Monitoring dissolved organic carbon concentration in coastal waters of Southeast Asia using remote sensing semi-analytical inversion models

Abstract

Southeast Asia contains most of the world’s tropical peat carbon pool. In recent decades, intensive peatland conversion for commercial activities has led to the rise in terrigenous dissolved organic carbon (DOC) concentration in coastal waters, which has many repercussions on coastal ecosystems. Optical remote sensing has emerged as an effective tool to monitor these impacts over large areas at high spatiotemporal resolutions. A recent study has applied a semi-analytical inversion model, parametrized with regional field measurements, to estimate the concentrations of DOC, colored dissolved organic matter (CDOM), and total suspended sediments (TSS) off peatland-rich Northwest Borneo. However, such studies remain lacking in other parts of Southeast Asia. To bridge this gap, we demonstrate the applicability of the aforementioned model in other coastal waters of Southeast Asia using MODIS-Aqua data. We incorporated field measurements from the Singapore Strait (2019–2020) into a regional spectral library and applied the algorithm to several parts of Southeast Asia. In NW Borneo, the model fit improved with the absolute mean relative percentage difference (MRPD) between the modelled and observed backscattering albedo decreasing by 2.72 ± 3.56 %. At the Malacca, Singapore, and west Karimata Strait, MRPD was generally small in coastal waters and estimates of DOC, CDOM, and TSS fell within the range of measurements from Sarawak and Singapore Strait. The model clearly reflected the spatial and temporal (multi-year) variability in each biogeochemical parameter. Overall, the semi-analytical model has a strong ability to monitor biogeochemical changes in Southeast Asia coastal waters beyond NW Borneo.