A tornado struck near Ellenboro, North Carolina, on 11 January 2012. Image taken by Dustin Wiser and reprinted from the NWS Facebook page.

Author's Note: The following report has not been subjected to the scientific peer review process.

1.  Introduction

During the early part of the evening on Wednesday, 11 January 2012, 
a series of severe thunderstorms produced three tornadoes across the 
eastern part of Rutherford, Burke, and Caldwell counties in North 
Carolina.  The first tornado touched down in Rutherford County 
northwest of Ellenboro at approximately 522 pm EST (Eastern Standard 
Time), or 2222 UTC (Universal Time Coordinated).  The tornado briefly 
made contact with the ground at two locations before leaving a 2.3 mile 
long path of destruction.  Six to ten homes were damaged and some mobile 
homes were completely destroyed.  Ten people were injured as a result.  
The damage was estimated at EF-2 intensity on the Enhanced Fujita Scale.  
The next tornado touched down at 604 pm EST (2304 UTC) about five miles 
south southwest of Icard in Burke County.  After a short period of 
intermittent contact with the ground, a more continuous trail of damage 
was produced along a 3.6 mile long path.  At least 75 homes were damaged 
near Icard, with several mobile homes totally destroyed, and eight people 
were injured.  The damage was also estimated at EF-2 intensity.  The 
final tornado briefly struck the northern shore of Lake Hickory in 
extreme southeast Caldwell County at 618 pm EST (2318 UTC).  The damage 
was limited to the area near the Lake Hickory Marina and was rated at 
EF-0 intensity.  No other reports of severe weather were received across 
the entire United States on that day (Fig. 1).

[Note:  Times in this report from this point forward are referenced to 
UTC, which is Eastern Standard Time plus five hours.]

Figure 1.  Preliminary severe wind, large hail, and tornado reports 
received at the Storm Prediction Center for the 24 hour period ending at 
1200 UTC on 12 January 2012.  Click on image to enlarge.

The events of 11 January were noteworthy because no tornadoes had been 
verified in the months of December, January, or February over the North 
Carolina foothills within the NWS Greenville-Spartanburg (GSP) County 
Warning Area (CWA) since detailed records began in 1950 (Fig. 2).  The 
lack of tornadic activity over the North Carolina foothills should come 
as no surprise as this area often experiences cold air damming with the 
passage of low pressure in the cool season.  Historically speaking, the 
only other tornado of record over the North Carolina foothills in the 
winter months occurred on 21 January 1959 over Surry County.  The last 
significant tornado in western North Carolina during "Meteorological 
Winter" occurred in Cleveland County on 10 February 1990 (an F2).  Thus, 
it would be fair to consider the tornadoes on 11 January 2012 to be a 
rare event.

Figure 2.  Tornadoes confirmed across western North Carolina during the 
months of December, January, and February for the period from 1950 to 2011.  
Click on image to enlarge.

2.  Synoptic Features and Pre-Storm Environment

The pre-storm environment on 11 January 2012 was characterized by deep 
forcing for upward vertical motion and high values of wind shear, but 
low values of Convective Available Potential Energy (CAPE).  In the 
morning of 11 January, an upper low at 500 mb centered over Mississippi 
at 1200 UTC (Fig. 3) was forecast to open up across the Cumberland 
Plateau and Tennessee Valley during the day, while a large area of upper 
difluence at 300 mb was expected to spread across the region ahead of 
the low.  Strong wind shear was forecast to translate across the western 
Carolinas during the middle of the day as the mid-level jet of 50-70 kt 
wrapped around the southern extent of the upper low and a southerly 850 mb 
jet of at least 50 kt (Fig. 4) moved overhead.  At the surface, a warm 
front was oriented west to east across central Georgia and coastal South 
Carolina (Fig. 5).  The warm front intersected a cold front over east 
central Alabama, from which point a cold front trailed down to the Florida 
Panhandle and an occluded front extended north to a center of low pressure 
over southwest Kentucky.  Although the upper air observation at Peachtree 
City, Georgia (FFC, Fig. 6), at 1200 UTC was taken in the more stable air 
mass to the north of the warm front, the modified convective parameters 
compared favorably with values computed in the warm sector later in the day.

Figure 3.  Objective analysis of 500 mb geopotential height (black 
contours), temperature (red dashed contours), and wind barbs at 1200 UTC 
on 11 January 2012.  Click on image to enlarge.

Figure 4.  Objective analysis of 850 mb geopotential height (black 
contours), temperature (red and blue dashed contours), dewpoint (green 
contours), and wind barbs at 1200 UTC on 11 January 2012.  Click on image 
to enlarge.

Figure 5.  Surface fronts and pressure analysis from the Hydrometeorological 
Prediction Center at 1200 UTC on 11 January 2012.  Click on image to enlarge.

Figure 6.  Skew-T, log P, diagram of the upper air sounding taken at FFC
(Peachtree City, Georgia) at 1200 UTC on 11 January.  The temperature 
sounding is shown in red and the dewpoint sounding is shown in green.  
A hodograph is shown in the upper right.  A table showing convective 
parameters is provided at the bottom.  Click on image to enlarge.

The best combination of shear and instability were forecast to occur along 
the Carolina coast, where a Slight Risk was included in the Day 1 Convective 
Outlooks issued by the Storm Prediction Center (SPC) at 1630 UTC and 2000 UTC.  
Convection the previous night and during the morning left behind a cool pool 
of relatively stable air across the western Piedmont and foothills of North 
Carolina.  A relative lack of instability was thought to be a limiting factor 
across the GSP forecast area as it was still unclear how far inland the warm 
front would penetrate during the afternoon.

In fact, the warm front lifted to the vicinity of the South Carolina – 
North Carolina border by 2100 UTC with a triple-point located over the 
North Carolina Mountains (Fig. 7).  Within a narrow warm sector over 
northern South Carolina, dew point temperatures surged into the mid-50s 
with a southerly wind of 10 kts.  The movement of the upper low to eastern 
Tennessee, seen on water vapor imagery at 2145 UTC (Fig. 8), and the 
northward movement of the warm front into North Carolina, had a pronounced 
effect on the environment ahead of shallow convection moving across 
Upstate South Carolina.  Differential positive vorticity advection at mid 
and upper levels combined with strong low level convergence along the 
South and North Carolina border in the vicinity of the surface triple-
point (Fig. 9) to produce strong upward vertical motion.  Colder air 
aloft associated with the upper low contributed to a steeper lapse rate 
in the 700-500 mb layer.  The northwestern extent of the effective bulk 
shear greater than 40 kt and the effective Storm Relative Helicity (SRH) 
of at least 200 m2s-2 reached into northern South Carolina and the southern 
Foothills of North Carolina, indicating that storms moving into that region 
would have a favorable environment to acquire rotation.  Superimposed on 
the strong shear, most unstable CAPE (MUCAPE) reached 500 J kg-1 within 
the warm sector.  The most unstable lifted index (MULI) reached its lowest 
negative number over the same region, while the core of the mid-level jet 
streak moved overhead.

Click here to view a 25 frame Java loop of GOES-13 water vapor satellite 
imagery from 0245 UTC on 11 January to 0245 UTC on 12 January.

Figure 7.  Surface fronts and pressure analysis from the Hydrometeorological 
Prediction Center at 2100 UTC on 11 January 2012.  Click on image to enlarge.

Figure 8.  GOES-13 water vapor imagery at 2145 UTC on 11 January 2012.  
Brightness temperatures are given by the color table at the bottom.  Click 
on image to enlarge.

Figure 9.  SPC objective mesoanalysis of (a) boundary layer wind (kt; barbs), 
streamlines (black vectors), and convergence (red contours), (b) 700-500 mb 
lapse rate (oC km-1; green contours, (c) effective bulk shear (kt; blue 
contours, (d) effective SRH (m2s-2; blue contours, inflow base (m AGL, 
color fill) , and storm motion (kt, barbs), (e) MUCAPE (J kg-1; orange 
contours), MUCIN, and effective bulk shear (kt; barbs), and (f) 850 and 
500 mb wind crossover (kt; barbs) and MULI (index; yellow contours) at 
2200 UTC on 11 January.  Click on image to enlarge.

3.  Radar observations

Thunderstorms developed close to the triple point over extreme northeast 
Georgia between 1800 UTC and 1900 UTC, then moved east and northeast 
across the northwestern part of Upstate South Carolina through 2100 UTC.  
Penny-sized hail was reported as a cell moved over Six Mile (Pickens County), 
South Carolina, at 2025 UTC.

Click here to view a 35 frame Java loop of KGSP Composite Reflectivity from 
1901 UTC to 2059 UTC on 11 January.

A persistent thunderstorm cell moved northeast across the middle of 
Greenville County, South Carolina, between 2100 UTC and 2130 UTC.  This 
storm gained strength for the next ten minutes.  The KGSP radar showed 
the vertical extent of the 55 dBZ reflectivity core reaching a maximum 
of 20,500 feet above Mean Sea Level near Lake Bowen at 2140 UTC (Fig. 10).  
The apparent echo overhang between Inman and Lake Bowen was probably the 
result of a new updraft developing on the southeastern (inflow) flank of 
the storm, and not a tilted updraft in the same storm.  The vertically-
integrated liquid product also peaked at 42 kg m-2 at that time.  A 
Tornado Vortex Signature (TVS) alarm was generated at 2144 UTC, but was 
thought to be false because of spurious low level radial velocity data.  
Penny-sized hail fell across the northern part of Spartanburg County 
between Fingerville and the state line.  A Severe Thunderstorm Warning 
was issued for this storm at 2155 UTC, including eastern Rutherford 
County, North Carolina, valid until 2300 UTC (Fig. 11).

Click here to view a 16 frame Java loop of KGSP Composite Reflectivity from 
2059 UTC to 2200 UTC on 11 January.

Figure 10.  Base reflectivity (dBZ) from the KGSP radar on the (a) 0.5 degree,
(b) 5.1 degree, and (c) 12.5 degree scans, and (d) vertically integrated 
liquid (kg m-2) at 2140 UTC.  Click on image to enlarge.

Figure 11. KGSP composite reflectivity at 2152 UTC with polygon for Severe 
Thunderstorm Warning #0001 (outlined in yellow) issued at 2155 UTC and 
valid until 2300 UTC.  Click on image to enlarge.

After 2140 UTC, the storm moved down-radial from the KGSP radar.  Between 
2144 UTC and 2204 UTC, the base velocity on the lowest four elevation 
scans on the KGSP radar showed evidence of developing rear to front flow 
across the upshear (southern) flank of the storm as it crossed southeastern 
Polk County, North Carolina, and into southwestern Rutherford County.  An 
expanding area of reflectors moving away from the radar at 45-50 kt was 
noted behind a bow in the higher reflectivity, while the targets ahead of 
the higher reflectivity were only moving away at 15-25 kt (Fig. 12).  By 
2204 UTC, the radar showed evidence of the rear to front flow moving 
around the eastern flank of the storm.  The leading edge of the stronger 
outbound velocity coincided with the southern extent of the reflectivity 
wrapping around the southeastern side of the storm at that time.  After 
2207 UTC, this began to manifest itself as broad rotation on the inflow 
side of the storm.

Figure 12.  Base reflectivity and base velocity from the KGSP radar on 
the (a, b) 1.8 degree scan at 2144 UTC, (c, d) 1.3 degree scan at 2156 UTC, 
and (e, f) 0.9 degree scan at 2204 UTC.  An alternate color scale was used 
for the base velocity which highlights outbound winds greater than 45 kt 
in pink.  The white arrow in (a), (c), and (e) is used as a tracer for 
the leading edge of the higher reflectivity.  The small black ‘x’ in (b), 
(d), and (f) is at the same height relative to MSL in each image. Click 
on image to enlarge.

The storm moved into the range of the Terminal Doppler Weather Radar (TDWR) 
located north of the Charlotte airport (the TCLT radar) around 2213 UTC.  
Rotational velocity and shear remained relatively weak until 2217 UTC and 
steadily increased thereafter on both radars (Figs. 13 and 14).  Although 
an inbound-outbound velocity couplet tightened on the KGSP radar through 
2224 UTC, values of rotational shear remained below values normally 
associated with tornadic activity.  The TCLT radar showed a more rapid 
increase in rotational shear to the point where it exceeded 0.02 s-1 on 
the 0.2 degree scan at 2221 UTC.  Other than some modest echo overhang as 
the storm was developing, reflectivity clues that a tornado was imminent 
were limited to evidence of an inflow notch on the southeastern flank of 
the storm (Fig. 15).  A gate-to-gate velocity couplet had formed on the 
TDWR storm relative velocity product, but rotation kept a broad appearance 
on the KGSP radar.  Tornadogenesis occurred at 2222 UTC.  Neither of the 
radars detected a mesocyclone or TVS during this phase of the event.  
Rotational shear remained on the transition between the "probable" and 
"likely" areas of the rotational shear nomogram (Falk and Parker 1988) 
through about 2232 UTC, although the Ellenboro tornado lifted around 
2227 UTC.  A Tornado Warning (#0001) was issued for northeastern Rutherford, 
northwestern Cleveland, northwestern Lincoln, southeastern Burke, and 
southwestern Catawba counties at 2235 UTC, valid until 2315 UTC (Fig. 16).

Click here to view a 13 frame Java loop of TCLT 1.0 Degree base reflectivity 
and storm relative motion from 2213 UTC to 2325 UTC on 11 January.

Figure 13. Maximum rotational velocity (kt) and rotational shear (10-3 s-1) 
on the lowest three scans from the KGSP radar from 2207 UTC to 2331 UTC on 
11 January 2012.  The yellow bar indicates the time of the Ellenboro tornado.  
The pink bars represent the times of the Icard and Lake Hickory tornadoes.  
Click on image to enlarge.

Figure 14. Maximum rotational velocity (kt) and rotational shear (10-3 s-1) 
for the two lowest scans from the TCLT radar from 2213 UTC to 2331 UTC on
11 January.  The 0.2 degree data was not available from 2301 UTC to 2316 UTC.  
The time of the Ellendale tornado is shown by the yellow bar and the times 
for the Icard and Lake Hickory tornadoes are shown by the pink bars.  Click 
on image to enlarge.

Figure 15.  Base reflectivity (a) and Storm Relative Motion (b) on the 
0.5 degree scan from the KGSP radar and base reflectivity (c) and storm 
relative motion (d) on the 0.2 degree scan from the TCLT radar at 2221 UTC 
on 11 January.  The KGSP radar is located off the bottom left corner of 
the top two images, while the TCLT radar is located off the right edge of 
the bottom two images.  The pink arrow points to the weak inflow notch on 
the eastern flank of the storm.   Click on image to enlarge.

Figure 16. Composite reflectivity from the KGSP radar at 2233 UTC with the 
polygon outline (in red) for Tornado Warning #0001, issued at 2235 UTC 
and valid until 2315 UTC.  Click on image to enlarge.

A new, detached updraft formed around 2227 UTC along the Rutherford and 
Cleveland County line, off the eastern flank of the old storm (Fig. 17).  
The new cell quickly acquired low level rotation after 2237 UTC.  Another 
surge in rotational shear was analyzed in the 0.2 degree scans from TCLT 
after 2241 UTC (Fig. 14) as the new cell moved past Casar in northern 
Cleveland County.  Values climbed back into the "tornado probable" range 
after 2245 UTC and peaked at 0.04 s-1 at 2252 UTC over southeastern Burke 
County.  Although a weather watch was contemplated by the SPC, one was
not issued because of the small extent and downward trend of weak 
instability at that time of day.

Figure 17.  KGSP radar scans at 2237 UTC of base reflectivity (a) and 
storm relative motion (b) at 0.5 degrees, and base reflectivity (c) and 
storm relative motion (d) at 0.9 degrees.  The "E" corresponds to the 
remnant of the Ellenboro storm, while the "I" indicates the development 
of the storm that would eventually produce the tornado near Icard.  Click 
on image to enlarge.

A weak echo channel developed behind the higher reflectivity associated 
with the main cell over eastern Burke County, perhaps associated with a 
rear flank downdraft (Fig. 18).  A cusp formed in the reflectivity where 
the weak echo channel began to wrap around the southeast flank of the 
storm.  This cusp was co-located with a prominent velocity couplet on the 
0.2 degree scan from TCLT at 2300 UTC.  Rotational shear on the lowest two 
elevation angles remained strong through 2304 UTC, when the second tornado 
touched down in eastern Burke County to the south southwest of Icard.  A 
new Tornado Warning (#0002) was issued at 2305 UTC to extend the warning 
area into Alexander County and more of Caldwell County (Fig. 19).  While 
the Icard tornado was on the ground, the weak echo channel continued to 
wrap around the eastern side of the main part of the cell and forced a 
split between the main cell and an arc of higher reflectivity that 
extended to the south (Fig. 20).  The resemblance to the "Broken-S" 
reflectivity evolution (e.g. Lane and Moore 2006) was noteworthy.  By the 
time the Icard tornado lifted at 2310 UTC near Hildebran, the fracture 
between the main cell and the flanking arc of reflectivity was complete.

Figure 18.  TCLT radar base reflectivity (a) and storm relative motion (b) 
at 0.2 degrees on the 2300 UTC scan, and base reflectivity (c) and storm 
relative motion (d) at 1.0 degrees on the 2301 UTC scan.  The white curved 
arrow denotes the weak echo channel.  The TCLT radar is situated off the 
bottom right corner of each graphic.  Click on image to enlarge.

Figure 19. KGSP composite reflectivity at 2304 UTC with outline of polygon 
for Tornado Warning #0002 (in red), issued at 2305 UTC and valid until 
2345 UTC on 11 January.  Click on image to enlarge.

Figure 20.  KGSP base reflectivity on the 0.5 degree scan at (a) 2301 UTC, 
(b) 2307 UTC, (c) 2313 UTC, and (d) 2319 UTC.  The curved white arrow 
denotes the weak echo channel wrapping around the southern and eastern 
flank of the main reflectivity core moving from eastern Burke into 
southeastern Caldwell counties.  Click on image to enlarge.

One final increase in rotational shear was analyzed on the 1.0 degree 
scan between 2313 and 2319 UTC as the main thunderstorm moved into extreme 
southeast Caldwell County.  The final tornado touched down briefly on the 
north shore of Lake Hickory at the Lake Hickory Marina at about 2318 UTC.  
No other severe weather was reported. 

4.  Summary 

In spite of a lack of reflectivity features commonly associated with 
miniature supercells, the storms on 11 January met the most basic 
definition of a supercell.  The environment had sufficient shear, 
helicity, and buoyancy to support supercell development.  While the 
cell had some low level characteristics of a mini-supercell, such as 
a low level pendant on the upshear flank, observed rotation was 
categorized as weak.  The rotation failed to trigger a mesocyclone 
or TVS alarm.  The WSR-88D at Columbia did not show any significant 
rotation, but was too far away to scan below approximately 8000 feet 
above ground level.  The Ellenboro storm appeared to have a quasi-steady 
updraft for over an hour before it acquired significant rotation.

The supercell that produced the Icard tornado was not the same cell as 
the one that produced the Ellenboro tornado.  In addition to the 
reflectivity evidence of a new and separate updraft core developing to 
the east of the cell that produced the Ellenboro tornado, the rotation 
track map (Fig. 21) obtained from the National Severe Storms Laboratory 
clearly showed two corridors of rotation across eastern Rutherford and 
northwest Cleveland counties.

Figure 21.  Rotation tracks for 11 January 2012 obtained from the KGSP and 
TCLT radars.  Image created by the National Severe Storms Laboratory.  
Click on image to enlarge.

Low-level rotation developed very quickly in the Ellenboro storm about 
five minutes before tornadogenesis.  There was a two minute lag in the 
increase of rotational velocity between the 0.2 degree scan and the 
1.0 degree scan, suggesting that rotation developed upward in the storm.  
Rotational shear jumped from ambient levels at 2217 UTC to a level above 
what has been observed in tornadic storms at 2221 UTC.  There were no 
hooks or appendages in the lowest reflectivity scans.  Thus, the issuance 
of a Tornado Warning was not supported by a preponderance of the evidence 
until about one minute before tornadogenesis.

Pictures of the Ellenboro Tornado Damage

Pictures of the Icard Tornado Damage

Pictures of the Lake Hickory Tornado Damage

References

Falk, K., and W. Parker, 1998:  Rotational shear nomogram for tornadoes.
     Preprints, 19th Conf. on Severe Local Storms, Minneapolis, MN, Amer. 
     Meteor. Soc., 733-735.

Lane, J. D., and P. D. Moore, 2006:  Observations of a non-supercell tornadic
     thunderstorm from a Terminal Doppler Weather Radar.  Preprints, 23rd Conf. 
     on Severe Local Storms, St. Louis, MO, Amer. Meteor. Soc.

Acknowledgements

The upper air maps, soundings, mesoscale analysis, and storm report 
plots were obtained from the Storm Prediction Center.  The surface
analysis was obtained from the Hydrometeorological Prediction Center.
Satellite imagery and surface observation plots were provided by
the University Corporation for Atmospheric Research.  The radar
images with warning polygon graphics were obtained from the Iowa
Environmental Mesonet web page, maintained by the Iowa State 
University Department of Agronomy.  Tornado track maps were created
using Google Earth and Google Maps.  The damage survey for the 
Ellenboro tornado was conducted by Tony Sturey and Larry Gabric.
The damage surveys for the Icard tornado and Lake Hickory tornado were
conducted by Harry Gerapetritis and John Tomko.  The map of rotation
tracks was obtained from the National Severe Storms Laboratory.