Debris flows are destructive movements of water and earth, strongly influenced by environmental factors. In South America, extensive tropical forests, mountain chains, and high annual rainfall create favorable conditions for these geodynamic processes. However, there is still a need for further research on the triggers, soil mechanics, climate influence, and travel distances of such events. This study explores three physically based models for simulating the propagation of shallow landslides and associated flows in data-scarce regions, with the aim of developing a comprehensive analysis methodology for these events, the affected zones, and the processes that cause them. Using the 2015 Salgar landslide in Antioquia, Colombia, as a calibration case, where over 40 landslides triggered by a storm resulted in significant loss of life and property damage, three models were calibrated and evaluated: RAMMS, Flow R, and GPP from SAGA GIS. Following this, the models were validated using a similar event in the La Argelia basin, Carmen de Atrato, Chocó. Statistical analyses and ROC curve evaluations were conducted to assess the predictive performance of these models, considering the available data and event documentation. The findings suggest that while the models are useful for hazard assessments in tropical mountain basins, further refinement and data collection are essential to improve predictive accuracy. This research contributes to understanding landslide dynamics in tropical regions and offers insights for future hazard mitigation efforts.
Hurtado et al. (Sun,) studied this question.