Disentangling Particle Composition to Improve Space-Based Quantification of POC in Optically Complex Estuarine and Coastal Waters.

In estuarine-coastal-shelf seas, particulate organic carbon (POC) shows the highest turnover rates of any organic carbon pool on the planet, playing a key role in the biological carbon pump. Compared with open ocean, estuarine and coastal waters are affected by large river inputs and show high hydrodynamic variability, which results in a mixture of diverse particles that includes inorganic mineral particles, living algal particles, and organic detritus. The highly complex and variable particle compositions in estuarine-coastal-shelf waters pose significant challenges in assessing their distinct roles in the carbon cycle and total POC. To overcome challenges, we collected biogeochemical and optical in situ data from 2014 to 2020 in estuarine-coastal-shelf waters of eastern China, which is one of the largest estuarine-coastal-shelf systems in the world, to develop an algorithm that can optically discriminate particle composition and estimate their respective contributions to POC. The algorithm combines the quasi-analytical algorithm and the semi-empirical radiative transfer algorithm to estimate total suspended particle concentrations and the mass fraction of organic particles from which both phytoplankton- and detritus-related POC fractions are derived. Compared to existing POC algorithms, this algorithm shows improved retrievals compared to in situ counterparts, with $r^{2}$ and root mean squared error (RMSE) values of 0.84 and $16.57~\mu \text{g}~\text{L}^{-1}$ , respectively. The algorithm is also applied to Sentinel-3/ocean and land color instrument (OLCI) images for the year of 2020. Applying the particle component discrimination method can enhance our understanding of the roles of different particle compositions in coastal carbon cycling affected by strong land-sea exchange.