Paper on how getting turbulence correct in forecast models can improve intensity forecasts in sheared storms published in Monthly Weather Review

Summary:  Though tropical cyclones (TCs) are usually hundreds of miles across, the strongest wind occurs in gusts that can be smaller than 100 yards across, or the size of a football field. We call features that cover such a small area the “turbulent scale.” Other features that are important in forecasting TCs can be as much as thousands of miles across, what we call the “environmental scale.” We have long known that one key environmental-scale factor for forecasting intensity is the wind shear, the difference in wind between the bottom and the top of the TC. TCs usually strengthen when the shear is small, and weaken when the shear is large, but in the middle, what we call “moderate” wind shear, is where forecasts are most difficult. We usually have good measurements of the wind shear, and we can now measure, with NOAA Hurricane Hunter aircraft, how wind on the turbulent scale (called “turbulence”) changes, but, until this study, we had not yet tested whether models can make these features look like what we see in the real storms, especially in moderate wind shear. This paper showed, for the first time, the link between the different scales of motion from turbulent scale all the way up to environmental scale and their importance in forecasting TC intensity.

We used the Hurricane Weather Research and Forecasting (HWRF) model, the same model that forecasters use to forecast where a TC will go, how strong (intensity) and large (size) it will be, and where the strongest winds will be, in this study. We ran different versions of HWRF many times, changing how the model makes turbulence, to see which forecasts are closest to these observations. Then, we looked at how these changes improved forecasts of the entire TC.

Screen Shot 2019-04-09 at 2.14.13 PM.png

Important Conclusions:

1. Accurate forecasts of turbulent-scale wind near the ocean surface (what we call the “boundary layer”) are important for forecasts of how quickly the TC will intensify and how strong it will get.

2. Boundary-layer turbulent-scale wind affects where the thunderstorms are in TCs. When there is just a little turbulent-scale wind in the model, the strong thunderstorms occur all around the storm; when the turbulent-scale wind in the model is accurately forecast, the thunderstorms occur where we see them in real TCs.

3. The center of the TC is not in exactly the same place from the bottom to the top, what we call “tilt.” Low-level turbulent wind influences how the tilt changes with time. When there is the wrong amount of turbulent-scale wind in the model, the tilt is not forecast well, so the model can not accurately forecast the TC intensity.

4. How cool and dry the boundary layer is affects how strong thunderstorms are in TCs. When the tilt is too large due to too much turbulent-sale wind in the model, the atmosphere is too cold and dry, creating thunderstorms that are too weak. Since TCs get energy from the thunderstorms, this makes the TC too weak in the model compared to reality.

5. This study emphasizes the importance of the all the different scales in and around TCs to improve model intensity forecasts.


You can read the paper in its entirety at