Comparison of Sentinel-1 SAR and Sentinel-2 for Assessing Mangrove Aboveground Carbon Stock in Pangpang Bay, East Java, Indonesia

Abstract

Mangroves are among the plants that effectively absorb CO₂ compared to other terrestrial vegetation. Monitoring their Aboveground Carbon Stock (AGC) in a specific area is essential for conservation and egological management. This research employed optical and radar imageries to estimate mangrove AGC. The radar can penetrate clouds and vegetation canopies, which will further be compared accordingly. Mangrove AGC in Pangpang Bay Mangrove Forest, East Java Province, Indonesia, was assessed using the polynominal regression model. The findings showed that the band 2 of Sentinel-2 exhibited the maximum accuracy level at 74.40%, accompanied by a standard error of 20.28 ton/ha. In this research, the VV (Vertical-Vertical) backscatter variable, identified as the most effective Sentinel-1 independent variable for estimating carbon stock, had an accuracy rate of 58.54% with a standard error of 32.84 ton/ha. Furthermore, the VV backscatter variable exhibited a higher determination coefficient (R² = 0.34) compared to band 2 (R² = 0.24), hence serving as the most effective predictor of AGC variations.