Remote estimation of harvested timber volume in the Congo Basin using Sentinel-1 data

A third of the world’s forest area is currently used for timber production, yet independent verification of selectively logged wood volumes remains challenging. Unsustainable logging undermines biodiversity, carbon storage, and forest resilience, especially under increasing land-use pressure. Here we test whether reported harvested timber volume can be reliably estimated using canopy gaps as detected using fine-scale disturbance data from Sentinel-1 radar across 93 concessions in the Congo Basin. We find a strong linear relationship (R 2 up to 0.86, bootstrap RMSE 8 . 67 ∗ 1 0 4 m 3 ) between reported harvested volume and detected disturbance area, independent of logging road signals. Our approach outperforms five widely used near-real-time forest disturbance alerting systems and annual change products, which largely capture only persistent canopy openings such as road corridors (R 2 0.28–0.74). Aggregating harvested volumes and disturbance area across multiple years improved the strength of the relationship while maintaining a consistent regression slope. Accurate harvest estimation based on logging gaps provides a stronger proxy for harvests than road-related disturbances and enables annual monitoring of harvested volume. In addition, the method allows detection of illegal encroachment by third parties through existing road networks. This advance offers a scalable, cost-effective tool for verifying logging operations, with important implications for sustainable forest management and international policy initiatives such as FLEGT and the EU Deforestation Regulation. • We demonstrate remote estimation of harvested timber volume using Sentinel-1. • We relate disturbance area to harvested volume in 93 concessions in the Congo Basin. • Our approach allows for annual instead of harvest cycle-based monitoring. • Cross-validation shows R 2 values as high as 0.83. • Our method outperforms alternative alerting and annual change products.