Study of inland flooding caused by tropical cyclones using TELEMAC at South Florida

Abstract:

A new extremely high-resolution (5 meter resolution) model is developed to encompasses the urban coastal region near Downtown Miami along the Miami River, based on open Telemac library developed by Électricité de France (EDF). This high resolution triangular unstructured grid combined with a high-resolution Lidar elevation dataset permits a street-by-street focus to freshwater inland inundation modeling. Robust flooding and drying of land in the model physics, which are based on the full non-linear shallow water equations without heuristic approximations, allows for the dynamic prediction of flood extents and inundation depth across land features during inundation events. The freshwater overland flooding inundation model was forced by rainfall from a number of different scenarios with 20, 50, 100, and 200mm rainfall imitating that due to a range of tropical cyclones. Flood maps corresponding to the above extreme rainfall events have been generated for an area near Downtown Miami. Preliminary validation of the simulated flooding was carried out by comparison with in-house runs of the USACE HEC-RAS specialist overland flood model.

If exploited by policy makers, this study has the potential to impact policy making decisions in a positive fashion. The work provides further improvement to what is already one of the most, advanced depth-averaged storm surge models in existence. In particular, improvements to the bathymetry in the South Florida basin, that include the automatic unblocking of rivers (a phenomenon caused by interpolation from bathy data with insufficient resolution) as well as the development of 5m resolution sub-grid that allow for the detailed simulation of the rain induced flooding of downtown Miami. Sub-grid models can be developed for any areas for which sufficient Lidar data is available. The ability of the model to combine tides, storm surge caused by hurricane wind, and over land flooding due to extensive rainfall (including at locations many miles inland) provides policy makers in South Florida with a unique modeling tool. The model can predict street level flooding developments included inundation extent, maximum flooding depth, duration of the flooding, and water velocity at given locations.