Reliably mapping low-intensity forest disturbance using satellite radar data

In the last decades tropical forests have experienced increased fragmentation due to a global growing demand for agricultural and forest commodities. Satellite remote sensing offers valuable tool monitoring loss, thanks coverage temporal consistency of acquisitions. regions, C-band Synthetic Aperture Radar (SAR) data from Sentinel-1 mission provides cloud-free open imagery on 6- or 12-day repeat cycle, offering unique opportunity monitor disturbances in timely continuous manner. Despite recent advances, mapping subtle losses, such as those small-scale irregular selective logging, remains problematic. A Cumulative Sum (CuSum) approach has been recently proposed applications, with preliminary studies showing promising results. Unfortunately, lack accurate in-situ measurements loss prevented full validation this approach, especially case low-intensity logging. study, we used high-quality field Forest Degradation Experiment (FODEX), combining unoccupied aerial vehicle (UAV) LiDAR, Terrestrial Laser Scanning (TLS), field-inventoried structural change collected two logging concessions Gabon Peru. The CuSum algorithm was applied VV-polarized ground range detected (GRD) time series canopy events, individual tree extraction clear cuts. We developed single metric using maximum distribution, retrieving location, time, magnitude disturbance events. comparison LiDAR reference map resulted 78% success rate test site 65% Peru, small 0.01 ha size height losses fine 10 m. correlation between above biomass (AGB) found R 2 = 0.95, 0.83 loss. From regression model directly estimated local AGB maps year 2020, at 1 scale percentages Comparison Global Watch (GFW) Tree Cover Loss (TCL) product showed 61% overlap when considering only deforested pixels, 504 deforestation by vs. 348 GFW. Low intensity captured method were largely undetected GFW SAR-based Detecting Deforestation (RADD) Alert System. results study confirm simple reproducible detection quantifying fine-scale high disturbances, even multi-storied forests.