WHAT IS A "MINI" OR "LOW-TOPPED" SUPERCELL?
-Members of the supercell spectrum of severe thunderstorms include 1) classic, 2) high precipitation (HP), and 3) low precipitation (LP). Supercells that typically occur over the Plains states have been well researched and documented, and often are large horizontally, tall vertically, contain deep mesocyclones, and frequently produce large hail, damaging winds, and weak to violent tornadoes.
-National Weather Service Doppler radars (WSR-88D), however, have revealed a number of supercells that are definitely smaller in both horizontal and vertical extent than those typical of the Plains. Nevertheless, these smaller storms often possess the same radar attributes, albeit sometimes more subtle, as their larger counterparts, including hook echoes, weak echo regions (WERs), bounded weak echo regions (BWERs), and mesocyclones. In essence, these storms are miniature versions of large, classic supercells and contain low echo tops; thus, they have been called "mini supercells" or "low-topped supercells." On radar, mini supercells can be isolated (similar to classic storms) or embedded within squall lines (similar to HP storms).
ENVIRONMENTAL CONDITIONS CONDUCIVE TO MINI SUPERCELLS
-A low equilibrium level (EL) above which the rising parcel is no longer positively buoyant, or a low tropopause level often is present which prevents/caps off deeper convection.
-The ambient instability usually is weak-to-moderate, i.e., CAPE (positive buoyant energy) values of 300-1500 J kg-1 with average values about 600-1000 J kg-1. Lifted index (LI) values typically range from 0 to -4.
-A modest CAPE value, however, may be misleading. Due to the low EL, the CAPE often is contained in a smaller vertical distance between the level of free convection (LFC, above which a parcel is positively buoyant compared to its surrounding environment) and the EL than for taller storms. Therefore, despite marginal values, the positive energy area (CAPE) on a sounding may still be relatively "fat" (horizontally wide) and similar to the amount of CAPE in the low-to-mid levels of environments conducive to taller storms (Fig. 1). As a result, the mini supercellís updraft will not extend as high as that of a taller supercell, nor will its maximum updraft strength be as strong as that of a large storm. However, the mini stormís updraft speed may be comparable at equal altitudes below the mini stormís EL as that of tall storms (Wicker and Cantrell 1996; Fig. 2). Thus, in assessing instability, be wary of only using the total CAPE in observed or model soundings as an indicator of severe storm potential; the vertical distribution of parcel buoyancy appears to be more important.
|Figure 1. Sample soundings used for a numerical simulation of mini and high-topped supercells. The CAPE (shaded area) for the simulated mini storm (left) is only 600 J kg-1 with a low EL of 7 km (just below 400 mb); that for the higher storm (right) is 2200 J kg-1 and 12 km (about 200 mb), respectively. However, note that the amount of CAPE below about 450 mb is nearly the same for both cases, i.e., the positive area is relatively "fat" (wide) in the low-to-mid levels of the mini stormís sounding.||Figure 2. Comparison of time averaged vertical velocity profiles for 3 simulated supercells using different CAPE (but similar shear) values. Note that the low CAPE mini stormís updraft strength is comparable to that of the high CAPE storm up to about 5 km height, then weakens above this level due to the low EL.|
-Moderate-to-strong vertical wind shear (at least 40 kts/20 m s-1 combined speed and directional shear), similar to large supercell environments, usually is required to promote mesocyclone development.
-Storm-relative helicity (0-3 km) values generally range from 200-500 m2 s-2 suggesting mesocyclone development is likely given storm occurrence.
-Bulk Richardson Number (BRN) values generally are about 10 or less, indicating the presence of strong vertical shear and relatively weak CAPE. This suggests supercell and mesocyclone development are likely given storm development.
REFLECTIVITY STRUCTURE ASSOCIATED WITH MINI SUPERCELLS
-In reflectivity data, mini supercells are smaller in horizontal extent than those of classic, taller storms. However, they still can exhibit hook echoes, WERs, and BWERs, although smaller and sometimes more subtle than those of large storms. Figure 3 shows a numerical simulation (Wicker and Cantrell 1996) of a low-level reflectivity display from a high-topped (CAPE=2200 J kg-1) versus low-topped (CAPE=600 J kg-1) supercell using the same amount of shear. Note that a WER and downwind V-Notch (indicative of blocking in the flow aloft) are evident in both storms, although those of the low CAPE (mini) supercell are more subtle.
-Mini supercell echo tops are relatively low, and typically range from 24000-32000 ft. The same simulation of a deep versus shallow supercell (Fig. 4) shows an average echo top of nearly 15 km (45000-50000 ft) for the higher CAPE storm, but only about 9 km (29000 ft) for the low CAPE storm. The horizontal extent of the high-topped supercell also was larger. However, despite the size difference, a low-level WER and downwind tilt aloft existed in both supercell storms.
-Maximum reflectivities often are not much more than 50 or 55 dBZ, and the horizontal and vertical coverage of 50 dBZ returns are smaller than that with high-topped supercells.
-An example of an actual mini supercell in low-level reflectivity data is shown in Fig. 5 (Burgess et al. 1995). The supercell, as viewed from the Sterling, VA WSR-88D on July 27, 1994, exhibited a hook echo, although it was smaller and a little more subtle than those typical of large, classic storms. Note that only a small area of 50 dBZ returns was evident in the storm in low-levels.
|Figure 3. Simulation of a low-level reflectivity horizontal plot of a low CAPE mini supercell (left) and a high CAPE traditional supercell (right). Both storms show similar features (WER on southern side of the storm and a V-notch downwind), although those of the mini storm are more subtle.|
|Figure 5. Base reflectivity image (0.5 deg elevation) from the Sterling, VA WSR-88D at 2317 UTC on July 27, 1994. A mini supercellís hook echo is evident in the region of greater than 45 dBZ values (orange). Note, however, that only a very small area of 50 dBZ or more (red) exists in this low-level reflectivity image. NOTE: The image was colorized from a black and white photo (right portion of image).|
|Figure 4. Simulation of a vertical cross-section of reflectivity through the updraft of a low CAPE mini supercell (left) and a high CAPE traditional supercell (right). Despite the smaller mini storm, both storms again show similar features (BWER and echo overhang).|
MESOCYCLONE STRUCTURE ASSOCIATED WITH MINI SUPERCELLS
-Mini supercell mesocyclones often have lower magnitudes of rotational velocity (Vr, i.e., maximum inbound plus maximum outbound winds within a cyclonic circulation divided by 2), smaller diameters, and shallower depths than those of classic storms studied in the Plains (see table below). Mini mesocyclones typically fall only within the "minimal" or "moderate" category on the mesocyclone strength nomogram developed for larger supercells (Andra et al. 1994). This may partly be due to radar sampling issues associated with a relatively small physical size of the mesocyclone (see below for more information). Nevertheless, mini supercells can produce tornadoes similar to their larger counterparts. The majority of tornadoes produced by mini storms are of weak-to-moderate intensity (F0-F2), although some can be strong (F3).
-The vertical extent of mesocyclones within supercells appears to be linked strongly to the depth of the stormís updraft. A simulation of 3 supercells using different CAPE values (but similar shear) revealed that the updraft speed for all 3 CAPE values was comparable below 5 km (16000 feet), although the low CAPE (600 J kg-1) stormís updraft peaked in strength around 5-6 km while that of the high CAPE (2200 J kg-1) storm peaked at higher values at about 10 km (Wicker and Cantrell 1996; Fig. 2). Thus, the mesocyclone was confined to a depth of 7 km for the simulated mini supercell but much deeper for the large storm. Foster and Moller (1995) noted that in a study of a Texas tornado-producing mini supercell, its mesocyclone was confined to a depth of 6 km (20000 ft).
-In an independent study of radar-observed mini (MINI) versus traditional (TRAD) supercells across the United States (Burgess et al. 1995), the following results were tabulated for their mid-level mesocyclones. Stated values are only averages and can vary from storm to storm. The study included storms within 55 nm of radar sites only.
|Mesocyclone Stage||Avg Max Vr (kt)||Diameter (km)||Height of Max Vr (ft)|
|Mature-Mid Level (MINI)||32||2.0||15000|
|Mature-Mid Level (TRAD)||48||3.3||30000|
|Mature-Low Level (MINI)||28||1.9||Sfc-3000|
|Mature-Low Level (TRAD)||44||2.9||Sfc-3000|
-In the table above, note that the mesocyclones of mini storms on average were weaker throughout their life cycle than the traditional classic mesocyclones, although both types increased in strength by about the same percentage (25 percent) from the organizing to mature stages. During the mature stage, the mini supercellsí low- and mid-level circulations were strongest, as expected. NOTE: The numerical simulation of mini versus traditional supercells discussed above (Wicker and Cantrell 1996) found less difference in the Vr values of their respective low-level mesocyclones versus those results shown in the table above for observed storms (Burgess et al. 1995).
-As with traditional supercells, tornadogenesis is most likely given the presence of low-level cyclonic convergence (in storm-relative velocity data), the presence of a relatively warm rear flank downdraft, and favorable baroclinic generation of horizontal vorticity capable of being tilted into the vertical updraft. In addition, tornadogenesis in mini storms correlates well with decreasing mesocyclone height, increasing low-level and maximum Vr values, and increasing maximum shear within the mesocyclone, i.e., when the highest Vr value is combined with the smallest diameter (Grant and Prentice 1996). The shear attribute seems to exhibit the greatest correlation to tornadogenesis.
MINI SUPERCELL RADAR SAMPLING ISSUES
-Due to their diminutive horizontal and vertical size, proper mini supercell detection and assessment may be more difficult than for larger, deeper (classic) supercells. In addition, classic and especially mini storm attributes will be more difficult to discern at long ranges, making knowledge of the storm environment and viewing of data from adjacent, closer radars very important.
-Due to the relatively small diameters of mini supercell mesocyclones, their radar-displayed strength (Vr values) may be lower than in reality due to velocity averaging in neighboring radar beams. This is a reason why apparent "minimal" or "moderate" mesocyclones are capable of producing tornadoes.
-In general, detection of mini supercells at ranges greater than 60 nm (110 km) may suffer from some loss of resolution (although still be viewable), while detections at ranges over 90 nm (170 km) may be difficult in certain situations (Burgess et al. 1995).
EXAMPLES OF MINI SUPERCELLS IN KENTUCKY AND SOUTHERN INDIANA
-Three examples of severe thunderstorms with mini supercell characteristics that occurred in Kentucky and southern Indiana are discussed below.
-Mini supercell embedded in a rain shield:
At 1856 UTC May 18, 1995, a large shield of rain was occurring over central and eastern Kentucky with convection in the southern portion of this area (Fig. 6). A severe storm with supercell characteristics was evident over extreme south-central Kentucky; an examination of Nashvilleís radar (not shown) would be useful to help assess this storm. However, the "main storm of interest" in Fig. 6 is circled and embedded within the rain shield. This mini supercell storm is examined more closely in Figs. 7 and 8.
Figure 6. Base reflectivity image (0.5 deg elevation) from the Louisville KLVX WSR-88D radar at 1856 UTC May 18, 1995. The area shown is central and eastern Kentucky. The subtle "main storm of interest" (circled) across south-central Kentucky had mini supercell characteristics and was embedded in the southern portion of a large area of rain.
A close-up 4-panel reflectivity display
of the "main storm of interest" at 1856 UTC revealed a mini supercell along
the border of Adair, Casey, and Russell counties in south-central Kentucky (Fig.
7). Note that a subtle hook echo and weak echo region (WER) were
evident in the 45-55 dBZ contours (orange and red colors) on the southern
inflow flank of the storm in extreme northern Russell county at 0.5 degrees
elevation (about 6800 ft). Aloft, however, the storm was relatively shallow
with only 30-40 dBZ returns noted at about 26000 ft altitude. Nevertheless,
the maximum returns at this level appeared tilted downwind from the maximum
The corresponding storm-relative
velocity map (SRM) data at 1856 UTC (Fig. 8)
showed a well-defined circulation at 0.5 degrees elevation coincident with the
subtle WER in reflectivity data. As measured by the Louisville KLVX WSR-88D
radar located 65 nm northwest of the storm, the mesocycloneís rotational
velocity (Vr) value at this level was approximately 31 kts, i.e., a
"moderate" mesocyclone. The mesocyclone also was detectable at 1.5 degrees
elevation (about 13500 ft) where its Vr was 24 kts (a "minimal"
mesocyclone). Outbound radial velocities were dominant at 1.5 degrees,
apparently due to radar sampling considerations. The circulation was weaker at
2.4 degrees and not apparent higher in altitude. Despite its subtle
appearance, this mini supercell produced a brief F1 tornado across far
northern Russell county resulting in uprooted trees and some structural
In this case, although limited instability and/or a low EL prevented deeper convection, strong vertical wind shear was present in the storm environment (Fig. 9). At the radar site, winds veered from southeast to southwest and increased from 15 to about 70 kts between the surface and 6 km altitude. The degree of shear may have been even greater over south-central Kentucky along the southern end of the rain shield. The storm environment, characterized by strong, deep-layered shear and reduced CAPE, was conducive to mini supercell and mesocyclone development. After the rain and storms moved out of central Kentucky, air mass destabilization occurred later in the day on May 18. Then during the evening, strong forcing combined with the greater instability and strong shear resulted in the development of 2 larger, taller classic supercells which produced tornadoes.
|Figure 7. Above: A 4-panel close-up of base reflectivity data at 1856 UTC May 18, 1995 from the Louisville KLVX WSR-88D radar. The "main storm of interest" (from Fig. 6) was along the Russell, Adair, and Casey County border in south-central Kentucky. Radar elevation angles and approximate beam heights over the main storm area are: 0.5 deg/6800 ft (upper left panel); 1.5 deg/13500 ft (upper right); 2.4 deg/19500 ft (lower left); and 3.4 deg/26000 ft (lower right). Values greater than 50 (35) dBZ are in red (yellow). A subtle hook echo was evident on the southern edge of the storm (it may have been more distinct at lower levels) with echo overhang aloft to the northeast of the low-level hook.|
Figure 9. Above: Vertical wind profile (VWP) from the Louisville KLVX radar near the time of the mini supercell shown in Figs. 7 and 8. The winds are in kts while the vertical coordinate numbers are in thousands of feet. The profile showed strong deep-layered wind shear supportive of rotating updrafts and supercells. In addition, significant low-level directional shear was present which provided a more conducive environment for low-level mesocyclone and tornado development.
|Figure 8. A 4-panel close-up of storm-relative velocity data at 1856 UTC May 18, 1995 (companion image to Fig. 7). Radar elevation angles and approximate beam heights are the same as those in Fig. 7, except that the 4.3 deg angle (about 30000 ft) is shown in the lower right panel. Red (green) colors are radial winds directed away from (toward) the radar located 65 nm northwest of the main storm. The mesocyclone associated with the mini supercell was evident over northern Russell County, especially in the top 2 panels where outbound red values were most pronounced. It should be noted that the circulation may have been stronger and tighter in low-levels below the radar beam (i.e., the 0.5 deg upper left panel was at about 6800 feet at this distance from the radar, and could not detect features below this level).|
-Mini supercell in southern Indiana:
An area of convection (not shown) over south-central Indiana organized into a mini supercell across southern Jackson County during the early morning hours on April 9, 1999. The organization was enhanced by an axis of low-level convergence along the outflow boundary (updraft/downdraft interface) from the convection, which strengthened storm-relative inflow and helped cause the development of low and mid-level storm rotation. As a result, supercell characteristics began to appear in reflectivity and storm-relative velocity data (SRM) (Fig. 10).
The 0.5 degree base reflectivity image
revealed a rather subtle weak echo region (WER; developing hook echo) on the
southern flank of the storm with weaker returns along the rear flank downdraft
just southwest of the WER. Despite the somewhat tenuous appearance, SRM data
showed a well-defined cyclonic circulation (low-level mesocyclone) in extreme
southern Jackson County within the WER. In addition, an axis of storm-relative
inflow over eastern Washington County was directed into the WER/hook echo,
apparently enhancing low-level mesocyclone development and maintenance.
A reflectivity vertical cross-section revealed a shallow, mini storm in that the 50 dBZ and higher core (red colors) only extended to just above 20000 feet. However, storm tilt was present along with a WER below 10000 feet. This storm produced a tornado in Jackson County, then a separate, stronger tornado near Cincinnati as the storm continued northeast and strengthened.
|Figure 10: Low-level (0.5 deg/4300 feet height) base reflectivity (left) and storm-relative velocity data (SRM; middle), and reflectivity vertical cross-section (right) from the Louisville WSR-88D radar at 0716 UTC April 9, 1999 over south-central Indiana. Although not prominent, base reflectivity suggested the presence of a WER and developing hook echo on the southern flank of the storm, with weaker convection apparently along the rear flank downdraft on the southwest edge of the storm. In the WER, a tight cyclonic circulation (bright red outbounds next to bright green inbound winds) was clearly evident in extreme southern Jackson County (middle image) with an axis of storm-relative inflow (red colors) across eastern Washington County extending into far southern Jackson County. The reflectivity cross-section showed this to be a mini storm as the 50 dBZ core (red colors) only extended to just over 20000 feet. However, supercell characteristics were quite noticeable given storm tilt with height and the presence of a WER below 10000 feet (just to the right of the 50 dBZ core).|
-Isolated mini supercell in south-central Kentucky:
During the evening of June 17, 1997, a mini supercell storm produced a brief F0 tornado in Butler County in south-central Kentucky. A 4-panel base reflectivity display of this storm at 0111 UTC June 18 (Fig. 11; black and white drawing taken from radar imagery) revealed that the echo top was below 30000 ft. At the lowest elevation slice (0.5 degrees), highest reflectivity values (50-55 dBZ) and a subtle WER were located on the southwestern inflow flank of the storm in southern Butler County. Aloft, the reflectivity core tilted slightly downwind to the northeast.
The corresponding SRM data at 0111 UTC (Fig. 12) showed a small, but definitive circulation coincident with the low-level WER. Displayed Vr values at 0.5 degrees (7500 ft) were about 26 kts (radar located 65 nm northeast of the main storm), placing the mesocyclone in the "upper minimal/lower moderate" range. A "minimal" mesocyclone (Vr about 22 kts) was noted at 1.5 degrees (15000 ft), with weaker values at 2.4 degrees as only inbound radial velocities were noted.
The vertical wind profile (VWP) from the Louisville KLVX WSR-88D radar (Fig. 13) showed that moderate wind shear, predominantly directional in low levels but with speed shear in upper levels, existed during the life span of the storm. The exact shear profile over Butler County was unknown, but likely was similar to that at the radar site. CAPE values were 1200-1500 J kg-1 in the storm environment. In this case, subtle reflectivity characteristics combined with SRM data suggested a severe storm.
|Figure 11. Above: A 4-panel display of base reflectivity data at 0111 UTC June 18, 1997 from the Louisville KLVX WSR-88D. Radar elevation angles and approximate beam heights over the main storm area are: 0.5 deg/7500 ft (upper left panel); 1.5 deg/15000 ft (upper right); 2.4 deg/20500 ft (lower left); and 3.4 deg/26000 ft (lower right). Values are in dBZ; values greater than 40 dBZ are shaded. An isolated mini supercell storm is shown in Butler ("B") County in south-central Kentucky.|
|Figure 13. Above: Vertical wind profile (VWP) from the Louisville KLVX radar near the time of the mini supercell shown in Figs. 11 and 12. The winds are in kts while the vertical coordinate numbers are in thousands of feet. The profile shows moderate deep-layered wind shear, including speed shear aloft near jet stream level and low-level directional shear. This shear was sufficient to support updraft rotation in the storm.|
|Figure 12. A 4-panel display of storm-relative velocity (SRM) data at 0111 UTC June 18, 1997 (companion image to Fig. 11). Radar elevation angles and approximate beam heights over the main storm are the same as those in Fig. 11. SRM values are contoured every 10 kts. Solid (dashed) lines are radial winds directed toward (away from) the radar located 65 nm northeast of the main storm. A "weak" mesocyclone is shaded in the southwest portion of the storm.|
-It is not necessary to have large and tall supercells and mesocyclones to produce tornadoes and damaging winds. Mini supercells are capable of producing weak to strong tornadoes, similar to their taller counterparts.
-Mini supercells usually have similar attributes (hook echo, WER, BWER, mesocyclone, storm tilt) as their taller, traditional counterparts. These features, however, appear to scale down as the storm depth decreases, thereby requiring careful radar analysis and knowledge of potential sampling issues.
-Mini supercells can be isolated storms (similar to large, tall classic cells), or can be embedded within a rain shield or squall line (similar to HP supercells).
-Forecasters must be able to anticipate days when small supercell events are possible, i.e., those environments with relatively low instability and ELs, but with strong wind shear and high storm-relative helicity values. In such environments, forecasters must alter their thinking and be ready to warn, if necessary, for smaller severe storms.
Andra, D., V. Preston, E. Quetone, D. Sharp, and P. Spoden, 1994: An operational guide to configuring a WSR-88D principal user processor. Operations Training Branch, Operational Support Facility, Norman, OK.
Burgess, D.W., R.R. Lee, S.S. Parker, and D.L. Floyd, 1995: A study of mini supercells observed by WSR-88D radars. Preprints, 27th Conference on Radar Meteorology, Vail, CO, Amer. Meteor. Soc., 4-6.
Foster, M.P. and A.R. Moller, 1995: The rapid evolution of a tornadic small supercell: observations and simulation. Preprints, 14th Conference on Weather Analysis and Forecasting, Dallas, TX, Amer. Meteor. Soc., 323-328.
Grant, B. and R. Prentice, 1996: Mesocyclone characteristics of mini supercell thunderstorms. Preprints, 15th Conference on Weather Analysis and Forecasting, Norfolk, VA, Amer. Meteor. Soc., 362-365.
Wicker, L.J. and L. Cantrell, 1996: The role of vertical buoyancy distributions in miniature supercells. Preprints, 18th Conference on Severe Local Storms, San Francisco, CA, Amer. Meteor. Soc., 225-229.