Squall Line Produces Severe Weather
Across the Western Carolinas - 4 March 2008
Patrick D. Moore
These well-developed cumulonimbus mammatus clouds were observed east of Asheville, North Carolina, after a severe squall line moved past the mountains.
Author's Note: The following report has not been subjected to the scientific peer review process.
A broken line of severe thunderstorms moved quickly east across northeast Georgia and the western Carolinas during the afternoon and evening of Tuesday, 4 March 2008. The thunderstorms produced numerous damaging wind gusts, mainly across the area south and east of the mountains, as well as a few reports of small hail. Notable damage occurred northeast of Greer, South Carolina, in the area between Highway 29 and Hammett Store Road. Numerous trees were toppled, a vehicle was overturned, and 10 to 15 homes experienced structural damage. A storm survey determined the damage was the result of straight line wind. A brief tornado touched down in the community of Cornatzer in Davie County, North Carolina. The damage was rated at EF-0 on the Enhanced Fujita Scale and was limited to a mobile home and nearby outbuildings on Cornatzer Road. The line of thunderstorms went on to produce widespread damage across the Piedmont and Sandhills of North Carolina and across the Tidewater area of Virginia and North Carolina, and more damage across the Mid-Atlantic region (Fig. 1) later that evening.
Figure 1. Large hail, damaging wind, and tornado reports compiled by the Storm Prediction Center for the 24 hour period ending 1200 UTC 5 March 2008. Click on image to enlarge.
The events of 4 March 2008 were well-anticipated by the National Weather Service (NWS). The severe weather potential was articulated in severe weather outlooks prior to the development of thunderstorms. The Storm Prediction Center (SPC) placed a large area from the northeast Gulf Coast to the Carolinas and Mid-Atlantic region in a Slight Risk of severe weather in the Day 1 Severe Weather Outlook issued at 0542 UTC. Tornado Watches were issued in advance of all severe weather events across the NWS Greenville-Spartanburg (GSP) county warning area. The GSP office issued 19 Severe Thunderstorm Warnings, three Tornado Warnings, and one Flash Flood Warning. All the warnings were verified with the exception of the Tornado Warnings. All reports of damage occurred while a warning was in effect, although a Severe Thunderstorm Warning was in effect during the Cornatzer Tornado and not a Tornado Warning. Using the legacy method for calculating severe weather statistics for this event, the Probability Of Detection was 0.98, the False Alarm Ratio was 0.14, and the Average Lead Time was 23.1 minutes.
Note: All times in this document are referenced to Coordinated Universal Time (UTC), which is Eastern Standard Time plus five hours.
2. Synoptic Features
The ingredients for a severe weather outbreak were expected to come together across the southern Appalachians, Carolinas, and Mid-Atlantic region on 4 March. At 1200 UTC 4 March, a strong 125 knot jet streak over east Texas and Louisiana at 300 mb was expected to lift northeast over the Appalachians late in the day (Fig. 2). This motion brought upper divergence associated with the right entrance region of the upper jet across the western Carolinas in the early evening. At 500 mb, a closed low over Arkansas was expected to move northeast over the Tennessee Valley and acquire a negative tilt (Fig. 3). This motion brought upper diffluence and a 70 knot mid level jet streak across the region late in the day. A dry slot at 700 mb over Texas and Louisiana (Fig. 4) was expected to move northeast and bring a mid-level moisture gradient across the western Carolinas. All of these factors favored large scale upward vertical motion while the nature of the forcing suggested that convection would be linear.
Figure 2. SPC objective analysis of 300 mb isotachs, streamlines, and divergence at 1200 UTC 4 March (left) and 0000 UTC 5 March (right). Click on images to enlarge.
Figure 3. SPC objective analysis of 500 mb geopotential height, temperature, and wind barbs at 1200 UTC 4 March (left) and 0000 UTC 5 March (right). Click on images to enlarge.
Figure 4. SPC objective analysis of 700 mb geopotential height, temperature, dewpoint, and wind barbs at 1200 UTC 4 March (left) and 0000 UTC 5 March (right). Click on images to enlarge.
Wind shear through a deep layer was expected to translate east across the region in the afternoon as the core of 65 knot winds at 700 mb lifted north and a low level jet of at least 50 knots at 850 mb developed across the southeast (Fig. 5). At the surface, low pressure was centered over middle Tennessee with a strong cold front extending south across Alabama to the western Florida Panhandle at 1200 UTC (Fig. 6). Based on this information, the situation remained favorable for severe weather across the region in the afternoon, but the threat increased over the eastern part of the Carolinas. A limiting factor over the western Carolinas was thought to be a relative lack of instability due to extensive cloud cover (Fig. 7) and a wide band of showers and embedded thunderstorms (Fig. 8) that was moving across the region in the morning ahead of the cold front.
Figure 5. SPC objective analysis of 850 mb geopotential height, temperature, dewpoint, and wind barbs at 1200 UTC 4 March (left) and 0000 UTC 5 March (right). Click on images to enlarge.
Figure 6. Hydrometeorological Prediction Center surface analysis of fronts and pressure valid 1200 UTC 4 March 2008. Click on image to enlarge.
Figure 7. GOES-12 visible satellite image at 1245 UTC 4 March 2008. Click on image to enlarge.
Figure 8. Regional radar reflectivity mosaic centered on GSP at 1257 UTC 4 March 2008. The intensity of precipitation is given by the color scale at the lower left. Click on image to enlarge.
By the late morning, it became apparent that some destabilization was likely where partial clearing occurred between the back edge of the rain band and immediately ahead of the cold front that stretched north to south across eastern Alabama at 1500 UTC. The corridor of developing instability, combined with the strong forcing for large scale uplift and deep layer shear as evidenced by the 1200 UTC upper air observation at Peachtree City, Georgia (FFC), suggested an increased threat of supercells and tornadoes. As such, the Day 1 convective outlook issued at 1609 UTC upgraded northeast Georgia and the western Carolinas to a Moderate Risk. When thunderstorms began to fire along the Georgia - Alabama border near the cold front, the SPC issued Tornado Watch #96 at 1705 UTC for the area generally west of a line from Asheville and Hendersonville to Anderson and Elberton.
3. Pre-Storm Environment
The prefrontal band of precipitation continued to weaken during the early afternoon across the western Carolinas while a line of thunderstorms strengthened over northwest Georgia (Fig. 9) along the cold front at 1800 UTC. A special upper air sounding taken at FFC at 1800 UTC was positioned perfectly to sample the environment ahead of the convective line (Fig. 10). Surface-based convective available potential energy (CAPE) of 1200 J/kg, strong shear in the surface to 3 km layer (almost 60 kt), and strong storm relative helicity (SRH) in the surface to 1 km layer (nearly 200 m2/s2) were all supportive of rotating updrafts. As the line of thunderstorms moved across north Georgia, the environment to the east remained favorable for rotating storms, as noted by the SPC in Mesoscale Discussions issued at 1903 UTC and 1930 UTC, and the updated Day 1 Convective Outlook issued at 1956 UTC. Although the mesoscale analysis showed considerably weaker buoyancy at 2100 UTC (Fig. 11) as the convective line was moving across extreme northeast Georgia, the low level helicity was considerably stronger (Fig. 12). Meanwhile, there was reason to expect further destabilization as a mid level dry intrusion moved in from the west (Fig. 13). For this reason, a new Tornado Watch (#98) was issued by the SPC at 2131 UTC to encompass the threat across the remainder of the GSP county warning area.
Figure 9. As in Figure 8, except for 1755 UTC. The location of the upper air sounding in Figure 10 is indicated by the "+" sign labelled FFC. Click on image to enlarge.
Figure 10. Skew-T, log P diagram and hodograph for upper air observation taken at FFC at 1800 UTC 4 March 2008. A table of severe weather parameters and indices is given at the bottom. Click on image to enlarge.
Figure 11. SPC objective analysis of mixed layer CAPE (contours) and mixed layer Convective Inhibition (CIN, shaded) at 2100 UTC 4 March. Click on image to enlarge.
Figure 12. SPC objective analysis of 0-1 km SRH (contours) and storm motion (barbs) at 2100 UTC 4 March. Click on image to enlarge.
Figure 13. GOES-12 water vapor imagery at 2045 UTC 4 March. The color scale at the bottom indicates the brightness temperature detected by the sensor. Warmer brightness temperatures correspond to drier air at mid levels. Click on image to enlarge.
4. Radar Observations
A broken line of strong to severe thunderstorms moved steadily east across northeast Georgia, the southern mountains of North Carolina, and the western half of the Upstate of South Carolina between 2030 UTC and 2300 UTC. Scattered reports of wind damage were received after the passage of the convective line, but none were outstanding. Most of the wind damage reports appeared to be associated with bowing segments of the line. As the line moved into western Greenville County at 2241 UTC, one particular bowing segment was observed over Travelers Rest, with a weak echo channel extending back across northwest Pickens County (Fig. 14). Although no damage was reported near Travelers Rest, this segment eventually moved east across the Tigerville area and produced wind gusts estimated at 60 mph and penny sized hail. The base velocity image at 2241 UTC showed motion of targets toward the radar in the 40 to 60 kt range on the leading edge of the line, which was probably representative of what was experienced in most places as the line passed.
Figure 14. KGSP base reflectivity on the 0.5 degree scan at 2241 UTC 4 March. The color scale at the bottom right indicates the intensity of the rain. Click on image to enlarge.
Figure 15. KGSP base velocity on the 0.5 degree scan at 2241 UTC 4 March. The color scale at the bottom right indicates the motion toward or away from the radar site, which is labelled KGSP. Red shades represent motion away from the radar and green shades indicate motion toward the radar. Click on image to enlarge.
a. The Greer downburst
The line of thunderstorms moved across Greenville County through 2300 UTC. The KGSP radar detected wind convergence on the leading edge of the line, both in a storm relative and an actual sense, but no significant storm relative cyclonic circulations were detected through the 2256 UTC scans when the radar beam cut across the line (Fig. 16). As the line moved past and into Spartanburg County around 2301 UTC, relatively strong cyclonic shear was observed as the radar beam scanned parallel to the line. However, rotation was minimal again at 2306 UTC as the line raced off to the northeast. The lack of rotation when scanning perpendicular to the line suggested a low potential for a tornado. However, a cross-section of reflectivity taken in the direction of motion showed a large (but shallow) echo overhang (Fig. 17), a common feature among severe wind producing squall lines. The line of thunderstorms produced a peak wind gust of 44 mph as it moved over the Greenville-Spartanburg International Airport. As the strongest part of the line moved northeast of Greer, the KGSP radar detected patches of outbound velocity above 50 kt (58 mph) between Greer and Duncan at 2306 UTC, roughly co-located with the reports of wind damage (Fig. 18). Brief tornadoes are sometimes associated with rapidly moving thunderstorm lines, but in this case a survey of the damage revealed numerous trees blown down in the same direction, which is indicative of a strong straight- line wind.
Figure 16. KGSP storm relative motion 1.8 degree scan at (A) 2251 UTC, (B) 2256 UTC, (C) 2301 UTC, and (D) 2306 UTC. Click on image to enlarge.
Figure 17. KGSP base reflectivity 12.5 degree scan (top) and cross section (bottom) at 2301 UTC. The white line in the top figure denotes the vertical plane of the cross section in the bottom figure. Click on image to enlarge.
Figure 18. As in Fig. 15, except for 2306 UTC. Note the patch of bright red east of Greer indicating outbound velocity of at least 50 kt. Note the small patch of bright green northeast of Greer is not inbound velocity but improperly dealiased data. Click on image to enlarge.
b. The Cornatzer tornado
The storm that produced the Cornatzer tornado strained the limits of predictability. Numerous discrete bowing line segments were observed across the western Piedmont of North Carolina through about 0100 UTC on 5 March, several of which had a legacy of producing wind damage. There was no indication in any of the severe weather reports up until that time that any tornadic activity had occurred. The convection was generally low-topped (radar echo tops from KGSP of 25,000 - 30,000 feet) and moving rapidly east northeast (45 - 55 kt). Based on the history of the storms and no coherent rotation seen on the imagery from either KGSP or the TCLT radar, a Severe Thunderstorm Warning was issued for Davie, Rowan, and Cabarrus counties at 0108 UTC in advance of the convective line (Fig. 19). In keeping with NWS policy, the warning included the phrase "severe thunderstorms can and occasionally do produce tornadoes with little or no advance warning" because of the Tornado Watch that was in effect.
Figure 19. Radar reflectivity mosaic at 0108 UTC 5 March. The yellow polygon outlines the Severe Thunderstorm Warning issued for Davie, Rowan, and Cabarrus counties valid until 0230 UTC. Click on image to enlarge.
The KGSP radar gave little notation that a tornado was imminent or occurring. The height of the 0.5 degree elevation beam over Davie County was approximately 12,000 to 15,000 feet MSL, which was too high to see any features in the bottom half of the storm in the reflectivity field (Fig. 20). Weak to minimal rotation was detected, but not identified as a mesocyclone at the extreme range from the radar (Fig. 21). The KRAX (Raleigh) and KFCX (Blacksburg) radars also did not yield any clues. The TCLT radar was in a much more favorable position to observe the structure of the storm, but it also failed to yield any meaningful clues. Although the height of the beam at 1.0 degrees was between 5,000 and 7,000 feet MSL, no discernable structure was seen in the reflectivity and the maximum value was only 45 dBZ in the cell thought to be responsible for producing the tornado. The storm relative motion was contaminated by range folding in all volume scans leading up to and including the time of the tornado (Fig. 22). The 2.4 degree scan was also largely contaminated, except for the scan at 0115 UTC which showed only minimal rotation. Reflectivity at that level was nondescript.
Figure 20. KGSP base reflectivity on the 0.5 degree scan at 0121 UTC 5 March. Click on image to enlarge.
Figure 21. KGSP storm relative motion on the 0.5 degree scan at 0121 UTC 5 March. Warmer colors represent motion away from the radar and cooler colors indicate motion toward the radar. Click on image to enlarge.
Figure 22. TCLT storm relative motion on the 1.0 degree scan at 0115 UTC 5 March. Warmer colors represent motion away from the radar and cooler colors indicate motion toward the radar. Click on image to enlarge.
The fact that no supercell structure could be seen in the TCLT reflectivity data in low to mid levels of the storm suggested that non-supercell processes were responsible for the tornado.
In the afternoon and evening of 4 March, a line of strong to severe thunderstorms moved east across northeast Georgia, the Upstate of South Carolina, and the Foothills and Piedmont of North Carolina. Numerous reports of wind damage were received as well as a few reports of large hail. The events of 4 March were well anticipated by the NWS with all damage reports occurring within a Tornado Watch and a Warning polygon.
Notable damage occurred northeast of Greer, South Carolina, and near the community of Cornatzer, in Davie County, North Carolina. The radar data did not show any significant rotation in the storm that moved over Greer, but had an echo overhang which is commonly associated with wind damage-producing storms. A patch of outbound velocity greater than 50 kt at 300 ft AGL was detected near the location of the damage. Thus, the radar data supported the conclusion that the damage northeast of Greer was the result of straight-line downburst winds.
Pictures of wind damage northeast of Greer, South Carolina, on 4 March 2008. Click on image to enlarge.
The radar data was not able to support the conclusion that a tornado produced the damage near Cornatzer. The beam from the KGSP radar passed through the Cornatzer storm at least 12,000 feet AGL, so the low level structure could not be observed. The TCLT radar did not show any significant rotation at the 1.0 degree elevation scan, but the 0.2 degree data was missing. The lack of rotation suggests the tornado was produced by non-supercell processes. Such tornadoes are very difficult to detect at long range from the radar.
Neil Dixon surveyed the wind damage northeast of Greer. Vince DiCarlo performed the damage survey for the Cornatzer tornado. The upper air analyses, soundings, mesoscale analyses, and severe weather report plot were obtained from the Storm Prediction Center. The satellite imagery and radar mosaics were obtained from the University Corporation for Atmospheric Research. The surface analyses were obtained from the Hydrometeorological Prediction Center. The KGSP radar images were created using the NCDC Java NEXRAD viewer. The reflectivity cross section was created using GR2Analyst. The radar mosaic and warning polygon image was obtained from the Iowa Environmental Mesonet.