Citation

Zhao, T., Peng, H., Xu, L., Sun, P. (2024). Statistical landslide susceptibility assessment using Bayesian logistic regression and Markov Chain Monte Carlo (MCMC) simulation with consideration of model class selection. Georisk: Assessment and Management of Risk for Engineered Systems and Geohazards, 18(1): 211-227. Link to paper

Abstract

This study proposes a Bayesian logistic regression (LR) method for landslide susceptibility assessment (LSA), together with Markov Chain Monte Carlo (MCMC) simulation for parameter estimation and model class selection. Landslide susceptibility mapping (LSM) plays an essential role in landslide management and contributes to decision-makers and planners in formulating landslide prevention policies. However, few studies explore the uncertainty and reliability of LR in LSM, and how to objectively determine the most relevant landslide conditioning factors (LCFs) remains an unsolved issue. The proposed method uses MCMC samples to determine the optimal model and to quantify the uncertainty associated with LSM. Real-life data from Shaanxi Province, China, are used for illustration. Results show that the proposed method works reasonably well in determination of the optimal model and in uncertainty quantification in LSM. The Bayesian approach provides a rigorous framework for incorporating prior knowledge, quantifying parameter uncertainty, and making probabilistic predictions. This research contributes to improving the reliability and transparency of landslide susceptibility assessment by explicitly addressing model uncertainty and parameter estimation through advanced statistical techniques.

AI Visual Interpretation (Beta)