Abstract

The reproduction of satellite-observed Remote-Sensing Reflectance (Rrs) within marine Biogeochemical (BGC) models is a key process in the development of ocean-color data assimilation frameworks. Assimilation approaches based on Rrs can avoid the uncertainties associated with satellite-derived water quality products and are therefore expected to be particularly applicable to coastal waters. However, in optically complex coastal environments, the reliability of the bio-optical models and radiative transfer processes used to reproduce the in-water light field remains limited, and the applicability of these processes within coupled BGC modeling frameworks is not yet fully established. In this study, we developed a coupled BGC–bio-optical modeling framework capable of reproducing Rrs in Tokyo Bay, an optically complex, eutrophic, semi-enclosed coastal system. To represent region-specific inherent optical properties, we conducted bay-wide shipboard observations and developed a bio-optical model that incorporates key processes, including the chlorophyll pigment package effect and chlorophyll fluorescence. This model was coupled to a BGC model tuned for Tokyo Bay. The simulated Rrs was evaluated against satellite observations from GCOM-C/SGLI and in-situ measurements from an autonomous radiometer system. The model reproduced the spatial distribution and temporal variability of Rrs, including enhanced long-wavelength reflectance during river flood events. These results demonstrate the feasibility of using Rrs as an observational constraint for ocean-color data assimilation in optically complex coastal waters, offering a more robust alternative to conventional methods.