Advancing Coastal Water Quality Modeling Through Deep Learning: State of the Art and Research Roadmap

Coastal water quality (CWQ) remains a complex and pressing issue in environmental engineering that should be considered with an in-depth understanding of sustainable water management. To facilitate well-informed decision-making and the adoption of effective management techniques, regular monitoring and precise forecasting for sustainable water management are critical. This review has provided a discussion on the applications of deep learning (DL) models in this context over the period (2010-2026). Despite the remarkable potential, DL models face extensive challenges due to the limited availability and heterogeneity of high-quality data, which complicates both training and validation learning processes. The current review has discussed the benefits and constraints of the applied DL models in CWQ, their strengths in managing spatiotemporal data, complex interactions and the provision of real-time environmental monitoring solutions. Additionally, the review identified the new trends, gaps in research, and future directions to enhance the model accuracy and interpretability.