A surrogate-assisted calibration framework for computationally expensive urban fluvial flood models

Key Points

• Improved calibration methods enhance model accuracy and forecasting for urban flooding scenarios.
• A reduction in computational cost was observed, showcasing a more efficient surrogate model process.
• The framework integrates simulation data to optimize fluvial dynamics predictions in urban areas.
• This approach may enable broader application of flood models, aiding in city planning and disaster management.