Particulate organic carbon (POC) is a central component of the ocean carbon cycle, yet its accurate retrieval from ocean colour satellites remains poorly constrained in optically complex, river-influenced coastal waters. The western Bay of Bengal (BoB) receives substantial freshwater, sediment, and nutrient discharge from major monsoon-fed rivers, including the Mahanadi, altering the region's optical properties and challenging global POC algorithms tuned primarily for open-ocean conditions. Despite this, comprehensive validation of satellite-retrieved POC in Indian coastal waters has not previously been undertaken.
This study evaluates the performance of widely used POC retrieval algorithms - the band-ratio model of Stramski et al. (2008) against an in-situ dataset from the estuary-coastal continuum off Paradip, at the mouth of the Mahanadi River, western Bay of Bengal. Monthly field sampling (n = 83) was conducted across contrasting monsoon and river-discharge regimes, comprising physicochemical parameters (temperature, salinity, dissolved oxygen), inorganic macronutrients, total suspended matter, and phytoplankton taxonomic/pigment composition, alongside remote-sensing reflectance measurements. In-situ POC ranged from 134 to 666 mg m⁻³ (mean ± SD: 329.5 ± 116.09), with the highest concentrations observed during post-monsoon and the lowest during pre-monsoon.
Global POC retrieval algorithms were applied to in situ hyperspectral remote sensing reflectance (Rrs) measurements and evaluated against laboratory-derived in situ POC concentrations. The Stramski et al. (2008) algorithm showed limited agreement with measured POC (R² = 0.2333, RMSE = 175.22 mg m⁻³, bias = -24.86 mg m⁻³), indicating poor transferability of globally calibrated algorithms to the optically complex coastal waters of the western Bay of Bengal. The positive relationship between the POC:chlorophyll-a ratio and river discharge (r = 0.57, p < 0.05) indicates a discharge-driven seasonal shift from phytoplankton-dominated to detritus-dominated particulate organic matter pools - a signal that would be systematically misrepresented by an uncorrected global algorithm.
These results provide the first regionally validated assessment of POC retrieval accuracy for the western Bay of Bengal and highlight the environmental controls on river discharge, suspended sediment, and coloured dissolved organic matter that govern satellite algorithm performance in this understudied, anthropogenically influenced coastal system. The findings offer a basis for developing a regionally tuned POC algorithm and contribute a benchmark in situ dataset to support long-term ocean colour climate data record efforts in monsoon-influenced tropical coastal seas.