Close Window

Florida State University: "Coastal winds and their effects on convection"

Final Report

Florida, as well as many other semi-tropical locations along the Gulf and Atlantic Coasts, receives much of its annual precipitation during the warm season. Convection usually is initiated by mesoscale coastal winds such as land/sea breezes. Although these circulations have been studied extensively, they have not been adequately monitored or described from a local forecasting perspective. Thus, improved summertime precipitation forecasting is considered to be the greatest local forecasting challenge. The original objective of this Cooperative Project was to assess the capabilities of the WSR-88D radar, combined with high-resolution GOES imagery, for investigating coastal wind regimes in a semitropical environment. Specific project tasks included: a) preparing a detailed climatology of coastal winds over the Florida panhandle using conventional data and radar and satellite products; b) establishing the structure and evolution of the land/sea breezes in the area; c) and developing aids for forecasting coastal winds and their associated convection. Related issues include forecasting severe weather and minimum temperature forecasting. The primary findings of the project were:

1) There are complex interactions between the sea breeze, bay and river breezes, and outflow boundaries formed by convective cells. These interactions were picked up very well by the WSR-88D and can be used to forecast areas of new development. They can also be used close to the radar site as an aid in forecasting wind shifts for maritime and coastal interests.

2) In order to better define the location of the sea breeze, it is suggested that “clear air mode” be utilized more frequently during radar operations. Under present guidelines, the WSR-88D must be operated in precipitation mode when a threshold reflectivity is exceeded within range of the radar. There are also interagency agreements which might conflict with the desire to switch to a routine basis and be of more use in future studies with goals of establishing a sea breeze climatology or other details of inland penetration and mesoscale interactions.

3) Used in conjunction with high-resolution satellite imagery (in time and space), the forecaster/analyst has a powerful combination of remote sensing technology to aid the understanding of mesoscale flow patterns. In most cases, the data were complementary, and there are situations where the satellite imagery aids interpretation of radar data and vice versa.

4) Utilizing a two-year archive of GOES 1 km visible imagery and radiosonde data, it was determined that certain flow patterns and vertical moisture distributions were more conducive than others to convection near Tallahassee. In addition, the 11 a.m. surface-based Lifted Index was found to be especially well-correlated with afternoon convection. These results have provided guidelines that are being used by the WFO to predict summer thunderstorms in the Tallahassee area.

5) Minimum temperatures at the Tallahassee official reporting site (TLH) have been studied extensively, and the results showed that TLH is colder than one would expect based on large scale geography. More specifically, TLH is most likely to be the coldest site when overall area temperatures are lowest. In another study, an in-depth comparison of the measurements from the official HO-83 station and an NWS-type minimum thermometer housed in a standard cotton region shelter suggest that the HO-83’s temperatures are 1-2 degrees F warmer than those from the “lower tech” instruments. Additionally, satellite image maps of early morning temperatures over the TLH area were created from GOES data. The existence of comparatively cold/warm areas and their evolution during the morning were also analyzed by using radiative transfer theory to obtain a relation between “split window channel” brightness temperatures and the surface skin temperature. Results showed that there were indeed regions of comparatively warm/cold temperatures in the Tallahassee area. Finally, sixteen days were characterized by a combination of weak synoptic scale forcing and clear nights for which a one-dimensional model of the planetary boundary layer (PBL) might be useful. The model outperformed the NGM and LFM model output statistics on these nights, indicated that the model is useful for predicting situations where PBL structure is the major factor in determining surface parameters.