Predictive Urban Coastal Flood Modeling: Accounting for Tides, Surge, Waves and Flood Control Infrastructure

Coastal flooding has been recognized as a significant challenge for human settlement. Global sea levels are expected to rise on the order of 1 m by 2100, exacerbating already episodic flooding in highly urbanized coastal communities. Flood mapping is the first step in risk assessment and represents a critical element of flood risk management, not only to identify vulnerable property but to evaluate the efficacy and potential consequences of proposed adaptation strategies. However, flood maps have not been rigorously validated to include information about model sensitivities and uncertainties relative to modeling methodology and infrastructure resolution. A two dimensional Godunov type hydrodynamic model that solves a local Riemann problem to accommodate weir-like overflow is successfully applied to simulate tidal flooding. The shallow water hydrodynamic model is then augmented with empirical overtopping models to investigate the effect of wave overtopping volumes. Simulation results are compared to two unique validation datasets and three critical issues in coastal flood prediction emerge; the effects of methodology (i.e. equilibrium vs. hydrodynamic), depiction of both weir-like overflow of walls and wave runup and overtopping volumes and finally, resolving flood control infrastructure and mitigation measures. Hydrodynamic modeling methodologies responsive to flood control infrastructure are found to outperform existing equilibrium flood mapping methodologies which ignore hydraulic connectivity and assume instantaneous filling of the backshore. Weir-like and wave runup and overtopping volumes are shown to significantly influence flood prediction, and paradoxically, mitigation strategies associated with weir-like flooding may exacerbate runup and overtopping floods. Finally, results show that accurate flood mapping requires inclusion of both flood control infrastructure such as drainage and flood defense walls resolved with centimetric accuracy. A real time kinematic (RTK) survey with approximately 1.5 cm vertical root mean square error (RMSE) is found to be suitable for barrier height measurement. However, errors consistent with aerial laser scanning (LiDAR) of approximately 15 cm RMSE are inadequate for urban flood mapping.