A comparative machine learning approach to identify landslide triggering factors in northern Chilean Patagonia

摘要: Worldwide landslides correspond to one of the most dangerous geological events due their destructive power and unpredictable nature. In Chilean Patagonia, SERNAGEOMIN (Chilean Geological Survey Service) has detected 2533 landslide in Northern Patagonia (42. 7∘S, 72. 4∘W) alone, a small area compared whole Patagonia. However, only 11 evens have known date. Consequently, it is not possible associate temporal triggers mechanisms that control such events, resulting lack understanding factors enable landslides. This work aims detect identify main environmental variables (climatic geomorphological) explain occurrence using machine learning methods. We will address following research questions: 1) How can dataset be built Landsat images Google Earth Engine Patagonia? 2) Once timing been detected, what are condition processes? our work, we developed for northern Chile. used three approaches, where was allow us predict generation. Statistical models show during last 19 years, there complex interaction between different influenced activity Climatic indices, indicators extreme high incidence events’ predictive capacity. important those linked Patagonia’s tectonic context. particular, time elapsed after eruptive event Chaiten volcano. Finally, Liquine-Ofqui Fault System’s presence extending throughout entire north generated discontinuities at general level, causing significant geomorphological instability. The study relief evolution reactive climatic conditions. Therefore, highlight need understand better processes future impact these natural hazards. emphasize importance analyzing controls, considering both sporadic events.