Advancing real-time coastal data monitoring: Bio-optical property analysis (chlorophyll-a and TSM) in the Northern Bay of Bengal using Sentinel-3 OLCI, IRS Oceansat-3, and artificial neural networks.
Journal:
Marine pollution bulletin
Published Date:
Jan 9, 2026
Abstract
Monitoring bio-optical attributes in estuarine environments is crucial for understanding ecosystem productivity, sediment dynamics, and climate-induced hydrological changes. This research introduces an integrated multi-sensor and machine learning approach to assess the spatiotemporal variability of Chl-a and TSM within the Northern Bay of Bengal. In-situ data gathered during pre-monsoon, monsoon, and post-monsoon seasons were integrated with satellite observations from Sentinel-3 OLCI and Oceansat-3 OCM to investigate seasonal trends and to train ANN models for enhanced parameter retrieval. Pre-processing intricacies included cloud masking, radiometric calibration, atmospheric correction, and spectral band selection to create matchup datasets within ±6-h intervals using 3 × 3 pixel spatial averaging. ANN models were developed and validated using stratified seasonal datasets with Monte Carlo cross-validation. Assessment of performance utilized R2, RMSE, and MAE metrics. The results highlighted significant seasonal variability, with moderate Chl-a (mean 7.08 mg/L) and TSM (mean 23.35 mg/L) during the pre-monsoon, notable monsoon peaks driven by nutrient input and sediment resuspension (Chl-a reaching up to 21.01 mg/L and TSM up to 87.71 mg/L), followed by substantial post-monsoon decreases attributable to tidal flushing and stabilization. Sentinel-3 provided finer spatial gradients, while OCM offered regional consistency. ANN regression markedly enhanced retrieval accuracy compared to conventional models, achieving R2 > 0.90 with RMSE <2.5 mg/L for Chl-a and RMSE <9 mg/L for TSM in monsoon conditions. This integrated method showcases a scalable framework for satellite-based monitoring of estuarine water quality. In addition to its scientific importance, the findings contribute to advancing SDG 6 (Clean Water), SDG 13 (Climate Action), SDG 14 (Life Below Water), and SDG 15 (Life on Land), aiding in effective management and climate resilience for the northern Bay of Bengal.
Authors
Keywords
No keywords available for this article.
