Answers are grounded in this paper and cite their sources.
Mangrove carbon budgets are usually derived from changes in forest extent, which cannot register carbon lost from forest that thins but remains mangrove. We partition the living-biomass carbon loss (aboveground plus belowground) of the Sundarbans over 1994–2023 between areal deforestation and in-situ canopy degradation; soil organic carbon is not resolved, and the fluxes are losses from the living-biomass pool rather than measured emissions. A Random Forest trained on the 300 m NASA/ORNL 2010 carbon product with fractional canopy cover, Landsat NDVI, canopy height and ALOS PALSAR L-band backscatter was applied to annual 30 m predictors (R2 = 0.80 at the reference scale, also under 10 km spatial block cross-validation). The domain stock showed no significant trend but a net decline of 1.72 Mt C. The remotely sensed carbon decline within persistent forest, interpreted as degradation, was 2.19 Mt C, against 0.79 Mt C from deforestation; correction for error rectification gave 1.70–2.12 Mt C, and a significance screen 0.93 Mt C (p < 0.05). Compared with the independently retrieved ESA CCI Biomass series, domain totals agreed over 10 years (r = 0.87) and declining pixels gained less biomass than other forest, although cell-level change did not agree. In GEDI lidar biomass for 2019–2023, cells dominated by significant decline carried 5.9 Mg ha−1 less biomass than cells of the same 1994 density. At 300 m only 34% of the deforestation flux was recovered, so the partition requires 30 m resolution. The contribution is the attribution of living-biomass carbon loss between the two pathways, with an explicit uncertainty range.
No takes yet. Share an insight, caveat, or question.
Melih Altay (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: