This study focuses on an outer rainband associated with TC Earl that affected west-central Florida on 2 September 1998. Low level convergence ahead of this outer rainband and strong CAPE values exceeding 2000 J kg-1 (over Tampa) produced mesocyclones 100 km to the north of Tampa. These were detected using WSR-88D data from Tampa (KTBW) and are being studied at Saint Louis University in collaboration with the Storm Prediction Center under the UCAR-COMET outreach program.

The objective of this research is to integrate the nonhydrostatic Penn State/NCAR MM5 model, Version 3 for 24 hours given an initial map time twelve hours prior to the mesocyclone initiation. The grid spacing for the four domains employed is respectively 81, 27, 9, and 3 kilometers. The grids are two-way interactive. The finest domain covers Tampa and an area approximately 150 km to the north, where WSR-88D data showed that mesocyclones occurred. The initial data was for 1200 UTC 2 September 1998, using NCEP Global Analysis “FNL” Data. Physics options in the forecast model include the Betts-Miller cumulus parameterization, a Mixed-Phase (Reisner) microphysics scheme, the Eta (Mellor-Yamada) boundary layer parameterization scheme, and a simple cloud-radiation scheme. However, the fine-mesh domain was explicitly resolved as opposed to using a convective parameterization scheme.

Preliminary model results show an area of confluence at 925 hPa north of Tampa on the 27-km domain. The location and orientation of this area agree with the initial location and subsequent propagation of the observed mesocyclones. The time of occurrence of these features is in general agreement with that of the observed mesocyclones.

On the 3 km domain, 925 hPa streamlines valid at 1800 UTC indicate a mesoscale cyclonic circulation approximately in the same area where such circulations were detected by the Tampa radar.

At the conference, detailed characteristics of the mesocyclones bearing tornadoes such as vorticity, helicity, and reflectivity will be presented. Model-derived characteristics of mesocyclones will be compared with those detected by the Tampa radar and conventional data. A plan is underway to study the sensitivity of changing the boundary layer and cumulus parameterization schemes in order to examine the effects of such changes on the characteristics of mesocyclonic circulations.