Earthquake-Induced Landslides Susceptibility Evaluation: A Point Process Modelling Approach

Abstract

Landslides pose a significant threat to human lives and infrastructure in mountainous regions such as the Chitral district of Northern KPK, Pakistan. This study aims to assess the earthquake induced landslides in the region by integrating the creation of a landslide inventory with the application of Point Pattern Analysis (PPA), Logistic Regression (LR), and Frequency Ratio (FR) models. The research began by compiling a comprehensive landslide inventory through the analysis of high-resolution satellite imagery and temporal study in Google Earth Pro. The inventory includes about 210 landslides that occurred due to the high seismicity in the region. To analyze the spatial distribution and clustering of landslides, Point Pattern Analysis (PPA) was applied. Furthermore, Logistic Regression (LR) and Frequency Ratio (FR) models were employed to develop Landslide Susceptibility Maps (LSMs). These models utilized the landslide inventory data along with relevant terrain, geological, and environmental variables as causative factors or predictors. The results of the study revealed significant spatial patterns and hotspots of landslide occurrence in the Chitral district. Logistic Regression model shows that 18.1 % of the region is highly susceptible towards landslide whereas the Frequency Ratio model reveals 26.1 % of the study area falls in Very high susceptible class in Landslide Susceptibility Mapping. Accuracy assessment through Receiver Operating Characteristics (ROC) Analysis reveals that the overall accuracy of LR model (85.34 %) was better than FR model (78.56 %). The findings of this research provide valuable insights into the spatial distribution of landslides in the Chitral district and contribute to improved understanding of the underlying factors influencing landslide susceptibility. The generated landslide susceptibility maps can be utilized for land-use planning, infrastructure development, and disaster risk management in the region.