A coupled optimized hedging rule-based reservoir operation and hydrodynamic model framework for riverine flood risk...

Area of Science:

• Environmental science
• Hydrology
• Climate change adaptation

Background:

• Climate change and human activities disrupt hydrologic stationarity, challenging reservoir operations during floods.
• Nonstationary conditions and complex flood dynamics necessitate improved reservoir management strategies.

Purpose of the Study:

• To develop and evaluate hierarchical reservoir operation frameworks for managing downstream flood risk under nonstationary conditions.
• To integrate advanced forecasting and hydrodynamic modeling for enhanced reservoir rule curves.

Main Methods:

• Developed single-stage hedging (1SH) and two-stage hedging (2SH) reservoir operation models optimized with Particle Swarm Optimization (PSO).
• Coupled PSO models with HEC-RAS-2D hydrodynamic modeling to assess flood risk at downstream critical levees (DCL).
• Utilized SWAT-Pothole and Wavelet-based Bidirectional Long-Short-Term Memory (WBiLSTM) for inflow forecasting with GFS weather data.

Main Results:

• The 2SH-PSOH2D framework demonstrated the highest efficiency in flood risk reduction.
• Achieved an average 21% reduction in peak flow depth and a 22.28% reduction in levee failure risk at the DCL station.
• Validated the effectiveness of the proposed models using the Rengali reservoir on the Brahmani River.

Conclusions:

• The developed 2SH-PSOH2D framework offers an effective approach for managing high flood events and reducing downstream flood hazards.
• The proposed methodology can be applied to other reservoir systems globally to enhance flood management and mitigate risks.