Close Window

Saint Louis University: "Distinguishing radar characteristics of tornado-producing and tornado-neutral mesocyclones associated with the Tropical Cyclone Floyd"

Final Report

SECTION 1: Project objectives and accomplishments

1.1 Objectives

The primary objective has been to examine the radar imagery of mesocyclones associated with the landfalling TC Floyd (1999) in North Carolina and list the characteristics of those that produced tornadoes and those that did not. The radar characteristics have been: a) echo top, b) maximum reflectivity, c) mean rotational velocity, d) diameter of the mesocyclone and e) shear. These characteristics are tabulated as a function of elevation angles at 0.5 and 1.5 degrees. Some characteristics based on conventional data and Velocity Azimuth Display (VAD) such as Convective Available Potential Energy (CAPE), Storm Relative Helicity (SRH), Bulk Richardson Number (BRN) are also noted. From this list, we will glean the distinguishing characteristics.

1.2 Accomplishments

An accomplishment has been to study in detail several mesocyclones associated TC Floyd in North Carolina on 15 September some of which produced tornadoes and some of which did not. The evolution of four mesocyclones, which produced tornadoes, was studied in detail depending on the availability of data. Storm Data (NOAA, 1999), which archives the tornado reports was our record of occurrences of tornadoes. Similarly, the characteristics of two mesocylones which had relatively longer lifetimes but which did not produce tornadoes were studied also. The distinguishing features of these two groups, which produced tornadoes and which did not, were noted.

A short report entitled, "Features of Mesocyclones, Tornadic and Non-tornadic Associated with TC Floyd (1999)" is attached.

1.3 Experience gained

a. While requesting archived radar data, several months should be allowed to obtain the data from NCDC on 8 mm tapes.

b. A tropical cyclone such as Floyd is prolific in producing mesocyclones. Only very subtle differences exist between the mesocyclones producing tornadoes and those not producing tornadoes. These differences are difficult to discern from radar data alone. Some results based on a local numerical model, such as MM5 should be used.

SECTION 2: Summary of university/NWS exchanges

2.1 Accomplishments of the University participant

GV Rao gave a seminar at NWS WFO St. Charles, Missouri on January 12, 2001 on TC tornadoes. Several NWS forecasters and three graduate students from SLU attended the presentation. Discussion on the differences between TC tornadoes and the midwestern tornadoes took place. The forecasters became familiar with the rich variety of mesocyclones present in the tropical cyclones.

During June 18-22, 2001, GV Rao and the three graduate students working on the Partners and related projects visited SPC, NSSL and the U of Oklahoma. Rao gave a seminar on the radar characteristics of tornadoes associated with TCs Frances (1998), Danny (1997) and Floyd (1999). Some of the graduate students also presented their preliminary numerical modeling and other work and discussed with the meteorologists at SPC. The graduate students practiced some meteorological analysis at SPC.

During January 6-11, 2002, GV Rao and his graduate student Daniel Gallagher visited Florida. GV Rao gave a seminar on TC tornadoes at NWS WFO Melbourne on January 7 and discussed at length with Bart Hagemeyer, David Sharp and Scott Spratt about the difficulty of delineating TC mesocyclones producing and not producing tornadoes. On January 9, GV Rao gave a seminar at the Tropical Prediction Center NHC and NWS Miami Florida on TC tornadoes. He discussed in detail with Max Mayfield, Ed Rappaport and Stacy Stewart on the importance of boundaries in generating TC tornadoes. The NHC scientists made many valuable comments on the SLU work. The next day GV Rao and Dan Gallagher participated in a seminar at NHC and discussed with the scientists of Hurricane Research Division the dropsondes in Floyd and how to access the data. This discussion helped SLU to use additional data in Floyd's research.

2.2 Accomplishments of NWS participants

The SPC forecaster R. Edwards visited Saint Louis University (SLU) on April 17, 2001 and gave a 2-hour workshop on mesoscale analysis. About eight graduate students, five undergraduate students and some faculty members attended. The attendees were given a set of maps for case study. Mesoanalysis of a tornado outbreak in the Carolinas was performed. The isobar, isotherm, isodrosotherms, frontal boundaries and wind shift boundaries were analyzed. Using the satellite imagery six-hour short range forecasts of severe weather were made and discussed.

The next day he and G. V. Rao discussed the difficulty of verification of tornado occurrences in North Carolina because of the mass evacuation of some counties when Floyd caused flooding. G. V. Rao and one of his students, Anthony Richey showed some radar imageries of mesocyclones of TC Floyd in North Carolina. Storm Data is one single source that gives the occurrence of tornadoes and the strategy of obtaining additional information was discussed. The accuracy of the time of occurrence was also a matter of contention in some instances.

During the SLU visit to Norman, Oklahoma in June 2001 Drs. J. Schaefer and M. Baldwin of SPC and Dr. R. Brown of NSSL spoke to SLU, respectively about parameterization of convection and radar strategies for detecting small mesocyclones such as the ones found in tropical cyclones. R. Edwards explained mesoscale convective discussions during a TC's landfall. He examined the structure of mesocyclones in TC Floyd and noted the difficulty of delineating the ones producing tornadoes from those not producing tornadoes. He saw the importance of a surface trough, extending from the center of Floyd northwards to North Carolina as a mechanism in intensifying mesocyclones occurring in an outer rain band as they traveled landwards from the sea.


GV Rao and his students became aware of the existence of numerous mesocyclones tornadic and nontornadic in an outer rain band of Floyd.

Benefits to the NWS partner

Although some delimiting features were ascertained, the general lack of distinct delineators between tornadic and nontornadic mesocyclones in the environment of TC Floyd -- a very prolific mesocyclone producer-- illustrates the difficulties field forecasters face in the warning mode. However, for SPC forecasting purposes, these nebulous characteristics indicate a need to monitor maintime radar trends upstream from potential landfall zones of peripheral rainbands for mesocyclone development, and if present offshore, to be prepared to provide mesoscale guidance to the possibility of tornadoes, in the form of mesoscale discussions (MDS) and watches covering coastal/inland target areas in the 2-6 hour nowcast time frame.

For TC Danny, the presence of mesocyclones and tornadoes in the landfall and exit phases illustrates the need for continual vigilance or residual TC circulations as they meander inland especially if the potential exists for them to access a second surge of maritime inflow air (originating from the Atlantic) as in Danny and other exit-tornado events documented by Edwards (1998). Radar derived wind trends and detailed surface analysis can help greatly here and alert the SPC forecaster to the need for an MD or watch as instability and wind trends increase, surface boundaries become evident in analysis, and mesocyclones begin to become evident. This research has, in particular, shown the importance of inflow airmass quality (for exiting systems) and interactions between outer rain bands and kinematic or baroclinic boundaries at the surface.


AMS presentations

Radar derived structures of tornadic mesocyclones from tropical cyclone (TC) Frances (1998) in Texas and Louisiana-- Rao, Scheck and Edwards. AMS 25th Hurricanes and Tropical Meteorology Conference, May 3, 2002 San Diego, CA. This preprint is enclosed.

Structures of tornadic mesocyclones of TC Danny (1957) in Alabama and South Carolina. Edwards, Rao and Scheck . AMS Severe Local Storms Conference, August 2002 San Antonio, TX. This preprint will be prepared in May 2002.