Applying the Multiscale Model-to-Model Cloud Comparison (M3C2) method to evaluate landslide source area metrics in Puerto Rico, in Applying the Multiscale Model-to-Model Cloud Comparison (M3C2) method to evaluate landslide source area metrics in Puerto Rico

The dataset found within this project is the result of the application of the M3C2 automatic change detection methodology of Bernard et al. (2021) utlizing pre- and post-Hurricane María LiDAR data from Puerto Rico. Through this approach, over 180,000 landslide source areas were initially identified across multiple watersheds, including: Lago Caonillas, Lago Dos Bocas, Lago Lucchetti, and the Río Grande de Añasco. To ensure data quality, these results were filtered based on geomorphic thresholds, removing any landslide source area with an average depth of less than 0.5 m and area smaller than 25 m². After this filtering process, over 96,000 landslide events remained within the study regions. Additionally, data from the study 'Effects of Widespread Mass Wasting Triggered by Hurricane María in the Lago Caonillas Watershed, Puerto Rico' served as a key reference within this project. Specifically, the over 10,000 landslide source areas identified within the Lago Caonillas watershed are used as a benchmark to evaluate and compare the automated results produced using the M3C2 method.