Landslide type inference based on statistical analysis of a high-resolution digital elevation model in Gorce National Park, Poland

Abstract

Abstract High-resolution digital elevation models are commonly utilized for detecting and classifying landslides. In this study, we aim to refine landslide detection and classification by analyzing the geometry of landslides using slope and aspect, coupled with descriptive statistics up to the fourth central moment (kurtosis). Employing the Monte Carlo method for creating terrain topography probability distributions and ANOVA tests for statistical validation, we analyzed 364 landslides in Gorce National Park, Poland, revealing significant kurtosis differences across landslide types and lithologies. This methodology offers a novel approach to landslide classification based on surface geometry, with implications for enhancing scientific research and improving landslide risk management strategies.