Conference Agenda

Traditional hydrodynamic flood models face significant limitations in real-time forecasting applications due to their computational complexity. High-resolution hydrodynamic flood simulations require extensive calculation times, often exceeding the skilful lead times of high-resolution short-term rainfall forecasts. This temporal mismatch proves particularly challenging for urban catchments due to their rapid response dynamics, resulting in delayed forecasts unsuitable for support of operational decision-making and impact mitigation. This study investigates a hybrid framework combining simplified hydrodynamic physics with machine learning to achieve accurate and timely flood depth mapping. While simplified models significantly accelerate computations by abstracting surface and sewer components, they sacrifice precision compared to the reference hydrodynamic simulations. To bridge this gap, a Gaussian Process (GP) regression model is trained on a comprehensive library of simulation results, enabling efficient bias correction without sacrificing computational speed. Implemented for a case study in Antwerp, Belgium, using radar-derived rainfall data, the framework demonstrated speed improvements with a factor of 30-100 depending on the simplified model configuration, while maintaining good accuracy (R² ≈ 0.85). Ongoing work expands the analysis to diverse storm events and model variants.