Komparasi Citra Multispektral Untuk Estimasi Simpanan Karbon Pada Kawasan Mangrove Nusa Lembongan

Abstract

Mangroves are vegetation capable of absorbing carbon and controlling atmospheric CO2 levels through sequestration. Remote sensing technology can estimate mangrove carbon stocks efficiently and quickly through radar and optical imagery. However, previous studies have shown that the spatial resolution of an image does not always correlate directly with the accuracy of the image. This study aimed to measure field carbon stocks, analyze the relationship between field carbon stocks and vegetation indices, and compare the capabilities of Sentinel-2 and Landsat-8 satellite imagery with different resolutions in estimating carbon stocks in Nusa Lembongan. Field research was conducted to assess and map mangrove carbon stock values. Mangrove carbon stock values were obtained using a general allometric equation for mangroves. The calculated results were then correlated with spectral vegetation indices using simple linear regression analysis. This study identified 13 mangrove species in Nusa Lembongan. Field measurements showed that the above-ground and below-ground carbon stocks were 100.81 tons/ha and 47.02 tons/ha, respectively. The relationship between Above-Ground Carbon (AGC) and vegetation indices resulted in the best R² values with IRECI (0.79) for Sentinel-2 and DVI (0.61) for Landsat-8. Meanwhile, the relationship between Below Ground Carbon (BGC) and vegetation indices showed the best indices as GCI (0.73) for Sentinel-2 and GCI (0.56) for Landsat-8. Accuracy measurements indicated that Sentinel-2 had lower error values than Landsat-8 across all vegetation indices concerning both AGC and BGC.