetation to prevent landslides caused by human activities and rainfall.
滑坡是世界上最具破坏性的自然灾害之一。准确地对滑坡灾害进行建模和预测，已被用作滑坡灾害防治的重要工具之一。本研究旨在探讨耦合模型在滑坡易发性评价中的应用。本文以威信县为研究对象，首先根据构建的滑坡目录数据库，研究区内有 345 个滑坡。选取了 12 个环境因子，包括地形（海拔、坡度、坡度方向、平面曲率和剖面曲率）、地质构造（地层岩性和距断裂带的距离）、气象水文（年平均降雨量和距河流的距离）和土地覆盖（NDVI、土地利用和距道路的距离）。然后，构建了基于信息量和频率比的单模型（逻辑回归、支持向量机和随机森林）和耦合模型（IV-LR、IV-SVM、IV-RF、FR-LR、FR-SVM 和 FR-RF），并对模型的准确性和可靠性进行了比较和分析。最后，讨论了最优模型下环境因素对滑坡易发性的影响。结果表明，9 个模型的预测精度范围为 75.2%（LR 模型）至 94.9%（FR-RF 模型），耦合精度普遍高于单模型。因此，耦合模型可以在一定程度上提高模型的预测精度。FR-RF 耦合模型的精度最高。在最优模型 FR-RF 下，距道路的距离、NDVI 和土地利用是三个最重要的环境因素，分别占 20.15%、13.37%和 9.69%。因此，威信县有必要加强对道路附近山体和植被稀疏地区的监测，以防止人类活动和降雨引发的滑坡。
