Remote estimation of harvested timber volume in the Congo Basin using Sentinel-1 data
This study uses Sentinel‑1 radar data to estimate selectively harvested timber volumes across 93 logging concessions in the Congo Basin. Canopy gap detection strongly predicts reported harvests and outperforms existing disturbance alerts, offering a cost‑effective tool for monitoring legal compliance, detecting illegal logging, and supporting sustainable forest management and international policy initiatives.
Subject Tags
- Land management
- Forest
Abstract
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 (R2 up to 0.86, bootstrap RMSE m3) 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 (R2 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.
Citation
Welsink, A. J., Fesenmyer, K., Slagter, B., Peña-Claros, M., Herold, M., & Reiche, J. (2026). Remote estimation of harvested timber volume in the Congo Basin using Sentinel-1 data. Ecological Informatics, 94, 103683.
TNC Authors
-
Kurt Fesenmyer
NCS Spatial Data Scientist. Tackle Climate Change
The Nature Conservancy
Email: kurt.fesenmyer@tnc.org