Integrated CRITIC–TOPSIS and Machine-Learning Framework for tectonic activity and landslide hazard assessment in Higher Himalaya

Understanding natural hazards in the Himalayan terrain requires an analysis of both tectonic forcing and landscape response, particularly in the Higher Himalaya where steep terrain, active tectonics and heavy rainfall combine to create a high potential for natural hazards. The present study integrates a Multi-Criteria Decision Making (MCDM) and Machine Learning (ML) framework to comprehend tectonic activity and landslide susceptibility in this high mountainous region. The MCDM methods, specifically CRITIC-TOPSIS, provide a consistent assessment of relative tectonic activity and surface deformation patterns. To complement this, multi-year (2020–2025) machine learning methods, Random Forest and XGBoost were applied to generate annual landslide susceptibility maps. These maps revealed a gradual increase in moderate to high susceptibility zones across the years, particularly along fault-controlled slopes and steep valley walls. This indicates an evolving environment that is being actively modified by both human and natural factors. Ultimately, the combined CRITIC–TOPSIS–ML approach provides a powerful, multi-parameter methodology for identifying tectonically active zones and slope instability hotspots, facilitating the early identification of emerging risk zones in rapidly evolving mountainous regions.