PELAGIX: An Intelligent Decentralized Framework for Verified Quantification and Transparent Trading of Blue Carbon Assets

Mangrove carbon assessments require spatially explicit methods whose assumptions, data support, and verification status remain traceable. This study presents an integrated workflow for operational mangrove mapping, modeled above-ground carbon (AGC) estimation, complementary carbon accounting, parameter sensitivity, and a prototype digital measurement, reporting, and verification (MRV) provenance layer for a Sundarbans area bounded by approximately 88.10–89.10°E and 21.50–22.35°N. Sentinel-2 surface-reflectance observations from 1 January to 30 April 2023 were cloud-screened, median composited, and used to calculate NDVI; NDVI > 0.4 defined the operational mangrove mask. The supplied Google Earth Engine workflow quality-screened GEDI monthly rh100, median-composited it, applied its original rh100/100 transformation, and filled unobserved structural values using 15× NDVI. Biomass and carbon were modeled as AGB = H× NDVI× 10.5 and AGC = 0.47× AGB. The retained raster footprint contains 15,793,243 mangrove cells (157,932.43 ha under the nominal 0.01 haper-cell convention). Mean AGC density was 2.8060 t C ha−1 (median 2.7935; standard deviation 0.6325), with total modeled AGC of 443,164.99 t C and derived AGB of 942,904.24 t. A separate InVEST baseline reports 409,383 t C. Combined variation of the project-specific AGB coefficient and carbon fraction produced 343,486.54–550,027.47 t C (-22.49% to +24.11%); this is parameter sensitivity, not a confidence interval. GEE and InVEST outputs are complementary, non-equivalent model results. The prototype demonstrates evidence handling and digital provenance workflows, not carbon-credit certification or independent verification. No independent field validation was available, and reported values represent modeled above-ground carbon rather than total ecosystem carbon.