Predicting Potential Fishing Zones (PFZs) is increasingly important for data-limited coastal fisheries, particularly where fishery advisories are constrained by sparse field observations, optically complex estuarine waters, and limited uncertainty validation. This study develops an integrated Sentinel-3- Artificial Intelligence-based framework for seasonal PFZ suitability prediction in the northern Bay of Bengal, with emphasis on the dynamic Sundarbans estuarine shelf. Sentinel-3 OLCI/SLSTR observations were combined with 150 in-situ measurements of chlorophyll-a (Chl-a), phytoplankton biomass, sea surface temperature (SST), nitrite, total dissolved solids (TDS), and biochemical oxygen demand (BOD) to generate a multi-source environmental predictor space. Four machine learning (ML) models, Artificial Neural Network (ANN), Random Forest (RF), Support Vector Machine (SVM), and K-Nearest Neighbors (KNN), were trained using 4700 georeferenced samples and evaluated through class-wise accuracy, ROC/AUC, confusion matrices, balanced accuracy, Cohen's kappa, Matthews correlation coefficient, spatial block validation, and external consistency with operational PFZ advisory patterns. The validation showed strong but realistic model skill, with RF and ANN providing the most stable seasonal performance, whereas SVM and KNN were more sensitive to mixed estuarine fronts and transitional PFZ boundaries. Chl-a, phytoplankton biomass, and SST emerged as dominant ecological signals, while nitrite, TDS, and BOD improved the interpretation of nutrient enrichment and water-quality stress. Post-monsoon PFZs showed greater spatial coherence, whereas monsoon suitability was fragmented by runoff, turbidity, and hydrodynamic mixing. The framework advances ocean-colour-based PFZ prediction from simple fish-location mapping toward transparent habitat-suitability intelligence for blue-economy resource sustainability.
- AI-driven ocean-colour prediction of potential fishing zones for blue-economy resource sustainability in the northern Bay of Bengal
- Ismail Mondal - University of CalcuttaSK Ariful Hossain - CSIR National Institute of Oceanography, Dona Paula, Goa, IndiaFelix Jose - Florida Gulf Coast University, Department of Marine & Earth SciencesJaved Akhter - University of CalcuttaMofareh D. Qoradi - King Saud UniversityMukhiddin Juliev - Turin Polytechnic University
- Physics and Chemistry of the Earth Parts A/B/C, Vol.144, 104655
- Elsevier
- 24
- Ongoing Research Funding program, King Saud University, Riyadh, Saudi Arabia: ORF-2026-1374 European Space Agency (ESA)
The authors extend their appreciation to the Ongoing Research Funding program (ORF-2026-1374), King Saud University, Riyadh, Saudi Arabia. Additionally, we would like to thank the European Space Agency (ESA) for providing the Sentinel-3 data, which has enabled us to extend our research work further. The authors would also like to acknowledge the use of artificial intelligence (AI) tools such as ChatGPT and Gemini for refining the English grammar, spelling, and sentence structure of this manuscript. We confirm that all content, analysis, and conclusions remain the original work of the authors.
- 99386027858206570
- © 2026 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
- Department of Marine & Earth Sciences
- English
- Journal article