Enhancing Landslide Vulnerability Mapping Through Automated Fuzzy Logic Algorithm-Based Methodology

The frequency and severity of worldwide landslides are increasing due to climate change and the human factor. That is why losses and damages to humans and materials are growing. These phenomena do not affect all regions similarly, especially the State of Guerrero (Mexico), an area vulnerable to landslides since the damage caused has been significant. However, the term vulnerability remains ambiguous and unclear in some studies since it is often confused with landslide susceptibility. Furthermore, there is no standard methodology to study landslide vulnerability, although a massive use of machine learning has been made in recent years. The methods are often heuristic, subjective, and rarely based on understanding the physical processes that control landslides. That is why this study proposes a methodology based on GIS, remote sensing, and fuzzy logic, making it a quantitative, automatic, scalable, and repeatable method, allowing the generation of multitemporal and multispatial cartography on vulnerability to landslides in the state of Guerrero. The vulnerability results calculated from a heuristic method have been compared with those generated with fuzzy logic, and despite not identifying significant differences in processing, differences in the final vulnerability categories can be seen, which can be attributed to less subjectivity.