• Blanchard, D. O., 1998: Assessing the vertical distribution of convective available potential energy. Wea. Forecasting,13, 870–877.

    • Crossref
    • Export Citation
  • Brady, R. H., and E. J. Szoke, 1989: Case study of nonmesocyclone tornado development in northeast Colorado: Similarities to waterspout formation. Mon. Wea. Rev.,117, 843–856.

    • Crossref
    • Export Citation
  • Brown, R. A., L. R. Lemmon, and D. W. Burgess, 1978: Tornado detection by pulsed Doppler radar. Mon. Wea. Rev.,106, 29–38.

    • Crossref
    • Export Citation
  • Crum, T. D., and R. L. Alberty, 1993: The WSR-88D and the WSR-88D Operational Support Facility. Bull. Amer. Meteor. Soc.,74, 1669–1687.

    • Crossref
    • Export Citation
  • Davies-Jones, R., 1984: Streamwise vorticity: The origin of updraft rotation in supercell storms. J. Atmos. Sci.,41, 2991–3006.

    • Crossref
    • Export Citation
  • Donaldson, R. J., Jr., 1970: Vortex signature recognition by a Doppler radar. J. Appl. Meteor.,9, 661–670.

    • Crossref
    • Export Citation
  • Djuric, D., 1994: Weather Analysis. Prentice-Hall, 304 pp.

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

    • Crossref
    • Export Citation
  • Fujita, T. T., 1981: Tornadoes and downbursts in the context of generalized planetary scales. J. Atmos. Sci.,38, 1511–1534.

    • Crossref
    • Export Citation
  • Hodanish, S., D. Sharp, W. Collins, and C. Paxton, 1997: A 10-yr monthly lightning climatology of Florida: 1986–95. Wea. Forecasting,12, 439–448.

  • Johnson, J. T., P. L. MacKeen, A. Witt, E. D. Mitchell, G. J. Stumpf, M. D. Eilts, and K. W. Thomas, 1998: The Storm Cell Identification and Tracking (SCIT) algorithm: An enhanced WSR-88D algorithm. Wea. Forecasting,13, 263–276.

    • Crossref
    • Export Citation
  • Klemp, J. B., 1987: Dynamics of tornadic thunderstorms. Annu. Rev. Fluid Mech.,19, 369–402.

    • Crossref
    • Export Citation
  • Lee, B. D., and R. B. Wilhelmson, 1996: The numerical simulation of non-supercell tornadogenesis. Preprints, 18th Conf. on Severe Local Storms, San Francisco, CA, Amer. Meteor. Soc., 200–204.

  • Leslie, L. M., 1971: The development of concentrated vortices: A numerical study. J. Fluid Mech.,48, 1–21.

    • Crossref
    • Export Citation
  • Markowski, P. M., E. N. Rasmussen, J. M. Straka, and D. C. Dowell, 1998: Observations of low-level baroclinicity generated by anvil shadows. Mon Wea. Rev.,126, 2942–2958.

    • Crossref
    • Export Citation
  • Mitchell, E. D., S. V. Vasiloff, G. J. Stumpf, A. Witt, M. D. Eilts, J. T. Johnson, and K. W. Thomas, 1998: The National Severe Storms Laboratory Tornado Detection Algorithm. Wea. Forecasting,13, 352–366.

    • Crossref
    • Export Citation
  • Mueller, C. K., and R. E. Carbone, 1987: Dynamics of a thunderstorm outflow. J. Atmos. Sci.,44, 1879–1898.

    • Crossref
    • Export Citation
  • NOAA, 1995: Storm Data. Vol. 37. National Climatic Data Center. [Available from National Climatic Data Center, Federal Building, 151 Patton Ave., Asheville, NC 28801.].

  • ——, 1997: Storm Data. Vol. 39. National Climatic Data Center. [Available from National Climatic Data Center, Federal Building, 151 Patton Ave., Asheville, NC 28801.].

  • NSSL, 1997: WSR-88D Algorithm Testing and Display System Reference Guide, Version 9.0, 179 pp. [Available from Storm Scale Research Application Division, National Severe Storms Laboratory, 1313 Halley Circle, Norman, OK 73069.].

  • Paxton, C. H., R. D. Pelatti, and M. E. Powell, 1996: Radiosonde climatology of west-central Florida. NOAA Tech. Memo. NWS SR-181, 19 pp. [Available from NTIS, 5285 Port Royal Rd., Springfield, VA 22161.].

  • Priegnitz, D. L., 1995: IRAS: Software to display and analyze WSR-88D radar data. Preprints, 11th Int. Conf. on Interactive Information and Processing Systems for Meteorology, Oceanography, and Hydrology, Dallas, TX, Amer. Meteor. Soc., 197–199.

  • Purdom, J. F. W., and K. Marcus, 1982: Thunderstorm trigger mechanisms over the southeast United States, Preprints, 12th Conf. on Severe Local Storms, San Antonio, TX, Amer. Meteor. Soc., 487–488.

  • Roberts, R. D., and J. W. Wilson, 1995: The genesis of three nonsupercell tornadoes observed with dual Doppler radar. Mon. Wea. Rev.,123, 3408–3436.

    • Crossref
    • Export Citation
  • Rogers, E., T. L. Black, D. G. Deaven, and G. J. DiMego, 1996: Changes to the operational “early” Eta analysis/forecast system at the National Centers for Environmental Prediction. Wea. Forecasting,11, 391–413.

    • Crossref
    • Export Citation
  • Rotunno, R., 1981: On the evolution of thunderstorm rotation. Mon. Wea. Rev.,109, 577–586.

    • Crossref
    • Export Citation
  • Saucier, W. J., 1955: Principles of Meteorological Analysis. University of Chicago Press, 438 pp.

  • Stumpf, G. J., A. Witt, E. D. Mitchell, P. L. Spencer, J. T. Johnson, M. D. Eilts, K. W. Thomas, and D. W. Burgess, 1998: The National Severe Storms Laboratory mesocyclone detection algorithm for the WSR-88D. Wea. Forecasting,13, 304–326.

    • Crossref
    • Export Citation
  • Trapp, R. J., and R. Davies-Jones, 1997: Tornadogenesis with and without a dynamic pipe effect. J. Atmos. Sci.,54, 113–133.

    • Crossref
    • Export Citation
  • ——, E. D. Mitchell, G. A. Tipton, D. W. Effertz, A. I. Watson, D. L. Andra, and M. A. Magsig, 1998: Descending and non-descending tornadic vortex signatures. Preprints, 19th Conf. on Severe Local Storms, Minneapolis, MN, Amer. Meteor. Soc., 128–131.

  • Wakimoto, R. M., and J. W. Wilson, 1989: Non-supercell tornadoes. Mon. Wea. Rev.,117, 1113–1140.

    • Crossref
    • Export Citation
  • ——, and J. K. Lew, 1993: Observations of a Florida waterspout during CaPE. Wea. Forecasting,8, 412–423.

  • Weisman, M. L., and J. B. Klemp, 1982: The dependence of numerically simulated convective storms on vertical wind shear and buoyancy. Mon. Wea. Rev.,110, 504–520.

    • Crossref
    • Export Citation
  • ——, and —— 1984: The structure and classification of numerically simulated convective storms in directionally varying wind shears. Mon. Wea. Rev.,112, 2479–2498.

    • Crossref
    • Export Citation
  • ——, and —— 1986: Characteristic of isolated convective storms. Mesoscale Meteorology and Forecasting, P. S. Ray, Ed., Amer. Meteor. Soc., 331–358.

    • Crossref
    • Export Citation
  • Wilczak, J. M., T. W. Christian, D. E. Wolfe, R. J. Zamora, and B. Stankov, 1992: Observations of a Colorado tornado. Part I: Mesoscale environment and tornadogenesis. Mon Wea. Rev.,120, 497–520.

    • Crossref
    • Export Citation
  • Ziegler, C. L., P. S. Ray, and D. R. MacGorman, 1986: Relations of kinematics, microphysics and electrification in an isolated mountain thunderstorm. J. Atmos. Sci.,43, 2098–2114.

    • Crossref
    • Export Citation
  • View in gallery

    Photograph of the large waterspout over Tampa Bay viewed from the west.

  • View in gallery

    WSR-88D locations of the precursor circulation (1744–1759 UTC), tornado (1804–1809 UTC), and waterspout (1814 UTC) relative to the Tampa Bay area and the foregoing radar.

  • View in gallery

    Eta Model initialization at 250 (upper-left), 500 (upper-right), 850 (lower-left), and 1000 mb (lower-right) for 12 Jul 1995 at 1200 UTC. At 250 mb, solid lines depict geopotential heights (in m); dashed lines indicate isotachs (in m s−1). At 500 mb, solid lines depict geopotential heights (in m); dashed lines reveal absolute vorticity (in ×105 s−1). At 850 mb, solid lines depict geopotential heights (in m);dashed lines reveal equivalent potential temperature (K). Wind barbs in each image are in m s−1, with full (half) barb equal to 10 (5) m s−1. The thick dashed lines in each image designate the wind shift associated with the trough.

  • View in gallery

    (a) Ruskin, FL, skew T–logp for 1200 UTC 12 Jul 1995. Ordinate (pressure) is in mb. Abscissa (temp) is in °C. Thick solid lines depict temperature (T) and dewpoint (Td). Thin sold line between T and Td is the wet-bulb temperature (Tw). Path of parcel is the most unstable path and is also based on virtual temperature. The half (full) wind barb represents 2.5 (5) m s−1. (b) Ruskin mean hodograph averaged over the period 1700–1800 UTC (12 Jul 1995). Data source is the Ruskin WSR-88D VWP output. The values along the hodograph represent the height (km) at each point. The M denotes the position of the 0–6-km pressure-weighted mean wind for this hodograph. The S is the position of the mean storm motion of the TC, averaged over its life.

  • View in gallery

    (a) Visible satellite image (which includes central FL) for 12 Jul 1995 at 1545 UTC. Letter A denotes the convective cell responsible for generating the OFB. The arrow labeled B points to the convection over the Tampa area that produced the QSB. Letter C depicts the drier air north of the surface trough.

  • View in gallery

    (Continued) (b) Visible satellite image (which includes central FL) at 1732 UTC 12 Jul 1995. Note the identifications of cells N and S, the QSB, and the OFB. The green letter T, located on the QSB, depicts the location of the tornado at 1804 UTC, and the blue TBW denotes the location of the TBW WSR-88D.

  • View in gallery

    Composite reflectivity (left column), 0.5° base reflectivity (center column), and 0.5° base radial velocity (right column) for the 1739 (top row), 1759 (center row), and 1804 UTC (bottom row) volume scans. Top legend denotes equivalent reflectivity (units of dBZ); truncated below 24 dBZ), and bottom legend depicts base radial velocity (units are m s−1). Equivalent lines in (a)–(c) denote the center of the cyclonic horizontal shear region depicted in (c). Equivalent lines in (d)–(i) refer to the corresponding cross sections in Fig. 8.

  • View in gallery

    (a) Base reflectivity and (b) base radial velocity at the 8.7° elevation angle for the 1809 UTC volume scan. (e) Base reflectivity and (f) base radial velocity at the 12° elevation angle for the 1809 UTC volume scan. Equivalent lines, which slice across the midlevel vortex pair in (a), (b) [(e), (f)], correspond to cross sections (c), (d) [(g), (h)]. For each cross section, the ordinate (abscissa) is height (horizontal distance) in km. Sets of numbers at the base of each cross section refer to the azimuth (°) and range (km), relative to the TBW WSR-88D, at the cross section edges. The distance between tick marks at the base of each cross section is 5 km. The top legend denotes reflectivity (dBZ), and bottom legend depicts base radial velocity (m s−1). The arrow in (b) points to the tornadic circulation.

  • View in gallery

    (Continued).

  • View in gallery

    RHI cross sections of base reflectivity (left column) and base radial velocity (right column) for 1759 (top row) and 1804 UTC (bottom row) volume scans. Sets of numbers at the base of each cross section refer to the azimuth (°) and range (km) relative to the TBW WSR-88D at the edge of each cross section. The distance between tick marks at the base of each cross section is 5 km. Height is in units of km. The arrows in (b) and (d) point to the position of the near-surface tornadic circulation. The top legend denotes base reflectivity (dBZ), and the bottom legend depicts base radial velocity (m s−1).

  • View in gallery

    Time (abscissa) vs height (ordinate) of base reflectivity (solid lines; units of dBZ) at the center of the developing tornadic circulation, and the azimuthal shear (dashed lines; units of ×10−3 5−1) of the circulation. The shaded region is base reflectivity ≥50 dBZ. Thick solid line at the base of the figure depicts the approximate tornado/waterspout duration. Time is in UTC. Height is in m.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 31 31 5
PDF Downloads 20 20 2

The 12 July 1995 Pinellas County, Florida, Tornado/Waterspout

View More View Less
  • 1 NOAA/National Weather Service, Corpus Christi, Texas
  • 2 NOAA/National Weather Service, Tampa Bay, Florida
  • 3 NOAA/Forecast Systems Laboratory, Boulder, Colorado
© Get Permissions
Full access

Abstract

On 12 July 1995, a tornado developed over south St. Petersburg, Florida, producing F1 damage and injuring one person before moving offshore. The tornado/waterspout was within 25 km of the Ruskin Florida WSR-88D, which provided detailed radar data. The preconvective environment was characterized by large CAPE and weak to moderate vertical wind shear, due in part to a weak upper-level cold core trough. The tornado parent cell developed rapidly in response to surface mesoscale boundary interactions. This cell was relatively short lived and nonsteady and, thus, classified as multicellular. Available data suggest that tornadogenesis occurred due to vertical stretching of preexisting vertical vorticity associated with one of the foregoing boundaries. Evidence suggests that the stretching was due to both storm updraft and convergence associated with storm downdraft. The parent cell contained a midlevel mesocyclone and mesoanticyclone pair, consistent with the proximity hodograph. This vortex pair and the tornadic circulation were separate and it is unclear what role the vortex pair contributed to tornadogenesis. This case is important since it demonstrates that a nonsupercell tornado can be anticipated before a single-Doppler radar tornado vortex signature (TVS) appears, using current nonsupercell tornadogenesis theories. Such anticipation is essential to operational forecasters in the National Weather Service, especially for cases when tornadoes are either undetectable by radar or when a radar-detected TVS does not provide sufficient lead time.

Corresponding author address: Waylon G. Collins, NOAA/NWS, 300 Pinson Drive, International Airport, Corpus Christi, TX 78406-1803.

Email: Waylon.Collins@noaa.gov

Abstract

On 12 July 1995, a tornado developed over south St. Petersburg, Florida, producing F1 damage and injuring one person before moving offshore. The tornado/waterspout was within 25 km of the Ruskin Florida WSR-88D, which provided detailed radar data. The preconvective environment was characterized by large CAPE and weak to moderate vertical wind shear, due in part to a weak upper-level cold core trough. The tornado parent cell developed rapidly in response to surface mesoscale boundary interactions. This cell was relatively short lived and nonsteady and, thus, classified as multicellular. Available data suggest that tornadogenesis occurred due to vertical stretching of preexisting vertical vorticity associated with one of the foregoing boundaries. Evidence suggests that the stretching was due to both storm updraft and convergence associated with storm downdraft. The parent cell contained a midlevel mesocyclone and mesoanticyclone pair, consistent with the proximity hodograph. This vortex pair and the tornadic circulation were separate and it is unclear what role the vortex pair contributed to tornadogenesis. This case is important since it demonstrates that a nonsupercell tornado can be anticipated before a single-Doppler radar tornado vortex signature (TVS) appears, using current nonsupercell tornadogenesis theories. Such anticipation is essential to operational forecasters in the National Weather Service, especially for cases when tornadoes are either undetectable by radar or when a radar-detected TVS does not provide sufficient lead time.

Corresponding author address: Waylon G. Collins, NOAA/NWS, 300 Pinson Drive, International Airport, Corpus Christi, TX 78406-1803.

Email: Waylon.Collins@noaa.gov

1. Introduction

Florida has the greatest mean annual tornado density of any state in the United States (NOAA 1997). In addition, half of all Florida tornadoes during the period 1950 through 1995 occurred from May through August (D. Imy 1998, personal communication). Further, Florida typically experiences persistent convective activity during the period June through August, owing to the interaction of mesoscale convergence, abundant moisture, and static/potential instability that occur on a nearly diurnal basis (e.g., Hodanish et al. 1997). Thus, a nearly daily challenge to National Weather Service forecasters in Florida during much of the summer is to determine which convective cell, typically among a myriad of cells, is likely to become tornadic.

Anticipating tornadogenesis requires knowledge of the various tornadogenesis theories. Tornadogenesis mechanisms disseminated in the literature are typically expressed in the context of storm type. Tornadoes associated with supercells can develop via descent of the radar-detected tornado vortex signature (TVS; Brown et al. 1978), within the midlevel mesocyclone, to the surface (Leslie 1971; Trapp and Davies-Jones 1997). In addition, supercell tornadoes can also originate within the subcloud layer as near-surface horizontal streamwise vorticity, associated with the forward flank downdraft, is tilted and stretched (e.g., Klemp 1987). Nonsupercell tornadoes can originate in the subcloud layer as near-surface preexisting vertical vorticity is stretched (Wakimoto and Wilson 1989, hereafter WW89; Brady and Szoke 1989, hereafter BS89) or when local near-surface horizontal streamwise vorticity is tilted and stretched (Wilzak et al. 1992). Further, Lee and Wilhelmson (1996, hereafter LW96) and Roberts and Wilson (1995, hereafter RW95) have demonstrated that convergence associated with outflow boundaries can help determine the timing and/or intensity of nonsupercell tornadoes. These foregoing mechanisms are not exhaustive.

On 12 July 1995, a tornado developed over south St. Petersburg, Florida (in extreme southern Pinellas County), between 1759 and 1804 UTC, moved southeast, and produced an estimated $200 000 in damage along a ∼2.8 km path. Numerous homes and businesses were damaged and one person was injured. Damage was classified in Storm Data (NOAA 1995) as F1 (Fujita 1981). The tornado partially damaged an apartment complex before moving offshore over Tampa Bay around 1810 UTC, becoming a distinctive waterspout with the condensation funnel in contact with the water surface (Fig. 1). Approximate locations of the tornado and waterspout, and the Ruskin (TBW) Weather Surveillance Radar 1988-Doppler (WSR-88D) radar (Crum and Alberty 1993) are shown in Fig. 2. The primary purpose of this paper is to determine, based on the data available, the tornadogenesis mechanism associated with this tornado/waterspout.

The parent storm developed in an environment composed of low to moderate vertical wind shear, high convective available potential energy (CAPE), and low convective inhibition (CIN). It was triggered by boundary interactions, then traveled along a quasi-stationary mesoscale boundary during its life. A midlevel mesocyclone and mesoanticyclone couplet developed within the parent storm, consistent with the proximity hodograph. This vortex pair was separate from the tornadic circulation and did not appear to contribute to tornadogenesis. Storm structure also included a weak echo region (WER), downshear V notch, and strong reflectivity gradient upshear. Although these radar-derived radial velocity and reflectivity features are present in supercells, the non–steady state nature of this storm suggests a multicellular classification. In particular, ∼40 min after storm initiation, the downdraft moved ahead of the storm, cutting off the supply of positively buoyant air to the updraft, resulting in storm dissipation.

Available data suggest the following tornadogenesis mechanism: Preexisting near-surface vertical vorticity existed along a quasi-stationary boundary. Ambient updraft, and convergence associated with a downdraft, stretched this near-surface vertical vorticity to tornadic strength.

The tornado/waterspout was within 25 km of the TBW WSR-88D, thus providing detailed radar analysis. The radar perusal software programs WSR-88D Algorithm Testing and Display System (WATADS, NSSL 1997); and Interactive Radar Analysis Software (IRAS, Priegnitz 1995) were used in radar analysis. The WATADS software includes the National Severe Storms Laboratory (NSSL) tornado detection (TDA; Mitchell et al. 1998), mesocyclone detection (MDA; Stumpf et al. 1998), and storm cell identification and tracking (SCIT; Johnson et al. 1998) algorithms. These algorithms, along with manual radar analysis, were used to analyze the storm in this case. The adjustable parameters for the TDA applied to the data for this study are included in appendix A. In addition, available satellite, thermodynamic, surface, and model intialization data were used.

2. Preconvective environmental conditions

a. Synoptic features and local thermodynamics

The synoptic-scale features were examined using the 12 July 1995 1200 UTC Eta Model (Rogers et al 1995) initialization for the 250-, 500-, 850-, and 1000-mb pressure levels (Fig. 3). Florida is typically under the influence of the subtropical ridge during July (e.g., Saucier 1955). However, a weak upper-level cold core trough developed/moved toward the south and west around the eastern and southeastern side of an upper ridge, into Florida and the eastern Gulf of Mexico. Figure 3 depicts this upper trough oriented approximately N–S along the southeastern U.S. coast, across northern Florida, and into the eastern Gulf of Mexico. Also note the 30 m s−1 jet streak at 250 mb entering Georgia from the north. At the lower levels, a trough was present and oriented approximately E–W across the north-central Florida peninsula. The 1200 UTC TBW Skew T–log p thermodynamic sounding (Fig. 4a), located at Ruskin, ∼25 km east of the location of subsequent tornadogenesis, was analyzed taking virtual temperature into account (Doswell and Rasmussen 1994). The sounding revealed an extremely moist environment with static and potential instability. The unmodified convective available potential energy (CAPE) measurement was 2674 J kg−1, based on the most unstable parcel. The normalized CAPE, or NCAPE (Blanchard 1998) was ∼3.70 J kg−1 mb−1. Both the TBW CAPE and NCAPE were greater than the TBW median values for the warm season (Blanchard 1998), likely due to colder temperatures aloft associated with the foregoing upper trough; the 500-, 400-, and 300-mb temperatures over Ruskin were 1.1°, 0.6°, and 0.7°C below the July normals, respectively (Paxton et al. 1996). The CIN value was −1.0 J kg−1, which implies early release of CAPE and subsequent convection before time of maximum heating (e.g., Djuric 1994). Convection did develop early over the extreme eastern Gulf of Mexico and over the Florida peninsula. Figure 4b reveals a proximity “mean” hodograph generated from the Ruskin WSR-88D velocity–azimuth display wind profile (VWP). This hodograph was constructed by calculating the resultant winds at each level for the 1-h period preceeding tornado commencement. Notice that the hodograph is quasi-straight in the lowest 3.5 km. Unidirectional shear in the lower levels is consistent with counterrotating midlevel circulations when such shear is tilted and stretched by storm updraft (e.g., Davies-Jones 1984, hereafter DJ84). The tornado parent cell did contain midlevel counterrotating circulations. The 0–6-km wind shear vector points toward ∼260°, with a magnitude of ∼2.5 × 10−3 s−1. Combining this hodograph to the 1200 UTC unmodified CAPE results in a bulk Richardson number (BRN) of 184. This BRN value suggests unsteady multicellular-type convection (e.g., Weisman and Klemp 1982; Weisman and Klemp 1984, hereafter WK). The tornado parent cell downdraft moved ahead of the storm, resulting in an unsteady system. Further, the mean hodograph reveals a 0–6-km pressure-weighted mean wind vector of 60° at 2.8 m s−1. The 0–6-km mean wind and the low-level vertical wind shear vector suggest that any multicellular convection would tend to propagate toward the west (Weisman and Klemp 1982). However, the storm moved toward the southeast (see Fig. 4b). Direction ofmovement of multicellular convection can be heavily influenced by surface boundaries (Weisman and Klemp 1986) and in this case the tornado parent cell moved along a preexisting mesoscale boundary (discussed in the next section).

b. Mesoscale boundaries

Two intersecting meso-β-scale (Fujita 1981) surface boundaries responsible for initiation of the tornado parent cell (TC) were clearly discernible on Geostationary Operational Environmental Satellite-8 visible satellite images prior to tornadogenesis.

Convection developed over Tampa, Florida, then produced an outflow boundary that became quasi-stationary and extended approximately SE–NW across central and southern Pinellas County, then E–W into Tampa Bay (Figs. 5a,b). This quasi-stationary boundary (QSB), and the associated horizontal shear, were also apparent from surface land and marine observations (not shown), and WSR-88D base velocity data (Fig 6c). Another outflow boundary (OFB), oriented SW–NE, developed from convection offshore, moved eastward, then onshore over central and southern Pinellas County after 1650 UTC (Fig. 5b). The OFB was identifiable on both satellite and radar data at least until the time of tornadogenesis.

3. Convection initiation, storm evolution, and structure

Lower-level convergence, associated with the synotic-scale trough, combined with instability and near-zero CIN to generate a convective cell over Tampa (identified by arrow labeled B in Fig. 5a). An additional trigger may have been provided by net radiational variations (e.g., Markowski et al. 1998) since Tampa was located near the edge of a storm anvil from earlier convection. As previously mentioned, convection over the Tampa area generated a QSB and convection offshore generated an OFB. Radar and satellite data revealed that the OFB moved onshore and triggered convection over Pinellas County and that the TC was triggered in response to the intersection of boundaries QSB and OFB (Fig. 5b). Such boundary interactions have been associated with strong convection (e.g., Purdom and Marcus 1982).

Two convective cells developed in close proximity to each other over extreme southern Pinellas County in response to the OFB and QSB. They are labeled N and S in the 1732 UTC satellite image (Fig. 5b) and in the 1739 UTC radar scan (Fig. 6a). Both cells began to merge in the upper levels at the 1749 UTC radar scan (not shown), owing to differential storm motion. In addition, counterrotating mesoscale circulations of similar azimuthal shear developed in the composite cell, with the cyclonic (anticyclonic) member to the north (south). The existence and orientation of the vortex pair can be explained by the tilting and stretching of 0–3-km environmental horizontal vorticity. The mean hodograph (Fig. 4b) is quasi-straight in the lowest 3.5 km with the 0–3.5-km vertical wind shear vector pointing toward the west. According to DJ84, an updraft would tilt the associated horizontal vorticity to create a vortex pair with the cyclonic (anticyclonic) member to the north (south), which is consistent with the observed radar features. The cells merged at all levels by the 1759 UTC radar scan (Figs. 6d,e). Also note that during this scan, WERs are implied on both flanks. An updraft is associated with the left flank (relative to the 0–6-km wind shear vector) during this scan (discussed in the next section). During the next volume scan, a well-defined hook, associated with the tornado, developed on the left flank of the composite cell (TC) (Fig. 6h). Also note the V-notch downshear in Fig. 6h. After tornadogenesis, the downdraft moved ahead of the TC resulting in a decrease in intensity as warm inflow into the storm updraft was cutoff. By the 1909 UTC radar volume scan (not shown), the TC was nearly nonexistent. Using a maximum reflectivity of ≥30 dBZ to define the existence of a convective storm, the time from the development of cells N and S to the demise of the TC was ∼10 min (1729–1909 UTC).

The midlevel vortex pair persisted until the 1835 UTC radar scan. (Figures 7a, b, d, f, h reveal the vortex pair in the 1809 UTC radar scan.) Both the cyclonic and anticyclonic members possessed azimuthal shears ranging generally from 5 × 10−3 s−1 to 1.0 × 10−2 s−1, with a maximum shear of 1.4 × 10−2 s−1 for each (not shown). The vortex depths and diameters ranged generally from 5 to 8 km and 3 to 5 km, respectively (not shown). These shear, depth, and time continuity values satisfy the mesocyclone recognition criteria established by Donaldson (1970).

The relatively short-lived (∼10 min) and nonsteady nature of this convective system suggest that it was multicellular. This multicellular classification is consistent with the high BRN in this case. The BRN is a nondimensional parameter that measures the relative contributions of buoyancy and vertical wind shear to storm structure. The buoyancy governs the strength of subsequent storm outflow, whereas the vertical wind shear governs the strength of the near-surface inflow and the ability of the storm to generate midlevel updraft rotation. Low BRN values (10–50) indicate that the near-surface inflow was strong enough to keep storm outflow from moving ahead of the storm and that the ambient 0–6-km vertical wind shear was strong enough (relative to buoyancy) to generate vertical pressure gradients sufficient to maintain a quasi-steady rotating updraft characteristic of supercells (WK). High BRN values (>50) indicate that the storm inflow was not strong enough to keep the outflow from moving ahead of the storm and that vertical wind shear was not strong enough to maintain a quasi-steady updraft, resulting in nonsteady multicells (WK).

4. Tornadogenesis

It is estimated that the tornado began between 1759 and 1804 UTC, based on the spatial correlation of the near-surface portion of the TVS to the location of the first tornado report and then determining the WSR-88D time this spatial correlation occurred. The tornado moved offshore by 1810 UTC, based on both the subsequent position of the foregoing TVS at the 1809 UTC volume scan and reports received from the United States Coast Guard. The waterspout diminished between 1814 and 1819 UTC, based on analysis of radial velocity.

Available evidence suggests that low-level vertical vorticity, associated with the QSB, was stretched by a convective updraft to help generate the tornado. Figure 6c denotes a near-surface meso-γ-scale horizontal shear region at the 1739 UTC scan. Comparing the location of this horizontal shear region, as depicted in the 1729 UTC and 1734 UTC 0.5° radial velocity plan position indicators (PPIs) (not shown), to the location of the QSB, based on the 1732 UTC satellite image (Fig. 5b), places this shear region within the central to western portions of the QSB. Analysis of radial velocity PPIs from the 1739 to 1804 UTC scans (some of which are depicted in Fig. 6) suggest that the tornado developed within this shear region. Figures 6a–c depict the foregoing N and S cores aloft (suggestive of updrafts beneath), almost directly above this shear region. Thus, stretching (via updraft) of vortices within this shear region may have contributed to torndogenesis. Further, Fig. 8 depicts the reflectivity and corresponding radial velocity range–height indicators (RHIs), at 1759 and 1804 UTC, which slice through the near-surface circulation as it developed. Note that just before the tornado formed (Figs. 8a,b), maximum reflectivities were almost directly above the circulation, which implies updraft and subsequent stretching of this low-level vorticity. The updraft is further implied by the narrow funnel-shaped region of outbound velocities (Figs. 8b,d) above the near-surface circulation. This radial velocity signature is physically consistent with an updraft (P. Ray 1998–99, personal communication; Ziegler et al. 1986). Note the apparent lower-level convergence and upper-level divergence. Also, note the increase in outbound radial velocity with height, which would be expected from an increasing radar beam tilt as the radial component detects an increasing vertical velocity component. Additional evidence that the circulation likely developed from the boundary layer and was stretched by an updraft is provided by following the reflectivity at the center of the circulation in time–height space (Fig. 9). Note that below 2 km, the shear associated with the circulation increased to 2 × 10−2 s−1 by 1752 UTC when the reflectivity aloft became 25–30 dBZ greater than the low-level value. By 1758 UTC the 2 × 10−2 s−1 isoline increased to a depth of 4 km. Thus, the higher reflectivities aloft suggest updraft that likely vertically stretched the lower-level circulation resulting in an increase in vortex strength and vertical depth. Additional evidence that the tornadic circulation in this case was stretched upward from the surface, as opposed to downward from aloft, is provided by the behavior of the TVS. The behavior of the TVS in this case would be classified as the nondescending type as described by Trapp et al. (1998). Hence, available data suggest that an updraft stretched near-surface vertical vorticity to near-tornadic strength, similar to that described by WW89 and BS89.

Additional stretching for tornadogenesis in this case was likely provided by downdraft convergence. Based on analysis of radial velocity/corresponding reflectivity RHIs for the 1739–1754 UTC volume scans (not shown), the inbound velocities below 5 km in Figs. 8b and 8d depict the leading edge of the outflow generated by TC. Further, the development of the outflow was coincident with the descent of high reflectivity values (≥50 dBZ) to the surface. Note that Fig. 8b depicts this downdraft while it approaches the near-surface circulation during the 1759 UTC scan. By the 1804 UTC scan (Figs. 8d and 8c), the near-surface portion (⩽1 km) of the tornadic circulation was embedded in both a region of inbound velocity and high reflectivity, which suggests that the downdraft was temporally and spatially correlated with tornado formation. This correlation is also apparent from Fig. 9. Note the high reflectivity values that suddenly appear below 1 km between 1800 and 1805 UTC. Further, Figs. 8b and 8d depict low-level radial convergence near the circulation. Both RW95, using both single- and dual-Doppler radar data, and LW96, using numerical model output, have demonstrated that stretching due to convergence associated with outflow boundaries can help determine the timing and/or intensity of nonsupercell tornadoes.

The authors conclude that both a convective updraft and convergence associated with a downdraft stretched preexisting low-level vertical vorticity. Since stretching acts on existing vorticity exponentially (Rotunno 1981), near-surface vertical vorticity was able to spin up rapidly to tornadic strength. Notice from Fig. 9 that below 1 km, the azimuthal shear increased by a factor of 3 from 1759 to 1803 UTC.

Since dual-Doppler analysis was not possible, the foregoing development sequence, although likely, is not certain. Other mechanisms are possible. For example, it is noteworthy to mention that additional near-surface vertical vorticity could have been generated from 1) possible preexisting vertical vorticity along the leading edge of the OFB and/or the TC outflow (e.g., Mueller and Carbone 1987), or 2) tilting and stretching of streamwise horizontal vorticity that may have existed along the OFB, the TC outflow, and/or the QSB (e.g., Wilzak et al. 1992).

Throughout the life span of the TC, the midlevel vortex pair and the tornadic circulation were separate entities. This vortex pair did not appear to contribute to tornadogenesis.

5. Discussion and conclusions

An F1 tornado developed over south St. Petersburg, Florida, on 12 July 1995, then moved offshore becoming a waterspout. The parent storm developed in an environment characterized by significant potential and static instability, low to moderate vertical wind shear, and high moisture content. The cold upper low contributed to the larger CAPE and NCAPE, and to the vertical wind shear. Intersecting boundaries provided the necessary lift for the release of CAPE and generation of the parent storm in this study. The tornado parent cell was nonsteady and relatively short lived and, thus, classified as multicellular. Available data in this case suggest that the source of vertical vorticity for the tornado was near-surface vertical vorticity associated with a quasi-stationary boundary. Updrafts from developing convection appeared to stretch the near-surface vorticity to near-tornadic strength; a process similar to the nonsupercell tornado genesis mechanism described by WW89 and BS89. In addition, the leading edge of an outflow boundary generated by the TC was spatially and temporally correlated with the tornado when the tornado developed. It is likely that this boundary provided additional stretching via convergence to strengthen the circulation to tornadic strength. This additional process is supported by RW95 and LW96, who found for nonsupercell tornadoes that the convergence on the leading edge of outflow boundaries can influence the timing and/or intensity of tornadoes. Data were not available to access whether other sources of vertical vorticity were important to tornadogenesis such as the tilting and stretching of horizontal streamwise vorticity. A midlevel mesocyclone and mesoanticyclone vortex pair developed within the TC, consistent with the proximity hodograph. The tornadic circulation was separate from this midlevel mesoscale vortex pair, and it is unclear what role is attributed to these midlevel circulations to tornadogenesis in this case.

This study is important to forecasters since it demonstrates that a nonsupercell tornado can be anticipated before a radar-defined TVS occurs, by applying the foregoing nonsupercell tornadogenesis theories. The region of near-surface horizontal shear, which the authors conclude to be the vertical vorticity source for this tornado/waterspout, was clearly visible on the WSR-88D 0.5° radial velocity output ∼30 min before tornadogenesis. When a convective updraft moved over the near-surface shear region, lower-level vertical vorticity was vertically stretched, then increased to tornadic strength when additional stretching was provided by a downdraft. Anticipation is necessary because not all tornadoes or waterspouts can be identified with current WSR-88D radars, even at close ranges, either from using radial velocity data alone to identify velocity couplets (e.g., Wakimoto and Lew 1993) or from TVS algorithms such as the NSSL TDA (Mitchell et al. 1998). In fact, applying the TDA to the data in this case, using the WSR-88D build 10 default parameter settings, would not have generated a TVS alarm. However, an elevated TVS (ETVS) alarm (see appendix B), would have appeared, but with zero lead time. The zero lead time is not suprising since the tornadogenesis mechanism suggested in this case can generate a tornado on a timescale on the order of one WSR-88D volume scan (in precipitation mode). Further, according to Trapp et al. (1998), a nondescending TVS within a convective line of cells typically has a lead time of only ∼5 min.

For tornadoes that originate in the boundary layer, anticipating such in an operational setting is not a simple process. This is because such tornadoes can develop owing to the stretching of preexisting vertical vorticity (e.g., WW89), or the tilting and stretching of horizontal streamwise vorticity (e.g., Wilzak et al. 1992). To further complicate matters, the stretching of preexisting vertical vorticity can occur from an updraft (e.g., WW89) and/or the leading edge of a downdraft (RW95, LW95). In addition, these processes can generate tornadoes in only a few minutes. In this case study, available data suggest that both updraft and the leading edge of a downdraft stretched preexisting vertical vorticity to generate the tornado on a timescale of only one radar volume scan. Nothwithstanding these difficulties, operational forecasters can improve their ability to anticipate many of these nonsupercell tornadoes/waterspouts by closely monitoring/anticipating situations where rapidly developing convective cells are collocated with surface boundaries. Surface boundaries can serve as a source for near-surface vertical vorticity, horizontal streamwise vorticity, and/or convergence. Strong updrafts could stretch low-level vertical vorticity, or tilt and stretch horizontal streamwise vorticity, to tornadic strength. Further, convergence at the leading edge of outflow boundaries can stretch preexisting vertical vorticity to tornadic strength. The lead author suggests that an ideal operational setting to anticipate the foregoing is to possess quasi–real time dual-Doppler lower-level microscale wind data (in order to identify regions of developing vertical or horizontal streamwise vorticity), then overlay such on radar reflectivity PPIs for several elevation angles for each scan. Thus, forecasters can quickly determine which convective cell will most likely generate a tornado via the foregoing mechanisms.

Finally, the lead author speculates that the study advanced by J. Davies (1998, personal communication) may be an additional tool to help anticipate nonsupercell tornadogenesis. Davies suggests that the level of free convection (LFC) may be useful in anticipating an environment condusive to these class of tornadoes. The LFC is the lowest level at which a parcel begins to accelerate upward. The lower the LFC, the lower the height at which parcels begin to accelerate upward and, thus, the greater the chance that lower-level vorticity can be tilted and/or stretched. Davies analyzed a case set of 242 tornadoes of F2 or greater intensity and found that tornadoes are more likely to occur with LFC heights below 2 km and much more likely below 1.5 km. (The corresponding LFC in this case was 1.8 km.) Thus, if rapidly developing cells are occurring over regions of near-surface vertical vorticity or streamwise horizontal vorticity, a lower LFC might increase the potential for tornadogenesis. Of course future studies may be needed to access whether this strategy can be successfully used in an operational environment.

Acknowledgments

The lead author thanks Morris Weisman for his invaluable comments relating to storm structure, Andrew Nash for his editorial comments in an earlier version of the manuscript, and DeWayne Mitchell for his comments relating to the TVS. The lead author further recognizes helpful comments from David Parsons, Jim Wilson, David Blanchard, Leslie Lemmon, and Michael Biggerstaff. We thank Peter Ray for his interpretation of selected WSR-88D data.

REFERENCES

  • Blanchard, D. O., 1998: Assessing the vertical distribution of convective available potential energy. Wea. Forecasting,13, 870–877.

    • Crossref
    • Export Citation
  • Brady, R. H., and E. J. Szoke, 1989: Case study of nonmesocyclone tornado development in northeast Colorado: Similarities to waterspout formation. Mon. Wea. Rev.,117, 843–856.

    • Crossref
    • Export Citation
  • Brown, R. A., L. R. Lemmon, and D. W. Burgess, 1978: Tornado detection by pulsed Doppler radar. Mon. Wea. Rev.,106, 29–38.

    • Crossref
    • Export Citation
  • Crum, T. D., and R. L. Alberty, 1993: The WSR-88D and the WSR-88D Operational Support Facility. Bull. Amer. Meteor. Soc.,74, 1669–1687.

    • Crossref
    • Export Citation
  • Davies-Jones, R., 1984: Streamwise vorticity: The origin of updraft rotation in supercell storms. J. Atmos. Sci.,41, 2991–3006.

    • Crossref
    • Export Citation
  • Donaldson, R. J., Jr., 1970: Vortex signature recognition by a Doppler radar. J. Appl. Meteor.,9, 661–670.

    • Crossref
    • Export Citation
  • Djuric, D., 1994: Weather Analysis. Prentice-Hall, 304 pp.

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

    • Crossref
    • Export Citation
  • Fujita, T. T., 1981: Tornadoes and downbursts in the context of generalized planetary scales. J. Atmos. Sci.,38, 1511–1534.

    • Crossref
    • Export Citation
  • Hodanish, S., D. Sharp, W. Collins, and C. Paxton, 1997: A 10-yr monthly lightning climatology of Florida: 1986–95. Wea. Forecasting,12, 439–448.

  • Johnson, J. T., P. L. MacKeen, A. Witt, E. D. Mitchell, G. J. Stumpf, M. D. Eilts, and K. W. Thomas, 1998: The Storm Cell Identification and Tracking (SCIT) algorithm: An enhanced WSR-88D algorithm. Wea. Forecasting,13, 263–276.

    • Crossref
    • Export Citation
  • Klemp, J. B., 1987: Dynamics of tornadic thunderstorms. Annu. Rev. Fluid Mech.,19, 369–402.

    • Crossref
    • Export Citation
  • Lee, B. D., and R. B. Wilhelmson, 1996: The numerical simulation of non-supercell tornadogenesis. Preprints, 18th Conf. on Severe Local Storms, San Francisco, CA, Amer. Meteor. Soc., 200–204.

  • Leslie, L. M., 1971: The development of concentrated vortices: A numerical study. J. Fluid Mech.,48, 1–21.

    • Crossref
    • Export Citation
  • Markowski, P. M., E. N. Rasmussen, J. M. Straka, and D. C. Dowell, 1998: Observations of low-level baroclinicity generated by anvil shadows. Mon Wea. Rev.,126, 2942–2958.

    • Crossref
    • Export Citation
  • Mitchell, E. D., S. V. Vasiloff, G. J. Stumpf, A. Witt, M. D. Eilts, J. T. Johnson, and K. W. Thomas, 1998: The National Severe Storms Laboratory Tornado Detection Algorithm. Wea. Forecasting,13, 352–366.

    • Crossref
    • Export Citation
  • Mueller, C. K., and R. E. Carbone, 1987: Dynamics of a thunderstorm outflow. J. Atmos. Sci.,44, 1879–1898.

    • Crossref
    • Export Citation
  • NOAA, 1995: Storm Data. Vol. 37. National Climatic Data Center. [Available from National Climatic Data Center, Federal Building, 151 Patton Ave., Asheville, NC 28801.].

  • ——, 1997: Storm Data. Vol. 39. National Climatic Data Center. [Available from National Climatic Data Center, Federal Building, 151 Patton Ave., Asheville, NC 28801.].

  • NSSL, 1997: WSR-88D Algorithm Testing and Display System Reference Guide, Version 9.0, 179 pp. [Available from Storm Scale Research Application Division, National Severe Storms Laboratory, 1313 Halley Circle, Norman, OK 73069.].

  • Paxton, C. H., R. D. Pelatti, and M. E. Powell, 1996: Radiosonde climatology of west-central Florida. NOAA Tech. Memo. NWS SR-181, 19 pp. [Available from NTIS, 5285 Port Royal Rd., Springfield, VA 22161.].

  • Priegnitz, D. L., 1995: IRAS: Software to display and analyze WSR-88D radar data. Preprints, 11th Int. Conf. on Interactive Information and Processing Systems for Meteorology, Oceanography, and Hydrology, Dallas, TX, Amer. Meteor. Soc., 197–199.

  • Purdom, J. F. W., and K. Marcus, 1982: Thunderstorm trigger mechanisms over the southeast United States, Preprints, 12th Conf. on Severe Local Storms, San Antonio, TX, Amer. Meteor. Soc., 487–488.

  • Roberts, R. D., and J. W. Wilson, 1995: The genesis of three nonsupercell tornadoes observed with dual Doppler radar. Mon. Wea. Rev.,123, 3408–3436.

    • Crossref
    • Export Citation
  • Rogers, E., T. L. Black, D. G. Deaven, and G. J. DiMego, 1996: Changes to the operational “early” Eta analysis/forecast system at the National Centers for Environmental Prediction. Wea. Forecasting,11, 391–413.

    • Crossref
    • Export Citation
  • Rotunno, R., 1981: On the evolution of thunderstorm rotation. Mon. Wea. Rev.,109, 577–586.

    • Crossref
    • Export Citation
  • Saucier, W. J., 1955: Principles of Meteorological Analysis. University of Chicago Press, 438 pp.

  • Stumpf, G. J., A. Witt, E. D. Mitchell, P. L. Spencer, J. T. Johnson, M. D. Eilts, K. W. Thomas, and D. W. Burgess, 1998: The National Severe Storms Laboratory mesocyclone detection algorithm for the WSR-88D. Wea. Forecasting,13, 304–326.

    • Crossref
    • Export Citation
  • Trapp, R. J., and R. Davies-Jones, 1997: Tornadogenesis with and without a dynamic pipe effect. J. Atmos. Sci.,54, 113–133.

    • Crossref
    • Export Citation
  • ——, E. D. Mitchell, G. A. Tipton, D. W. Effertz, A. I. Watson, D. L. Andra, and M. A. Magsig, 1998: Descending and non-descending tornadic vortex signatures. Preprints, 19th Conf. on Severe Local Storms, Minneapolis, MN, Amer. Meteor. Soc., 128–131.

  • Wakimoto, R. M., and J. W. Wilson, 1989: Non-supercell tornadoes. Mon. Wea. Rev.,117, 1113–1140.

    • Crossref
    • Export Citation
  • ——, and J. K. Lew, 1993: Observations of a Florida waterspout during CaPE. Wea. Forecasting,8, 412–423.

  • Weisman, M. L., and J. B. Klemp, 1982: The dependence of numerically simulated convective storms on vertical wind shear and buoyancy. Mon. Wea. Rev.,110, 504–520.

    • Crossref
    • Export Citation
  • ——, and —— 1984: The structure and classification of numerically simulated convective storms in directionally varying wind shears. Mon. Wea. Rev.,112, 2479–2498.

    • Crossref
    • Export Citation
  • ——, and —— 1986: Characteristic of isolated convective storms. Mesoscale Meteorology and Forecasting, P. S. Ray, Ed., Amer. Meteor. Soc., 331–358.

    • Crossref
    • Export Citation
  • Wilczak, J. M., T. W. Christian, D. E. Wolfe, R. J. Zamora, and B. Stankov, 1992: Observations of a Colorado tornado. Part I: Mesoscale environment and tornadogenesis. Mon Wea. Rev.,120, 497–520.

    • Crossref
    • Export Citation
  • Ziegler, C. L., P. S. Ray, and D. R. MacGorman, 1986: Relations of kinematics, microphysics and electrification in an isolated mountain thunderstorm. J. Atmos. Sci.,43, 2098–2114.

    • Crossref
    • Export Citation

APPENDIX A

Selected NSSL TDA Parameters

The following are selected NSSL TDA default parameters used in build 10 of the WSR-88D [see Mitchell et al. (1998); WSR-88D Handbook, Vol. 1, Radar Product Generator (RPG)].

  • Minimum required shear segment velocity difference:11 m s−1,
  • Minimum number of shear segments for a 2D detection: 3,
  • Maximum aspect ratio (radial diameter/azimuthal diameter): 4,
  • Minimum required number of 2D detections to declare a 3D detection: 3,
  • Minimum required depth for a 3D detection to declare a TVS or ETVS1: 1.5 km,
  • Minimum base height for TVS detection (see footnote 1): 0.5° or below 0.6 km ARL,
  • Minimum velocity difference required at the base or at any level within a 3D detection to declare a TVS:25 m s−1 and 36 m s−1, respectively.

APPENDIX B

Misocyclone Criteria

The following are the criteria used to identify and to calculate the azimuthal shear associated with the circulation depicted in Fig. 9.

The diameter2 of the circulation must be ⩽4 km to identify as a misocyclone as opposed to a mesocyclone (Fujita 1981).

The aspect ratio (radial diameter/azimuthal diameter) must be ⩽2 to identify as a circulation as opposed to simple horizontal shear.

The azimuthal shear was calculated by dividing the maximum differential velocity3 by the diameter (see footnote 2) of the circulation.

Fig. 1.
Fig. 1.

Photograph of the large waterspout over Tampa Bay viewed from the west.

Citation: Weather and Forecasting 15, 1; 10.1175/1520-0434(2000)015<0122:TJPCFT>2.0.CO;2

Fig. 2.
Fig. 2.

WSR-88D locations of the precursor circulation (1744–1759 UTC), tornado (1804–1809 UTC), and waterspout (1814 UTC) relative to the Tampa Bay area and the foregoing radar.

Citation: Weather and Forecasting 15, 1; 10.1175/1520-0434(2000)015<0122:TJPCFT>2.0.CO;2

Fig. 3.
Fig. 3.

Eta Model initialization at 250 (upper-left), 500 (upper-right), 850 (lower-left), and 1000 mb (lower-right) for 12 Jul 1995 at 1200 UTC. At 250 mb, solid lines depict geopotential heights (in m); dashed lines indicate isotachs (in m s−1). At 500 mb, solid lines depict geopotential heights (in m); dashed lines reveal absolute vorticity (in ×105 s−1). At 850 mb, solid lines depict geopotential heights (in m);dashed lines reveal equivalent potential temperature (K). Wind barbs in each image are in m s−1, with full (half) barb equal to 10 (5) m s−1. The thick dashed lines in each image designate the wind shift associated with the trough.

Citation: Weather and Forecasting 15, 1; 10.1175/1520-0434(2000)015<0122:TJPCFT>2.0.CO;2

Fig. 4.
Fig. 4.

(a) Ruskin, FL, skew T–logp for 1200 UTC 12 Jul 1995. Ordinate (pressure) is in mb. Abscissa (temp) is in °C. Thick solid lines depict temperature (T) and dewpoint (Td). Thin sold line between T and Td is the wet-bulb temperature (Tw). Path of parcel is the most unstable path and is also based on virtual temperature. The half (full) wind barb represents 2.5 (5) m s−1. (b) Ruskin mean hodograph averaged over the period 1700–1800 UTC (12 Jul 1995). Data source is the Ruskin WSR-88D VWP output. The values along the hodograph represent the height (km) at each point. The M denotes the position of the 0–6-km pressure-weighted mean wind for this hodograph. The S is the position of the mean storm motion of the TC, averaged over its life.

Citation: Weather and Forecasting 15, 1; 10.1175/1520-0434(2000)015<0122:TJPCFT>2.0.CO;2

Fig. 5.
Fig. 5.

(a) Visible satellite image (which includes central FL) for 12 Jul 1995 at 1545 UTC. Letter A denotes the convective cell responsible for generating the OFB. The arrow labeled B points to the convection over the Tampa area that produced the QSB. Letter C depicts the drier air north of the surface trough.

Citation: Weather and Forecasting 15, 1; 10.1175/1520-0434(2000)015<0122:TJPCFT>2.0.CO;2

Fig. 5.
Fig. 5.

(Continued) (b) Visible satellite image (which includes central FL) at 1732 UTC 12 Jul 1995. Note the identifications of cells N and S, the QSB, and the OFB. The green letter T, located on the QSB, depicts the location of the tornado at 1804 UTC, and the blue TBW denotes the location of the TBW WSR-88D.

Citation: Weather and Forecasting 15, 1; 10.1175/1520-0434(2000)015<0122:TJPCFT>2.0.CO;2

Fig. 6.
Fig. 6.

Composite reflectivity (left column), 0.5° base reflectivity (center column), and 0.5° base radial velocity (right column) for the 1739 (top row), 1759 (center row), and 1804 UTC (bottom row) volume scans. Top legend denotes equivalent reflectivity (units of dBZ); truncated below 24 dBZ), and bottom legend depicts base radial velocity (units are m s−1). Equivalent lines in (a)–(c) denote the center of the cyclonic horizontal shear region depicted in (c). Equivalent lines in (d)–(i) refer to the corresponding cross sections in Fig. 8.

Citation: Weather and Forecasting 15, 1; 10.1175/1520-0434(2000)015<0122:TJPCFT>2.0.CO;2

Fig. 7.
Fig. 7.

(a) Base reflectivity and (b) base radial velocity at the 8.7° elevation angle for the 1809 UTC volume scan. (e) Base reflectivity and (f) base radial velocity at the 12° elevation angle for the 1809 UTC volume scan. Equivalent lines, which slice across the midlevel vortex pair in (a), (b) [(e), (f)], correspond to cross sections (c), (d) [(g), (h)]. For each cross section, the ordinate (abscissa) is height (horizontal distance) in km. Sets of numbers at the base of each cross section refer to the azimuth (°) and range (km), relative to the TBW WSR-88D, at the cross section edges. The distance between tick marks at the base of each cross section is 5 km. The top legend denotes reflectivity (dBZ), and bottom legend depicts base radial velocity (m s−1). The arrow in (b) points to the tornadic circulation.

Citation: Weather and Forecasting 15, 1; 10.1175/1520-0434(2000)015<0122:TJPCFT>2.0.CO;2

Fig. 8.
Fig. 8.

RHI cross sections of base reflectivity (left column) and base radial velocity (right column) for 1759 (top row) and 1804 UTC (bottom row) volume scans. Sets of numbers at the base of each cross section refer to the azimuth (°) and range (km) relative to the TBW WSR-88D at the edge of each cross section. The distance between tick marks at the base of each cross section is 5 km. Height is in units of km. The arrows in (b) and (d) point to the position of the near-surface tornadic circulation. The top legend denotes base reflectivity (dBZ), and the bottom legend depicts base radial velocity (m s−1).

Citation: Weather and Forecasting 15, 1; 10.1175/1520-0434(2000)015<0122:TJPCFT>2.0.CO;2

Fig. 9.
Fig. 9.

Time (abscissa) vs height (ordinate) of base reflectivity (solid lines; units of dBZ) at the center of the developing tornadic circulation, and the azimuthal shear (dashed lines; units of ×10−3 5−1) of the circulation. The shaded region is base reflectivity ≥50 dBZ. Thick solid line at the base of the figure depicts the approximate tornado/waterspout duration. Time is in UTC. Height is in m.

Citation: Weather and Forecasting 15, 1; 10.1175/1520-0434(2000)015<0122:TJPCFT>2.0.CO;2

1

If a 3D detection meets the minimum required depth, yet does not meet the minimum base height, the detection would be classified as an elevated TVS (ETVS).

2

The diameter of the radar-observed circulation at a given elevation and range is defined as the distance separating the maximum inbound and outbound radial velocities.

3

The difference between the maximum inbound and outbound radial velocities is defined as the maximum differential velocity.

Save