Comparing Sentinel-2 vegetation indices for optimal estimation of aboveground carbon stock in a tropical community forest of Nepal

Résumé fourni par la source

Accurate monitoring of forest carbon stocks is essential for effective climate change mitigation. This study aimed to identify the optimal Sentinel-2 vegetation index (VI) for estimating aboveground carbon (AGC) stock in the Raktamala Namuna Community Forest, Nepal. Field data from 53 circular plots (500 m² each) were used to compute AGC based on tree-level dendrometric measurements and species-specific wood density. Ten VIs, including traditional (e.g., NDVI, EVI) and red-edge-based indices (NDVIre1–NDVIre4), were derived from a cloud-free Sentinel-2 Level-2A image (April 7, 2023). Five regression models (linear, logarithmic, quadratic, power, and exponential) were tested for each VI–AGC relationship. The average AGC was 63.88 t·ha-¹. The red-edge index NDVIre1 (using Band 5, 705 nm), modelled with a logarithmic function, yielded the highest predictive accuracy (R² = 0.7205, r = 0.848, p < 0.001), outperforming traditional indices like NDVI (R² = 0.609). This study demonstrates the superior sensitivity of Sentinel-2’s red-edge band (705 nm) to canopy structure in dense tropical forests. The study concluded that the NDVIre1 logarithmic model provides a novel, cost effective tool for operational and scalable carbon monitoring in community-managed forests, directly supporting REDD+ implementation and localized forest management.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.