Enhancing Long-Term Reservoir Inflow Forecasting: An Integrated Approach Combining Switch Prediction Method, Ensemble Rainfall Forecasts, and Machine Learning Techniques
Abstract. This study makes a unique contribution by evaluating the effectiveness of the Switch Prediction Method (SPM) in integrating ensemble rainfall forecasts, significantly improving the accuracy of long-term inflow forecasting for reservoirs—a crucial aspect of hydrological forecasting. The proposed approach combines Numerical Weather Prediction (NWP) data with advanced machine learning techniques, specifically Support Vector Machine (SVM) and Long Short-Term Memory (LSTM) models, to develop a robust forecasting framework. The study utilizes comprehensive datasets from the ShihMen Reservoir in Taiwan to assess the performance of the proposed model. The SPM dynamically integrates multiple meteorological forecasts to reduce uncertainty and improve rainfall input accuracy. These enhanced inputs are then used in multi-step forecasting (MSF) models with a 72-hour lead time. The results demonstrate that the LSTM-based model, combined with SPM-integrated forecasts, delivers accurate and stable inflow predictions. For instance, in the case of Typhoon Soudelor—the test event with the highest observed peak inflow (5,634.1 cms)—the proposed SPM-LSTM-MSF model achieved a Mean Absolute Error (MAE) of 178.8 cms and a Coefficient of Efficiency (CE) of 0.87, demonstrating superior accuracy and temporal stability compared to the SVM-based approach. These findings highlight the potential of SPM and machine learning techniques in enhancing reservoir management and flood control strategies, offering a robust and adaptable solution for complex hydrological forecasting tasks.
The manuscript addresses an important problem in hydrology and water resources management. The integration of ensemble rainfall forecasts with the Switch Prediction Method (SPM) and machine learning techniques is relevant, and the ShihMen Reservoir case study provides a solid application context. The methodological framework is clearly presented, and the paper is generally well organized. These are important strengths of the study.
Nevertheless, there are several areas where improvements would help to increase the clarity, novelty, and broader impact of the work:
1. Research novelty not sufficiently highlighted
Although the Abstract and Introduction provide a comprehensive description of the study background and methodology, the novelty of the proposed approach compared with existing methods is not clearly articulated. I recommend that the authors strengthen the concluding part of both sections by explicitly stating the uniqueness of this study and clarifying how the proposed framework improves upon conventional approaches. This will better highlight the scientific contribution.
2. Outdated references
The list of references added by the authors appears to rely heavily on older studies. A scientific article should also reflect the state-of-the-art. It is recommended to include more recent literature and to provide a review of previous research on similar issues to enhance the timeliness and relevance of the study.
3. Insufficient technical details of SPM
The Switch Prediction Method (SPM) is only briefly described, making it difficult for readers to fully understand its operational mechanism. More details are needed. This will also help ensure reproducibility.
4. Limited discussion on practical implications
The Discussion and Conclusion are focused mainly on the technical aspects of the model. However, the practical significance for reservoir operation remains underdeveloped. It is suggested to elaborate on how 72-hour inflow forecasts—even with certain errors—can still provide critical lead time for reservoir managers to adjust release strategies and enhance operational safety. This would substantially strengthen the applied value of the study.
5. Future research directions too narrowly technical
The future work outlined in the Conclusion mainly emphasizes model improvement. To broaden the impact, the authors should consider including wider perspectives such as cross-basin applicability, feasibility under climate change conditions, or hybridization with physically-based models. This would enhance the forward-looking dimension of the paper and its relevance to a broader research community.