Application of Certainty Factor and Bayesian statistics models for evaluation of landslides and environmental factors at Bao Thang district and Lao Cai city, Lao Cai province

Abstract

The study area is heavily affected by landslides with increasing frequency and intensity, causing serious damages and affecting the sustainable socio-economic development of the region. The use of mathematical methods in landslide research is increasingly interested due to the quantitative nature of parameters and calculation results. This study aims to apply the Certainty Factor (CF) and Bayesian statistics models for geological hazard evaluation. Landslide distribution is identified from remote sensing images and field surveys. Landslide inventory maps (428 landslides) were compiled by reference to historical reports, Google Earth, and field mapping. All landslides were randomly separated into two data sets: 70% were used to establish the models (training data sets) and the rest for validation (validation data sets). Fifteen environmental factors from geology, topography and hydrological information of the studied area were extracted from the spatial database. Results show that the group of factors of slope angle, Terrain Ruggedness Index, fault/lineament density, stratigraphy, geoengineering characteristics, weathering types, and maximum daily rainfall play the most important role in the formation of landslides in the study area. Validation from Certainty Factor (CF) and Bayesian statistics models show 87% and 92% prediction accuracy between hazard maps and existing landslide locations. These models show reasonably accurate landslide predictions in the study area and can be served as the basis of landslide risk-management studies in the future.