A novel hybrid dual-stream deep learning architecture integrating multi-state LSTM, CNN, and multi-head self-attention for water level and discharge prediction (Missouri River Basin, USA)

The Missouri River Basin, located between the states of Nebraska and Iowa, USA. Water level (WL) and discharge (Q) are key hydrological variables, accurate prediction of which in both long-term and short-term (extreme events) scenarios is essential for water resources and flood risk management. We propose a novel hybrid deep learning architecture, CNN-NLSTM-SA, which integrates a Convolutional Neural Network (CNN) branch for extracting local features and a Multi-State LSTM (NLSTM) branch for capturing long-term temporal dependencies. NLSTM, with its child-parent structure, enhances memory propagation and mitigates vanishing and exploding gradients. The outputs of these two branches are fused through a multi-head Self-Attention (SA) mechanism, enabling the model to automatically emphasize the most informative representations. The proposed model is evaluated for both long-term and short-term forecasting scales. The long-term scenario leverages extensive historical data to provide a large set of training data, whereas the short-term scenario focuses on extreme events with limited training samples. To mimic real-world operational challenges in poorly gauged or data-scarce basins, the model is also tested under varying station-availability conditions using a Leave-n-Station-Out (LnSO) validation strategy. An ablation study comparing CNN-NLSTM-SA with several single- and dual-branch alternatives (CNN-LSTM-SA, LSTM-SA, CNN-SA, CNN-LSTM) shows the superior performance of the proposed architecture. Overall, CNN-NLSTM-SA demonstrates strong potential for WL and Q prediction in both data-rich and data-limited environments. • A novel deep learning architecture, termed CNN-NLSTM-SA, is proposed. • It integrates Convolutional Neural Networks (CNN) and multi-state LSTMs (NLSTM) through a multi-head Self-Attention (SA) mechanism. • The proposed model is evaluated for long-term forecasting using extensive historical data. • Short-term forecasting performance is also assessed, with a focus on extreme events. • An ablation study with several single- and dual-branch alternatives demonstrates its strong predictive performance.