Probability mapping of debris flows triggered by multiple mechanisms in the Himalayas

Debris flows pose a significant hazard in the Himalayas due to the region's diverse climatic conditions and complex topography. However, previous studies have predominantly focused on individual debris flow types, often neglecting the multi-triggering mechanisms that influence their occurrence. This limitation has reduced the accuracy of probability assessments and hindered the development of effective risk management strategies for vulnerable areas. To address this gap, we developed an indicator system that incorporates multi-triggering mechanisms and applied three hybrid machine learning models to comprehensively assess debris flow probability. These models generated probability maps for Rainfall-Triggered Debris Flow (RTDF), Glacier Debris Flow (GDF), Glacial Lake Outburst Debris Flow (GLODF), and multi-type debris flows. The results indicate that high RTDF probability is concentrated in the Yarlung Zangbo River Valley, the Indus River Valley, and the southern slope. High GDF probability is primarily located in the Western Himalayas, while high GLODF probability is predominantly distributed along the Central and Eastern Himalayan ridge. Notably, 52.98 % of catchments are vulnerable to at least one type of debris flow, with 2.04 % at risk from all three types. This study addresses a critical gap in debris flow probability assessment by integrating multi-triggering mechanisms, offering valuable insights to improve risk management and enhance resilience strategies in the Himalayas.