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.

1.  Introduction

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.

Click here to view a list of local storm reports for this event

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.

Click here to view a 20 frame Java loop of GOES-12 water vapor 
satellite imagery from 0845 UTC 4 March to 0345 UTC 5 March.

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.

Click here to view a 7 frame Java loop of GOES-12 visibile satellite 
imagery from 1245 UTC to 1845 UTC.

Click here to view a 10 frame Java loop of the regional reflectivity 
mosaic centered on GSP from 0958 UTC to 1856 UTC.

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.

Click here to view a 33 frame Java loop of 0.5 degree base 
reflectivity from the KGSP radar from 2031 UTC to 2301 UTC.

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.

Click here to view a 14 frame Java loop of base reflectivity from 
the 2301 UTC volume scan from the KGSP radar.

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.

Click here to view an 8 frame Java loop of 0.5 degree base 
reflectivity from the KGSP radar from 0102 UTC to 0135 UTC.

Click here to view an 8 frame Java loop of 0.5 degree storm 
relative motion from the KGSP radar from 0102 UTC to 0135 UTC.

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.  

5.  Summary 

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.

Acknowledgements

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.