DETER-R: An Operational Near-Real Time Tropical Forest Disturbance Warning System Based on Sentinel-1 Time Series Analysis

Continuous monitoring of forest disturbance on tropical forests is a fundamental tool to support proactive preservation actions and stop further destruction native vegetation. Currently most the systems in operation are based optical imagery, thus flaw-prone areas with frequent cloud cover. As this, several Synthetic Aperture Radar (SAR)-based have been developed recently, aiming all-weather detection. This article presents main aspects results first year SAR Near Real-Time Deforestation Detection System (DETER-R), an automated deforestation detection system focused Brazilian Amazon. DETER-R uses Google Earth Engine platform preprocess analyze Sentinel-1 time series. New images treated analyzed daily. After analysis, vectorizes clusters deforested pixels sends corresponding polygons environmental enforcement agency. 12 months operational life, has produced 88,572 warnings. Human validation warning showed extremely low rate misdetections, less than 0.2% detected area false positives. During operation, provided 33,234 warnings interest national agencies which were not by its counterpart DETER same period, 105,238.5 ha, or approximately 5% total detections. rainy season, additional detections increased as expected, reaching 8.1%.