The Practical Predictability of Storm Tide from Tropical Cyclones: A Look at Recent Storms in the Gulf of Mexico.

The active Atlantic hurricane season of 2017 has resurrected the national interest in hazards from landfalling hurricanes. As evidenced by recent storms making landfall along the United States coastline, forecasts of the potential for storm surge are critical to emergency management evacuation preparations. Yet storm surge hazards remain challenging to predict, due in large part to limits in current capabilities in hurricane prediction and the ways in which the resulting hurricane forecast errors propagate into surge predictions. This research seeks to understand the practical predictability of storm surge by using best track data and systematic perturbations to four storm parameters (track, intensity, size, and translation speed) to drive an ensemble of ADCIRC storm tide (surge + tide) simulations. The resulting hurricane-surge ensemble is used to investigate the sensitivity of storm surge inundation to these four storm parameters, and thus the sensitivity of storm surge forecasts to errors in forecasts of those parameters. To facilitate examining practical predictability at different lead times, the spread of storm perturbations was designed to represent the average error in current operational tropical cyclone forecasts. Results will be shown for experiments with three storms in the Gulf of Mexico (Ike, Charley, and a hypothetical storm Charike), analyzed using location-specific storm surge inundation metrics as well as storm-following, volume-integrated metrics of inundation. The results indicate that the practical predictability of storm surge inundation is limited foremost by errors in hurricane track forecasts, followed by intensity errors, then speed errors. Errors in forecasts of hurricane size also play an important role, especially at short lead times. Initial results from Hurricane Irma and Hurricane Harvey will also be presented.