Abstract
Mapping susceptibility of landslide disaster is essential in subtropical area, where abundant rainfall may trigger landslide and mudflow, causing damages to human society. The purpose of this paper is to propose an integrated methodology to achieve such a mapping work with improved prediction results using hybrid modeling taking Chongren, Jiangxi as an example. The methodology is composed of the optimal discretization of the continuous geo-environmental factors based on entropy, weight of evidence (WoE) calculation and application of the known machine learning (ML) models, e.g., Random Forest (RF), Support Vector Machine (SVM) and Logistic Regression (LR). The results show the effectiveness of the proposed hybrid modeling for landslide hazard mapping in which the prediction accuracy vs the validation set reach 82.35–91.02% with an AUC [area under the receiver operating characteristic (ROC) curve] of 0.912–0.970. The RF algorithm performs best among the observed three ML algorithms and WoE-based RF modeling will be recommended for the similar landslide risk prediction elsewhere. We believe that our research can provide an operational reference for predicting the landslide hazard in the subtropical area and serve for disaster reduction and prevention action of the local governments.
Highlights
Mapping susceptibility of landslide disaster is essential in subtropical area, where abundant rainfall may trigger landslide and mudflow, causing damages to human society
Processing based on weight of evidence (WoE) can effectively reduce the collinearity among the factors
The collinearity among the geo-environmental factors selected for this research is low, and they can be used for susceptibility modeling
Summary
Mapping susceptibility of landslide disaster is essential in subtropical area, where abundant rainfall may trigger landslide and mudflow, causing damages to human society. The calculation process of the statistical models such as frequency ratio (FR), certainty coefficient (CF), information value (IV) and weight of evidence (WoE) is simple; and qualitative or categorical factors can be converted into quantitative weights by these approaches, and thence, they are widely employed for landslide risk a ssessment[15,21,22,23]. Based on the target definition, or rather, collection of samples for training, ML approaches can automatically analyze and extract rules from the input data to make predictions[14] It is highly efficient in calculating high-dimension data and can fit the nonlinear relationships between target and factors[8,29,30,31]. The prediction accuracy of the most studies, even including those harnessing the hotspotted deep learning techniques[32,33,34,35], comes between 75 and 85%, except for those of Huangfu et al.[36], Ou et al.[26], Zhang et al.[27] and Zhou et al.[28], who have achieved landslide risk prediction
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
