With the advent of geoinformation technology including remote sensing (RS), Geographic Information System (GIS), Global Positioning System (GPS) or Beidou System (BDS) and powerful computer processing facility, acquisition and processing of geo-environmental factors with high resolution have been greatly facilitated 8, 12. Accurate and reliable mapping of landslide risk is a key step for local decision-makers and authorities to plan reasonable land use and implement disaster reduction and prevention measures to reduce the massive damage 5, 6, 7, 8, 9.Īctually, a number of scientists have been exploring reliable approaches for landslide hazard mapping 10, 11. With the socioeconomic development and the continuous expansion of human activities into the natural environment, landslide occurs more and more frequently and constitutes the main disaster threatening the safety of life and restricts the economic development in the hilly and mountainous areas 1, 2, 3, 4. Landslide is a common geological disaster leading to destruction and damages to human society in subtropical areas. 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. 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. 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 of 0.912–0.970. 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 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.
Mapping susceptibility of landslide disaster is essential in subtropical area, where abundant rainfall may trigger landslide and mudflow, causing damages to human society.
0 kommentar(er)
