<![CDATA[The Penunggul mangrove ecosystem has undergone continued rehabilitation, but the extent to which recent spatial changes reflect changes in aboveground carbon remains unclear. This study evaluated changes in mangrove extent and canopy-density classes between 2021 and 2025, estimated aboveground carbon stock (AGC) using the InVEST model and field-based allometry, and examined the relationships of NDVI and field-measured canopy cover with plot-level AGC. Sentinel-2 imagery was classified using a Random Forest algorithm, while vegetation structure and canopy cover were measured in ten field plots. Mangrove area increased from 18.94 ha in 2021 to 19.95 ha in 2025 (an increase of about 1.01 ha, or 5.33%), accompanied by an expansion of the high-density canopy class from 16.62 to 18.37 ha. The classification produced apparent Overall Accuracy and Kappa coefficients of 1.00 in the validation dataset. Field measurements recorded tree and sapling densities of 3,460 and 1,100 individuals ha⁻¹, respectively, with Rhizophora mucronata as the dominant taxon. Field-derived AGC was 71.92 Mg C ha⁻¹, whereas the 2025 InVEST estimate reached 280.10 Mg C ha⁻¹, indicating substantial overestimation by the class-based model. NDVI showed virtually no relationship with field-derived AGC, while canopy cover exhibited a positive linear association that was not supported by the Spearman correlation and should therefore be interpreted cautiously. These findings indicate that Sentinel-2 and InVEST are useful for assessing broad spatial patterns, but field measurements and spatially matched validation remain necessary for reliable local carbon estimation.]]>
Copyrights © 2026
