Author(s)

Qiusheng Wang ¹, Youwen Cui ¹, Zhang Kai ²

Affiliation(s)

¹ Beijing University of Technology, Beijing, 100124, China
² China Renewable Energy Engineering Institute, Beijing, China

Corresponding Author

Zhang Kai

ABSTRACT

Landslides are one of the most destructive geological hazards in the Plateau, and their susceptibility assessment is of great significance for regional disaster prevention and mitigation. This study took the Great Bend region of the Yarlung Zangbo River-an area with complex terrain, frequent geological activities, and frequent landslide hazards-as the research area. Based on GIS technology and multi-source data, 12 landslide-influencing factors including elevation, slope angle, aspect, NDVI, and stratigraphic lithology were selected. After eliminating 2 factors with multicollinearity through Pearson correlation analysis and variance inflation factor test, categorical variables were processed by one-hot encoding, and a landslide susceptibility prediction model based on the eXtreme Gradient Boosting (XGBoost) algorithm was established. The model performance was evaluated by confusion matrix and ROC curve, and the factor influence mechanism was analyzed by the SHAP method, finally generating a four-level spatial zoning map of landslide susceptibility. The results show that the XGBoost model has excellent prediction performance, with an accuracy, precision, and recall rate of 0.8778 on the test set and an AUC value of 0.92, which can effectively identify high-risk areas. The extremely high susceptibility areas are mainly distributed along the main and tributary valleys of the Yarlung Zangbo River, with a landslide density of 0.430 landslides/km², which is highly consistent with the actual landslide distribution. The research results provide a scientific basis for the early warning and prevention of landslide hazards in the study area, and also offer reference for relevant assessments in alpine and canyon regions of Southwest China.

KEYWORDS

Susceptibility Assessment; XGBoost Model; Spatial Zoning Map

CITE THIS PAPER

Qiusheng Wang, Youwen Cui, Zhang Kai. Landslide Susceptibility Assessment in the Plateau Based on the XGBoost Model. Journal of Civil Engineering and Urban Planning (2026). Vol. 8, No.1, 52-61. DOI: http://dx.doi.org/10.23977/jceup.2026.080106.

REFERENCES

[1] Zhai Wenhua, Wang Xiaodong, Wu Mingtang, et al. Evaluation of geological disaster susceptibility based on frequency ratio model and random forest model coupling[J]. Journal of Natural Disaster Science, 2023,32(06):74-82.
[2] Zhang Zhipeng, Wei Zaihao. Evaluation of landslide disaster susceptibility based on weighted information quantity model: A case study of Baqiao District [J]. Science and Technology in Engineering, 2020,20(09):3492-3500.
[3] Huang F. Prediction of landslide displacement and susceptibility assessment based on 3S and artificial intelligence [D]. China University of Geosciences, 2017.
[4] Deng Shujun. Comparative study on quantitative evaluation methods for landslide susceptibility [D]. China University of Geosciences (Beijing), 2022.
[5] Liu Rui, Shi Shuxian, Sun Deliang, et al. GIS and Random Forest-Based Zoning of Landslide Susceptibility in Wushan County [J]. Journal of Chongqing Normal University (Natural Science Edition), 2020,37(03):86-96.
[6] Cui B, An Hui-Lun, Chen W-L, et al. Research on arch dam stress prediction model based on PCA-SSA-XGBoost algorithm [J]. Hydropower, 2024,50(05):45-53.
[7] Huangfu Wenchao. Application of logistic regression model in landslide disaster susceptibility assessment [D]. Donghua University of Science and Technology, 2021.
[8] Feng QR, Zhang JF, Zhu JJ, et al. Prediction method for formation fracture width based on LightGBM algorithm [J]. Energy and Environmental Protection, 2025,47(01):65-72.
[9] Goetz J N, Brenning A, Petschko H,  et al. Evaluating machine learning and statistical prediction techniques for landslide susceptibility modeling[J]. Computers & geosciences, 2015, 81: 1-11.
[10] Wu Hongyang, Zhou Chao, Liang Xin, et al. Evaluation and zoning of landslide susceptibility in Yanshan Township, Three Gorges Reservoir Area based on XGBoost model [J]. China Journal of Geological Hazards and Prevention, 2023,34(05):141-152.
[11] Zhao Shuai, Zhao Zhou. Evaluation of geological hazard susceptibility based on information quantity model [J]. Hydropower, 2019,45(03):27-32.
[12] Geng Haopeng. (2025). Catalog Database of Landslides in the Yajiang Large Bend Section (1987-2021). National Tibetan Plateau Science Data Center.
[13] Peng Shouzhang, Ding Yongxia, Liu Wenzhao, Li Zhi. 1 km monthly temperature and precipitation dataset for China from 1901 to 2017. Earth System Science Data, 2019, 11, 1931–1946.
[14] Jie Yang, & Xin Huang. (2026). The 30 m annual land cover datasets and its dynamics in China from 1985 to 2025 [Dataset]. Published in Earth System Science Data (1.0.5), Vol. 13, No.1, pp. 3907–3925. Zenodo.
[15] Ke C, He S, Qin Y. Comparison of natural breaks method and frequency ratio dividing attribute intervals for landslide susceptibility Mapping [J]. Bulletin of Engineering Geology and the Environment, 2023, 82(10).
[16] Agboola G, Beni L H, Elbayoumi T,  et al. Optimizing landslide susceptibility mapping using machine learning and geospatial Techniques [J]. Ecological Informatics, 2024, 81: 102583.
[17] Hu Qilong, Wang Yunsheng. Evaluation of landslide susceptibility based on weighted information quantity and GIS [J]. Hydropower, 2018,44(08):31-35.

Sponsors, Associates, and Links

---

• Journal of Sustainable Development and Green Buildings
  
  
• Landscape and Urban Horticulture
  
  
• Bridge and Structural Engineering
  
  
• Soil Mechanics and Geotechnical Engineering
  
  
• Journal of Municipal Engineering
  
  
• Heating, Ventilation and Air Conditioning
  
  
• Indoor Air Quality and Climate
  
  
• Computer Aided Architecture Design