Developing an Integrated Remote Sensing and Ground-Truth Framework for National Forest Monitoring in Nigeria

Nigeria lacks a nationally consistent, operationally robust forest monitoring system capable of generating annual deforestation and degradation estimates at the accuracy levels required for international reporting and results-based payment mechanisms under REDD-plus, representing a critical gap in the country's climate change commitments. This study develops and validates an integrated remote sensing and ground-truth framework for national forest monitoring in Nigeria, addressing methodological limitations in current national greenhouse gas inventory approaches. The framework combines multi-sensor satellite data (Sentinel-2, Landsat-8/9, and ALOS PALSAR-2) in a wall-to-wall change detection algorithm capable of distinguishing deforestation, forest degradation, and forest recovery at a 10-metre resolution. A stratified systematic sample of 4,200 ground reference plots across Nigeria's six ecological zones was used for validation, with field teams collecting tree basal area, canopy cover, and disturbance history data over two field seasons. Algorithm performance was evaluated against independent ground truth using the Good Practices Guidance for national forest monitoring. Overall accuracy for the deforestation class exceeded 94%, with degradation detection at 81%, significantly above the 70% threshold for REDD-plus eligibility assessments. A novel operational workflow integrating near-real-time alert generation with quarterly forest inventory updates is described and costed. The framework's open-source architecture allows adaptation by state forestry agencies for sub-national monitoring. Implementation recommendations are structured around the National REDD-plus Programme's existing institutional arrangements and technical capacity. Keywords: national forest monitoring, remote sensing, REDD-plus, Nigeria, deforestation detection