EVALUATION OF FOREST DEGRADATION METHODOLOGIES USING LANDSAT TIME SERIES ON ARGENTINEAN DRY CHACO FOREST

Nowadays, several methodologies allow disturbance detection in tropical forests with good results. On the other hand, subtropical forests represent a challenger as the robustness of these tools can be reduced by seasonality. In this work, we test three disturbance detection methods to assess their performance in a biome with a dry season like the Argentinean dry Chaco forest. We focused on the evaluation of both the accuracy and the number of hits by type of disturbance. The methods show acceptable accuracy values for the complexity of the problem and proved to be a very useful tool as a proxy of deforestation.