Adapted from Preprints, 20th AMS Conference on Severe Local Storms, Orlando, Sep 2000


Roger Edwards

Storm Prediction Center

Norman, OK

Gandikota V. Rao and Joshua W. Scheck

Saint Louis University

St. Louis, MO


Forecasting the typically short-lived tornadoes associated with tropical cyclones (TCs) is a major concern for National Weather Service (NWS) forecasters (e.g., Spratt et al. 1997, hereafter S97, and Edwards 1998a). Previous studies of TC tornadoes were largely concerned with climatological and/or synoptic aspects (e.g., Hill et al. 1966 hereafter H66, Weiss 1987, McCaul 1991, Vescio et al. 1996). Only recently did S97 and Hagemeyer (1997) use the WSR-88D radar observations to discuss the structures of tornadic mesocyclones in TCs.

Given at least marginally favorable ambient vertical shear and instability conditions, supercell tornadogenesis is largely tied to storm scale and meso beta scale processes (e.g., Markowski et al, 1998a and 1998b). In tropical cyclones, however, tornadic mesocyclones are around 1 km in diameter (S97) and often fast moving. As such, small scale conditions suitable for tornadoes escape most analytic and radar diagnoses, and still are largely inferred from the meso-alpha scale TC environment.

Hurricane Earl made landfall in the Florida panhandle on 3 September 1998 and exited into the Atlantic near the Virginia/North Carolina border the following day. Its tornado production was bimodally clustered (Edwards 1998b): about 11 hours before landfall in central Florida, and beginning about 8 hours after landfall through the early exit phase over eastern Georgia and the eastern Carolinas (Fig. 1). The latter tornadoes killed one person and caused several million dollars in damage. Rao et al. (2000) analyzed mesocyclones associated with Earl during the Florida phase. For this portion of our study, we are using conventional surface and upper air data, satellite imagery, Eta model upper air analyses and WSR-88D radar imagery to examine the environments of the tornadoes.


After identifying tornado occurrences, WSR-88D base reflectivity and storm-relative velocity imagery were examined to determine associative mesocyclone structure and morphology using WATADS (WSR-88D Algorithm Testing and Display System) software (NSSL 2000). Because the mesocyclones were about 130 km or less from the Tampa Bay FL and Charleston SC radars, their reflectivity and velocity characteristics could be well-resolved and -analyzed.

Figure 1. Path of TC Earl (1998) in six hour increments (4-point stars), locations of Tampa and Charleston WSR-88D units (five-point stars) and locations of tornadoes listed in Storm Data (NCDC, 1998) and investigated here (triangles).

Observed soundings from the rawinsonde sites nearest the tornado events in space and time were modified kinematically for observed storm motions associated with each tornadic supercell, to estimate the magnitude of storm-relative helicity (SRH, after Davies-Jones et al. 1990), a parameter commonly used in operational forecasting to prognose supercell and tornado potential.

Thermodynamic parameters were computed using a virtual temperature correction (Doswell and Rasmussen 1994), which yields the most thermodynamically representative lifted parcel characteristics, and increases CAPE values for nearly saturated boundary layer environments, such as those in TCs.


The 12-hour forecast of the operational Eta model from 12 UTC 02 September (Fig. 2) captured the asymmetry of the wind field, as well as the ascent over Florida connected with some combination of Earl's outer rainband and of the surface warm front. While the degree of ascent associated with each in the model is unknown, the maximum lift corresponded closely with the tornado occurrences (Fig. 1).

Figure 2. Eta 12-hour 850 mb forecast of omega (mbar s-1) valid 00 UTC 03 September, in the form of solid, bold lines (ascent) and dashed (descent). Light dashed lines are isotherms (C) and light, solid lines represent isohypses (dam).

In the 00 UTC 3 September Tampa Bay sounding (Fig. 3), a dry layer was evident around 600 mb, supporting the empirical associations of H66 between midtropospheric drying and TC tornadoes. Surface-based CAPE values approached 2100 Jkg-1, 1.3 times higher than, and in a similar center-relative position as, the largest composite CAPE from 1296 climatological hurricane soundings analyzed by McCaul (1991).

Figure 3. Skew-T diagram of Tampa Bay, FL, sounding and associated derived parameters from 00 UTC, 3 September 1998.

Vertical shear was also pronounced; the 0-3 km AGL SRH of 241 m2s-2 was well above the 150 m2s-2 guideline suggested by Davies-Jones et al. (1990) for tornado occurrence in nontropical supercells. In contrast, the Tallahassee sounding from the same time (not shown), north of the center of Earl, yielded CAPE and SRH of only 6 Jkg-1 and 156 m2s-2, respectively. This also supports the findings of McCaul (1991) with respect to spatial distribution of SRH maximized in the middle portion of the eastern semicircle of TCs.

Both soundings were taken around the periphery of the azimuthal northeastern quadrant of Earl, the sector identified as the most favorable region for tornado occurrence by several decades of cumulative climatological study (e.g., H66; Novlan and Gray 1974; Gentry 1983). However, in this event, the Florida tornadoes occurred entirely near the southern edge of that sector. One factor for enhanced vertical shear, instability and enhanced tornado potential east of Earl's center, as opposed to elsewhere in the northeastern quadrant, was a pronounced frontal zone extending across central Florida. Tornadoes were most common within 100 km north of this front, implying that the results of Markowski et al. (1998a) regarding middle latitude mesoscale boundaries and tornadogenesis may apply to TCs interacting with fronts as well.

At the time of the Tampa and Tallahassee soundings, Charleston SC (not shown) had a relatively high CAPE of 3356 Jkg-1 but low SRH of 89 m2s-2. No tornadoes were reported in that area until about 12 hours later, when 5 tornadoes occurred as the center of Earl approached, and its associated enhanced wind fields and vertical shear structures moved overhead. Fig. 4 shows a nearly saturated Charleston sounding at 12 UTC, 3 September, with a reduced CAPE of 76 Jkg-1 but SRH increased greatly to 426 m2s-2.

Figure 4. As in Fig. 3, except for Charleston, SC, at 12 UTC, 3 September 1998.

Mesocyclones in Georgia and South Carolina associated with the exit phase of TC Earl (Fig. 5) occurred during a 12-hour period, and moved generally northward within the peripheral bands of the TC. The reflectivity structure of mesocyclone E (from Fig. 5) is presented in Fig. 6, sampling the persistent and enhanced reflectivity commonly associated with TC supercells (S97). The relatively intense reflectivity on the storm's rear (southern) flank, and cyclonic curvature therein, have been long identified as characteristic features among nontropical supercells as well (Lemon and Doswell 1979). A cross section of radial storm-relative velocity through this supercell (Fig. 7) illustrates the vertical structure of the associated tornadic mesocyclone, with the most intense velocities and shear evident around 4 km AGL. The convergence axis of the mesocyclone is tilted rightward (eastward) approximately 1 km per km in height, or 45 o from the vertical, through the lowest 5 km AGL. This shear intensified downward through tornadogenesis.

Figure 5. Paths of tornadic mesocyclones associated with the exit phase of TC Earl, 3-4 September 1998.

Figure 6. Base reflectivity at 0.5o elevation from Charleston SC (KCLX) WSR-88D, 1244 UTC, 03 September 1998. Tornadic supercell E (Fig. 5) is the area of enhanced reflectivity at upper middle.

Figure 7. Vertical cross section of storm-relative velocity from the Charleston, SC WSR-88D, same time as Fig. 6. The signature associated with the supercell in Fig. 6 is annotated here using the "MESO" label.


This study will be expanded to include more intensive examination of surface and upper air data for Earl and several other TC events of the late 1990s, in order to determine the extent and magnitude of various parameters associated with TC supercells and tornadogenesis. Such parameters include dry air entrainment into the cyclones, vertical shear in the lowest 1 km, and various derived kinematic and bulk parameters. The possible role of middle tropospheric drying will be examined in light of the empirical proposal of H66, and of the findings of Gilmore and Wicker (1998) related to middle latitude supercells. This study will include more thorough analyses for each event of observed proximal baroclinic features as possible contributors to the environmental shear and instability profiles favorable for TC tornadoes.


We thank the computer support staffs of SPC Scientific Services Branch and SLU Dept. of Earth and Atmospheric Sciences for data and imagery, and John Kobar of NCDC for supplying radar data. We also appreciate the logistical support and reviews of Dave Imy, Joe Schaefer and the rest of the SPC administrative staff. Special thanks goes to UCAR for their generous support of our COMET cooperative partners program studying tropical cyclone tornadoes.


Davies-Jones, R.P., D.W. Burgess and M. Foster, 1990: Test of helicity as a tornado forecast parameter. Preprints, 16th Conf. on Severe Local Storms, Kananaskis Park, AB, Canada, Amer. Meteor. Soc., 588-592.

Doswell, C.A. III, and E. N. Rasmussen, 1994: The effect of neglecting the virtual temperature correction on CAPE calculations. Wea. Forecasting, 9, 625-629.

Lemon, L. R., and C. A. Doswell III, 1979: Severe thunderstorm evolution and mesocyclone structure as related to tornadogenesis. Mon. Wea. Rev., 107, 1184-1197.

Edwards, R., 1998a: Storm Prediction Center forecast support for landfalling tropical cyclones. Preprints, 23rd Conf. On Hurricanes and Tropical Meteorology, Dallas, Amer. Meteor. Soc., 53-56.

_____, 1998b: Tornado production by exiting tropical cyclones. Preprints, 23rd Conf. On Hurricanes and Tropical Meteorology, Dallas, Amer. Meteor. Soc., 485-488.

Gentry, R.C., 1983: Genesis of tornadoes associated with hurricanes. Mon Wea. Rev., 111, 1793-1805.

Gilmore, M. S., and L. J. Wicker, 1998: The influence of midtropospheric dryness on supercell morphology and evolution. Mon. Wea. Rev., 126, 943-958.

Hagemeyer, B.C., 1997: Peninsular Florida tornado outbreaks. Wea. Forecasting, 12, 399-427.

Hill, E.L., W. Malkin and W.A. Schulz, Jr., 1966: Tornadoes associated with cyclones of tropical origin -- practical features. J. Appl. Meteor., 5, 745-763.

Markowski, P. M., E. N. Rasmussen, and J. M Straka, 1998a: The occurrence of tornadoes in supercells interacting with boundaries during VORTEX-95. Wea. Forecasting, 13, 852-859.

_____, J. M. Straka, E. N. Rasmussen, and D. O. Blanchard, 1998b: Variability of storm-relative helicity during VORTEX. Mon. Wea. Rev., 126, 2959-2971.

McCaul, E.W., Jr., 1991: Buoyancy and shear characteristics of hurricane-tornado environments. Mon Wea. Rev., 119, 1954-1978.

Novlan, D.J., and W.M. Gray, 1974: Hurricane-spawned tornadoes. Mon Wea. Rev., 102, 476-488.

NSSL, 2000: WATADS (WSR-88D Algorithm Testing and Display System) 2000: Reference guide for version 10.2 [Available from Storm Scale Applications Division, National Severe Storms Laboratory, 1313 Halley Circle, Norman, OK 73069].

Rao, G.V., R. Edwards and J. Scheck, 2000: Case studies of tornadoes associated with tropical cyclones based on conventional and WSR-88D data. Preprints, 24th Conf. On Hurricanes and Tropical Meteorology, Ft. Lauderdale FL, Amer. Meteor. Soc.

Spratt, S.M., D.W. Sharp, P. Welsh, A.C. Sandrik, F. Alsheimer and C. Paxton, 1997: A WSR-88D assessment of tropical cyclone outer rainband tornadoes. Wea. Forecasting, 12, 479-501.

Vescio, M.D., S.J. Weiss, and F.P. Ostby, 1996: Tornadoes associated with Tropical Storm Beryl. Preprints, 21st Conf. Hurricanes and Tropical Meteor., Amer. Meteor. Soc., Miami, 469-471.

Weiss S.J., 1987: Some climatological aspects of forecasting tornadoes associated with tropical cyclones. Preprints, 17th Conf. Hurricanes and Tropical Meteor., Amer. Meteor. Soc., Miami, 160-163.