Thursday, 21 April 2016: 8:00 AM
Ponce de Leon C (The Condado Hilton Plaza)
While the processes that are responsible for tropical cyclone (TC) intensity change are fairly well understood, at least from an idealized perspective, it is not clear how predictable this process should be. On one hand, theory suggests that the convection, which drives the secondary circulation, should only be predictable for hours. By contrast, statistical TC intensity models, whose inputs are parameters related to the large-scale environment, exhibit skill out to five days, suggesting a larger-scale control of TC intensity change. Not surprisingly, there are some cases where the rapid intensification of a TC are well predicted by both statistical and dynamical models, but others where it is uncertain whether a TC will undergo significant intensification. As a consequence it is of interest to determine under what large-scale environment and vortex properties TC intensity change is more or less predictable.
This study explores the vortex and large-scale parameters that are associated with reduced predictability of 48 h TC intensity change using a sample of 700 sets of ensemble forecasts from the quasi-operational HWRF ensemble prediction system from 2014-2015. Here, predictability is defined as the standard deviation in the maximum wind speed. Cases with the largest intensity standard deviation are then compared to cases characterized by relatively small intensity standard deviation. Preliminary results suggest that cases characterized by the highest standard deviation in 48 h maximum wind speed are small, weak TCs that are close to the time of genesis. Moreover, TCs that are initially far from their MPI also tend to exhibit greater ensemble standard deviation, though there is no relationship to SST or vertical shear. The remainder of this talk will evaluate potential mechanisms that explain these results.
![](images/flag.png)
- Indicates paper has been withdrawn from meeting
![](images/blue_ribbon.png)
- Submission entered in competition