15C.6 Analysis of NRL COAMPS simulated boundary layer of Hurricane Isabel (2003)

Friday, 14 May 2010: 9:15 AM
Arizona Ballroom 10-12 (JW MArriott Starr Pass Resort)
Shouping Wang, NRL, Monterey, CA; and Y. Jin, P. Black, and J. Zhang

The objective of this study is to understand and evaluate NRL COAMPS simulated atmospheric boundary layer structure of Hurricane Isabel (2003) using in-situ observations obtained during CBLAST field experiment. A tropical cyclone (TC) version of COAMPS is configured to include three nested grids (45km, 15km, and 5km) with 15/5km nests moving with TC. The COAMPS features relevant to TC boundary layers include a turbulent-kinetic energy (TKE) based dissipative heating parameterization; sea-spray representation; and options of different mixing length formulation for a prognostic TKE turbulence scheme.

The complete TKE budget from COAMPS is derived. The results show that shear production dominates the TKE production and its values are one order of magnitude larger than the buoyancy production. The TKE advection contributes considerably to the budget near the eye wall due to the strong winds and horizontal gradients, indicating that traditional one-dimensional TKE budget may not be sufficient for TC boundary layers.

The evaluation includes three parts: mean profiles, turbulence variables such as fluxes, and turbulence generation processes. The COAMPS thermodynamic profiles in general agree with the observed, although the observations show layered structure below 750m. The magnitudes of the turbulent fluxes are comparable with those observed. The COAMPS fluxes, however, extend to the levels above the mixed-layer heights, in contrast to the observation that diminishes just above the mixed-layer height. This indicates that the turbulence parameterization gives excessive turbulence production.

A mixing length calculation based on the Bougeault convective free-path concept is implemented. Simulations are also performed to evaluate the sensitivity of TC intensity and boundary layer structure to different turbulence mixing length formulations. It is found that this sensitivity is significant. A larger mixing length enhances the simulated hurricane intensity. We are currently analyzing these results with CBLAST in-situ observations.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner