<![CDATA[Seagrass is one of the ecosystems that absorbs large amounts of carbon in the blue carbon ecosystem. This ecosystem also has many benefits, such as becoming the habitat for fish and improving water quality. However, there is degradation of the seagrass ecosystem that requires annual carbon inventory with Sentinel-2 which covers benthic habitat areas and has high spatial resolution. This study aims to map changes in the area and distribution of seagrass meadows, as well as to map and estimate changes in aboveground carbon stocks (AGC) of seagrass meadows in part of the coast of Bontang City, East Kalimantan annually with a range of 2020-2024. The methods used in this study include integration with field data, benthic habitat classification and AGC regression with the Random Forest (RF) algorithm, accuracy testing, and multitemporal mapping to detect changes. The results of this study indicate that the benthic habitat class is dominated by seagrass which remains from 2020 to 2024 covering an area of 2643.92 hectares and the area of seagrass tends to decrease from 2808.14 hectares to 2745.94 hectares. The benthic habitat classification used to classify the seagrass area has an Overall Accuracy of 65.9%. Meanwhile, the total estimation of Seagrass AGC decreased from 287.12 tons of carbon in 2020 to 265.6 tons in 2024 of carbon from an area of 3118.6 hectares. The model from the regression results was tested for accuracy with an RMSE of 2.09 gC/m2, R2 of 0.65, and a Plot Goodness of Fit of 1:1 which explains that the estimated AGC sample is positively correlated with the field AGC sample.]]>
Copyrights © 2026
