Surface Wind Reconstructions for Hurricane Michael at Landfall with a New Parametric Model

On the 10^th of October 2018, nearly Category-5 Hurricane Michael struck the Florida Panhandle Gulf of Mexico coast near Panama City with maximum sustained winds estimated to be 135 kts (70 m/s). Two days prior to landfall, Michael was a minimal Category-1 storm with maximum sustained winds of 70 kts when a period of rapid intensification commenced in which winds increased by 65 kts over the 2 days and ended at landfall. At that time, Michael was forecast to strike as a 100 kt, Category-3 storm. Using a new parametric model of the tropical cyclone surface wind field for real time and stochastic hazard risk assessment developed at AIR Worldwide (AIR), the impact of the forecast uncertainty on the full wind field at landfall and inland is examined. A unique component of this model explicitly quantifies the variability of model parameters from a mean climatological relationship with intensity, location, and vertical shear. A statistical model for these “error” terms represents the full range of natural structural variability captured in the wind data set upon which the model is constructed. This enables pairing the new model with formal Bayesian data assimilation methods for real time application. Assimilated surface wind observations include aircraft-based (SFMR, GPS dropwindsonde, Coyote UAV), coastal (CMAN), and land-based (ASOS, state-operated mesonets, research towers). The impact of overland friction on the local wind, derived from high resolution land use/land cover data, is also explicitly forward modeled. In this talk we will present the evolution of the expected impact of Michael at landfall as the intensity changed, both in the forecasts and the advisory analyses. Additionally, we will present the resulting impact on estimates of insurance risk as the landfall situation rapidly evolved by coupling the wind field model to an insured-loss simulation.