Supercell Development over North-central Mississippi:
Storm-scale Changes to an Unfavorable Pre-storm Environment
NWSFO Jackson, Mississippi
NWSO Wichita, Kansas
Severe thunderstorms developed across northwest Mississippi during the mid-afternoon hours of May 14, 1997. The thunderstorms moved east southeast across north-central Mississippi during the late afternoon hours, while other strong storms developed over western Mississippi. With the exception of one storm, the WSR-88D radar at Jackson, Mississippi, showed no supercell characteristics with the thunderstorms that developed on this day. The exception developed over Holmes County and tracked southeast across Holmes, Attala and Leake counties in north-central Mississippi.
This storm produced golfball-size hail, damaging straight-line winds, and a short-lived F1 intensity tornado as it moved across eastern Holmes County and western Attala County. The pre-storm environment appeared to be somewhat favorable for severe thunderstorms, but unfavorable for supercellular convection. This paper will suggest, using WSR-88D imagery, how storm-scale environmental changes in the vicinity of an apparent outflow boundary led to unexpected supercell formation.
2. Pre-Storm Environment
The pre-storm airmass over Mississippi appeared to be only marginally unstable. The May 15, 1997, 0000 UTC Jackson sounding showed convective available potential energy (CAPE) values below 1000 J Kg-1 for surface-based parcels. However, modifying the sounding (Fig. 1a) using slightly higher surface dew point values of 60 to 65 degrees pooled just ahead of an approaching surface pre-frontal trough showed CAPE values of 1300 to 1500 J Kg-1 for surface-based parcels.
The Jackson hodograph (Fig. 1b) showed moderate speed shear but very little directional shear in the lowest 0-3km layer. The 0-3km storm-relative helicity (SRH) varied from near 60 m2s-2 using the average storm motion vector observed on May 14, to around 110 m2s-2 (Fig. 2a) using the motion vector for the lone supercell detected. These values are below the typical threshold of 150 m2s-2 for mesocyclone formation (Davies-Jones and Burgess, 1990). Furthermore, the 0-6 km shear vector, considered a better indicator of supercell potential than storm-relative helicity (Weisman 1996), was about 22 m s-1. This met the typical threshold of 20 m s-1 but was likely insufficient given the low instability.
An environment with insufficient shear, when coupled with high instability can support supercells (Johns and Doswell 1992). The environment in the May 14 case exhibited low values of both shear and instability, suggesting weak updraft rotation would be much more likely than deep persistent mesocyclones.
The pre-storm environment showed no substantial synoptic or meso-scale features over Mississippi to focus severe thunderstorm development, as illustrated by the May 15, 1997 0000 UTC composite chart (Fig. 3). The upper jet pattern, and the location of an upper-level trough well north of Mississippi, suggested little in the way of synoptic scale vertical motion over the state during the afternoon of May 14. The only apparent focus for thunderstorm development was a weak surface pre-frontal trough over northwest Mississippi and northeast Louisiana during the afternoon.
3. Supercell Development
Reflectivity products from the WSR-88D radar in Jackson, Mississippi, showed the development of a multicellular band of convection over Sunflower, Leflore, Carroll and Holmes counties through 2130 UTC (Fig. 4a). The thunderstorms rapidly became severe within 30 minutes, producing large hail and strong straight-line winds. Between 2130 and 2150 UTC, the westernmost storm in the band became quite large, with maximum reflectivity values greater than 65 dBZ existing through a depth of 5-6 km, and maximum reflectivity values in the 72 to 74 dBZ range.
By 2155 UTC, the Jackson WSR-88D indicated supercell characteristics associated with the large thunderstorm, including a bounded weak echo region (Fig. 5) and the development of a moderate low-level mesocyclone (Fig. 6) with rotational velocities around 15 m s-1 up to a height of 4 km above ground level (AGL). Between 2155 and 2215 UTC, a hook-like feature developed in the reflectivity pattern (Figs. 4b-4c), and the low-level mesocyclone extended upward to a height around 8 km AGL (Fig. 6). The supercell produced hail up to 50 mm (2 in) in diameter, damaging straight-line winds, and a short-lived F1 intensity tornado as it moved across eastern Holmes County and western Attala County. The supercell gradually weakened after 2230 UTC, though it continued to produce severe weather through 2330 UTC.
4. Role of Storm-Scale Environmental Changes
The pre-storm environment revealed CAPE values below 1500 J kg-1, and storm-relative helicity values well below 150 m2s-2, characteristics not typically favorable for supercell thunderstorms. Even so, one of the thunderstorms that developed over north-central Mississippi on May 14, 1997, did exhibit supercell characteristics, including a bounded weak-echo region and a deep, persistent mesocyclone. The modification to the pre-storm environment likely resulted in the transition of a multicell storm to a supercell. WSR-88D imagery suggested the formation of this supercell was likely the result of storm-scale changes to the pre-storm environment in the vicinity of this storm, not the result of the pre-storm environment itself.
The supercell developed on the western edge of a band of multicellular convection over north-central Mississippi (Figs. 4a-4c). An apparent outflow boundary produced by this band of thunderstorms may have played an important role in supercell development. As the westernmost storm in the band moved southeast along this boundary (Fig. 4b), strong horizontal streamwise vorticity produced by the boundary was apparently ingested into the storm's updraft, leading to the development of a strong low-level mesocyclone with rotational velocities of 15-20 m s-1. The mesocyclone developed below 3km AGL around 2150 UTC, and extended upward into the middle portion of the storm (around 8 km AGL) by 2205 UTC (Fig. 6). As this occurred, a hook-like feature developed in the low-level reflectivity pattern (Fig. 4c). Other studies (e.g. Moller et al 1990) have documented the important role of low-level boundaries in low-level mesocyclone development. Such a boundary would have likely provided the strong low-level horizontal streamwise vorticity necessary for low-level mesocyclone generation in the case presented here.
Fig. 2a is the actual hodograph for Jackson, Mississippi, at May 15, 1997, 0000 UTC using the storm motion vector for the supercell. Fig. 2b is the hodograph modified for estimated changes on the cool-side of the apparent outflow boundary being intercepted by the intensifying cell. Assuming low-level easterly inflow on the cool-side of the outflow boundary, the modified hodograph reveals a storm-relative helicity value of about 250 m2s-2. and a 0-6km shear vector of 30 m s-1. In addition, the modified hodograph produces a 0-1 km horizontal streamwise vorticity value of about 32 s-1, much greater in magnitude than the unmodified hodograph value of about 5 s-1. The storm-scale environmental changes in the vicinity of the apparent outflow boundary seems to have made low-level mesocyclone development much more likely.
The thunderstorm gradually lost its supercellular characteristics after 2230 UTC, likely the result of the apparent outflow boundary moving well to the south of the storm (Fig. 4d). As the apparent boundary was moving southward, low-level rotation within the updraft of the supercell weakened from nearly 20 m s-1 to around 10 m s-1 in the lowest 3 km by 2217 UTC (Fig. 6). At the same time, the mid-level rotation remained in the 15 to 20 m s-1 range. The entire mesocyclone quickly dissipated after 2230 UTC.
Supercell thunderstorm development is obviously not just a function of the pre-storm environment given by local soundings. Changes to such environments, which can be substantial especially on the storm-scale, can alter the convective mode. On May 14, 1997, a non-supercell thunderstorm became supercellular as it tracked southeast along an apparent outflow boundary produced by organized convection just to the east. As this cell tracked along the boundary, strong horizontal streamwise vorticity was ingested into the updraft, leading to the formation of a strong low-level mesocyclone.
Knowledge of the pre-storm environment, and severe weather threat, is important for successful warning operations. Equally important is the knowledge of environmental changes, even on the storm-scale level, that may be occurring and how such changes may alter the storm type and primary severe weather threat. Without detailed data such as mesonet observations, radar operators must attempt to determine such changes from other available data, including satellite, wind profiler, and WSR-88D data.
Davies-Jones, R. and D.W. Burgess, 1990: Test of helicity as a tornado forecast parameter. Preprints, 16th Conf. Severe Local Storms, Kananaskis Park, Alberta, Canada, Amer. Meteor. Soc., 588-592.
Johns, R.H. and C.A. Doswell III, 1992: Severe local storms forecasting. Wea. Forecasting, 7, 588-612.
Moller, A.R., C.A. Doswell III and R.W. Przbylinski, 1990: High-precipitation supercells: a conceptual model and documentation. Preprints, 16th Conf. Severe Local Storms, Kananaskis Park, Alberta, Canada, Amer. Meteor. Soc., 52-57.
Weisman, M.L., 1996: On the use of vertical wind shear versus helicity in interpreting supercell dynamics. Preprints, 18th Conf. Severe Local Storms, San Francisco California, Amer. Meteor. Soc., 200-204.
AcknowledgmentsThe authors would like to thank Russell Pfost, science and operation officer and Alan Gerard for reviewing this paper.
1. The May 15, 1997, 0000 UTC Jackson, Mississippi, modified sounding (a) and hodograph (b). The arrow points to the equilibrium level on the sounding and to the storm motion vector on the hodograph.
2. The May 15, 1997, 0000 UTC Jackson, hodograph modified using the observed supercell motion (a), and modified for expected conditions on the cool-side of an apparent outflow boundary (b). The arrow points to the storm motion vector.
3. The May 15, 1997, 0000 UTC synoptic composite chart. Large, thick arrows represent upper-level jets; the large thin arrow represents the low-level jet; small arrows indicate surface flow; surface fronts, 500 mb heights (thin contours), and the 16 degree Celsius surface dew point contour (gray) are also shown.
4. The 0.5 degree base reflectivity product at 2131 UTC (a), 2152 UTC (b), 2217 UTC © and 2232 UTC (d). rear-flank downdrafts (RFD), hook echos (HOOK), and apparent outflow boundaries are labeled.
5. Reflectivity cross-section through the supercell at 2152 UTC showing bounded weak echo region.
6. Time-height series plot of rotational velocity (in m s-1) for the supercell. A thin 14 m s-1 contour is drawn. Bold numbers indicate rotational velocities greater than 15 m s-1, while "<" represents rotational velocities less than 9 m s-1.