Paper on how accurately predicting turbulence can improve forecasts of tropical cyclones published in the Journal of Advances in Modeling Earth Systems – Hurricane Research Division

Accurately predicting the distribution and transport of wind energy is important for accurate forecasting of hurricanes by computer models. This study aims to improve our understanding of the energy associated with small-scale eddies and gusts that are known as turbulence. Computer simulations were analyzed to study the distributions of turbulent energy and how hurricane structure and intensity are affected. This study suggests that enabling the realistic transport of turbulence will improve simulations and forecasts of hurricanes.

Left: TKE at 1 km altitude (shaded) in the simulation of a category-5 hurricane (top) without TKE advection and (bottom) with TKE advection. The dashed circle shows the location of the maximum TKE.  The storm motion (arrow) along with the front-right (FR), front-left (FL), rear-right (RR), and rear-left (RL) are indicated. This shows how including TKE advection moves the maximum values downwind closer to where it occurs in real TCs.  Right: TKE averaged around the TC center (shaded) divided by the maximum value, the velocity of the part of the wind that flows around the TC center (black contour), the velocity of the part of the wind that flows toward and away from the TC center (white contour), and the velocity of the part of the wind that flows upward or downward (gray contour; positive=solid line, negative=dashed line) for the category-5 simulation (top) without TKE advection and (bottom) with TKE advection. The red ‘x’ shows where the maximum TKE is. This shows how including TKE advection increases the TKE in the region of the dashed circle, so that it resembles measurements in real hurricanes.

Turbulence is rapidly changing wind across small distances.  Turbulence in the lowest 1-2 km of the atmosphere and in clouds affects TC intensity and structure by impacting how much warm, moist air can fuel the TC. The size of turbulence is sometimes less than 100 meters across, or about the length of a football field. 

We use computer models to forecast the weather, including tropical cyclones. These models forecast the weather on grids with the distance between points much larger than the size of the turbulence.  This means that the models themselves cannot predict turbulence. Thus, turbulence is estimated in the models using what we call parameterization schemes. An important measure of how much turbulence exists is turbulence kinetic energy (TKE), or the amount of energy in the turbulence. Current schemes assume that the turbulent winds are the same in all horizontal directions and neglect how the wind moves the turbulence around, a process called advection. This study investigates the effect of including advection of TKE in parameterization schemes on simulations of TCs that range from category-1 to category-5 intensity. The study uses a realistic computer model that includes both the atmosphere and the underlying ocean, and how the two react with each other. 

Conclusions:

1. The inclusion of TKE advection in the parameterization scheme pushes the region of maximum turbulent energy downwind of its location without the advection being included (left panels of figure).
2. The inclusion of TKE advection in the parameterization scheme increases the turbulence in upper levels of the eyewall where the strongest wind speeds occur and reduces it near the surface (right panels of figure). 
3. Further important changes to TKE are seen when the TC is allowed to move, and when the computer model includes how the ocean and atmosphere interact with each other.
4. The inclusion of TKE advection in the parameterization scheme therefore allows the model to predict turbulence more accurately than without the advection, so that it looks like measurements from hurricane hunter aircraft and small uncrewed aerial systems (drones). Thus, the process should be used in model parameterization schemes.

For more information, contact aoml.communications@noaa.gov. The full text can be found at https://doi.org/10.1029/2022MS003230. This research was carried out [in part] under the auspices of the Cooperative Institute for Marine and Atmospheric Studies (CIMAS), a Cooperative Institute of the University of Miami and the National Oceanic and Atmospheric Administration, cooperative agreement NA20OAR4320472. Computational time was generously provided by the Center for Computational Sciences at the University of Miami. We also appreciate program support from the NOAA Office of Marine and Aviation Operations, Uncrewed Systems Operations Center 2021 Request for Proposals. Jun Zhang is supported by NOAA Grants NA21OAR4590370, NA22OAR4590178, and NA22OAR4050669D.