Hurricanes have been the costliest natural disaster in US history, causing billions of dollars in damage so far. Despite recent progress in modeling hurricanes, the turbulence dynamics of hurricanes is not well understood due to the lack of sufficient measurements and high-resolution simulations. In this talk, we aim to bridge this knowledge gap by using high-resolution turbulence-resolving large-eddy simulations (LESs). To this end, a novel high-fidelity LES solver will be used to reveal distinctive turbulence dynamics of hurricane boundary layers (HBLs) compared to regular atmospheric flows. Despite the unique features of HBLs, the current turbulence schemes in numerical weather prediction (NWP) models are neither specifically designed nor comprehensively tested for hurricane flows. We will present the implications of these new insights into the turbulence parameterization of NWPs by conducting real hurricane simulations in the Weather and Research Forecasting (WRF) model. Our results indicate that the current turbulence schemes in WRF are overly diffusive for simulating major hurricanes mainly since they do not account for turbulence suppression effects due to rotation in hurricanes. Reducing the vertical diffusion in the default planetary boundary layer (PBL) schemes led to ~30-40% improvements on average in hurricane intensity forecasts compared to the default models in the five considered cases. Finally, the implications of these diffusion changes on hurricane precipitation and flood forecasts will be shown by coupling the WRF atmospheric model with the WRF hydrological model. Modifying the diffusion not only impacts hurricane intensity but also modulates precipitation patterns and streamflow forecasts. Intensified hurricanes were shown to generate more intense and localized precipitation. This improved representation of hurricane dynamics led to enhanced flood forecasts in the considered cases. In total, we found that reducing the vertical diffusion led to ~16% and 34% average improvements in streamflow bias and correlation forecasts, respectively, in four landfalling hurricanes. This research provides new insights into the turbulence dynamics of hurricanes and can guide the development of more accurate models for forecasting hurricane winds and floods.

*Refreshments: 10:30 AM - 11:00 AM (Atrium)