Assessing Physics-Informed Neural Networks in Streamflow Predictions for Tropical Monsoonal Basins in Sri Lanka
Globally, escalating floods and droughts demand accurate streamflow prediction for effective water management and early warning. Yet, this remains challenging in tropical monsoon regions. Traditional process-based models (e.g., HEC-HMS) underperform given extensive data needs, while purely data-driven models (e.g., LST...