Suspended particulate matter and secchi disk depth in the Chilika Lagoon from in situ and remote sensing data: a modified semi-analytical approach

Increasing sediment load and deteriorating water clarity are the key challenges for many inland water bodies. The uncertainty associated with the bio-optical complexity of these water bodies limits remote-sensing approaches to monitor such fragile ecosystems. Therefore, we measured optical properties along with water quality parameters in the Chilika lagoon, the second-largest brackish water lagoon in the world. We evaluated exiting quasi-analytical algorithms (QAAs) to retrieve particle back-scattering coefficients (bbp) and diffuse attenuation coefficient (Kd) from surface remote-sensing reflectance (Rrs) data. However, existing QAAs underestimated the observed bbp and Kd values in the Chilika lagoon. Therefore, a modified inversion algorithm (MIA) using the combined features of the sixth version of QAA (QAA v6) and the IOPs inversion model of inland waters (IIMIW) is proposed and validated. Values of bbp at 665 nm and Kd at 490 nm were then used to estimate suspended particulate matter (SPM) and Secchi disk depth (ZSD), respectively. The bias for bbp and Kd was significantly reduced using MIA, which improved the accuracy of SPM and ZSD estimation. We further used MIA to retrieve bbp(665) and Kd(490) values from the Sentinel 2A/2B (S2A/S2B) multispectral instrument (MSI) Rrs data for each pixel over the Chilika lagoon to estimate SPM and ZSD respectively. Better root-mean-squared error (RMSE) value of 10.91 mg L−1 for SPM (ranging from 11.54 to 99.00 mg L−1) and 0.21 m for ZSD (ranging from 0.15 to 1.2 m) was observed when compared with in situ measurements (n = 77). The mean SPM trend from 90 different S2A/S2B MSI acquisition dates from 2017 to 2021 indicated that SPM concentration in the lagoon is closely related to wind velocity. Matching fish landing trends with SPM concentrations was also observed, which may open new opportunities to manage inland water bodies using remote-sensing observations.