Khozani et al. (2026) Advancing River Water Level Prediction: A Comparative Machine Learning and Deep Learning Approach

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Water Resources Management
• Year: 2026
• Date: 2026-08-18
• Authors: Zohreh Sheikh Khozani, Monica IONITA
• DOI: 10.1007/s11269-026-04826-6

Research Groups

• Department of Hydrology and Water Resources, University of Freiburg
• Institute for Computational Science, University of Zurich
• German Federal Institute of Hydrology (BfG)

Short Summary

This study developed four standalone machine learning models and a hybrid model to predict weekly water levels in the Rhine River, achieving remarkable accuracy with the hybrid LSTM-RF model outperforming standalone models.

Objective

• Investigate the efficacy of machine learning models for predicting river water levels

Study Configuration

• Spatial Scale: The study focused on the Rhine River basin.
• Temporal Scale: Data was collected between 2006 and 2024, with a weekly time step.

Methodology and Data

• Models used:
  • M5Rules
  • Random Forest (RF)
  • Sequential Minimal Optimization (SMO)
  • Long Short-Term Memory (LSTM)
  • Hybrid LSTM-RF model
• Data sources: Historical water level data from the Rhine River between 2006 and 2024.

Main Results

• The hybrid LSTM-RF model achieved a Nash-Sutcliffe Efficiency (NSE) of 0.98, significantly outperforming standalone models.
• Standalone models performed as follows: M5Rules and SMO (NSE = 0.78), RF (NSE = 0.76), and LSTM (NSE = 0.72).

Contributions

• This study demonstrated the efficacy of a hybrid machine learning model for predicting river water levels, outperforming standalone models.
• The findings have implications for effective planning and flood risk mitigation in the Rhine River basin.

Funding

• This research was funded by the German Federal Ministry of Education and Research (BMBF) under grant number 03G0866A.

Citation

@article{Khozani2026Advancing,
  author = {Khozani, Zohreh Sheikh and IONITA, Monica},
  title = {Advancing River Water Level Prediction: A Comparative Machine Learning and Deep Learning Approach},
  journal = {Water Resources Management},
  year = {2026},
  doi = {10.1007/s11269-026-04826-6},
  url = {https://doi.org/10.1007/s11269-026-04826-6}
}