Thia Griffin-Elliott

Thia Griffin-Elliott

Published on: September 13, 2021

to Uncategorized

Study showing the impact of turbulence in computer forecasts of hurricanes published in The Journal of the Atmospheric Sciences

Summary:  Turbulence is made up of random and continuously changing wind.  It is important in tropical cyclones because turbulence in the lowest 1-2 km of the atmosphere (the planetary boundary layer or PBL) and in clouds affects tropical cyclone intensity and structural change.

Meteorologists use computer models to forecast the weather, including tropical cyclones.  These models forecast the weather on grids with the distance between grid points typically much larger than turbulent circulations or eddies in the PBL, so the model cannot predict turbulence. Thus, turbulence is typically estimated in the models, what we call parameterization, using techniques known as PBL schemes. One important measure of the intensity of the turbulence is turbulence kinetic energy (TKE), the amount of energy in the turbulence, which is the foundation of many PBL schemes. 

Current schemes assume that the turbulent winds are the same in all horizontal directions (homogeneity) and neglect how the steady, non-turbulent wind moves the turbulence (advection). However, we know from observations that the PBL in tropical cyclones is not homogeneous, especially in the eyewall, but we do not have observations of how the eddies advect, or push, momentum upward or downward, and this effect has not been studied before.  This study investigates the effect of advection of TKE on TC simulations using one particular PBL scheme, the Mellor-Yamada-Nakanishi-Niino scheme, where we could turn the advection on and off.  We ran two sets of simulations, one includes TKE advection (CTL) and one that does not (NoADV).

TKE (shading, m2 s–2) averaged around the TC center during 20 h of the simulation for the CTL experiment that includes TKE advection (a) and the NoADV experiment that excludes TKE advection (b). These show the largest values of TKE in the eyewall, and the black line marks how far from the center the maximum wind speed at each level occurs. Vertical profiles of how much TKE is produced or dissipated in the TC eyewall, showing the turbulent transport term (trans), the sum of the shear production and dissipation (shear+diss), buoyancy production (buoy), and the advection (tkeadv) for (c) CTL and (d) NoADV experiments.

■ Important Conclusions:  

  1. For the first time, we found that the advection of TKE must be included in TC models.
  2. The observed large TKE values above the PBL in the eyewall (see figure) are predominantly due to vertical advection of TKE. It had been assumed that the TKE was created locally by the buoyancy that drives the convection there.
  3. The simulation including TKE advection produces a slightly stronger and smaller TC than the one excluding TKE advection. Differences in the two sets of simulations are mostly due to the vertical advection of TKE.

For more information, contact aoml.communications@noaa.gov. The full article can be accessed at https://doi.org/10.1175/JAS-D-21-0088.1.

The authors acknowledge high-performance computing support from Cheyenne (doi:10.5065/D6RX99HX) provided by NCAR’s Computational and Information Systems Laboratory, sponsored by the National Science Foundation. Xiaomin Chen is supported by a NRC Research Associateship award. George Bryan is supported by the National Center for Atmospheric Research, which is a major facility sponsored by the National Science Foundation under Cooperative Agreement No. 1852977, and by Office of Naval Research grant N00014-20-527 1-2071.