Several trees were blown down by the tornado that moved across northern
Gaston County on 26 May 2006.  The tornado also damaged the roof of the 
structure seen in the distance.

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

1.  Introduction

The active springtime continued across the Carolinas through late May 
with another severe weather outbreak on Friday, 26 May 2006 (Fig. 1).  
The National Weather Service (NWS) Weather Forecast Office (WFO) in 
Greer, South Carolina (GSP), issued 25 Severe Thunderstorm Warnings, one 
Tornado Warning, and one Flash Flood Warning on that day.  In particular, 
a series of strong to severe multicell thunderstorms moved across the 
southern Piedmont of North Carolina during the late afternoon and early 
evening.  The fourth and last in the series of severe storms produced a 
tornado across the extreme northwestern part of Gaston County in a rural 
area between Cherryville and High Shoals.  The tornado touched down about 
four miles east northeast of Cherryville, near the intersection of 
Hepzibeth Road and Saint Marks Church Road, at approximately 614 pm EDT 
(2214 UTC).  [All times in this document are referred to in Universal 
Time Coordinated (UTC), which is Eastern Daylight Time plus four hours.]  
The tornado produced intermittent damage along a path 2.2 miles long and 
40 yards wide that ended near the intersection of Landers Chapel Road 
and Gaston-Webb Chapel Road.  A damage survey rated the tornado at 
F1 intensity on the Fujita Scale.  A Tornado Warning was issued by 
WFO GSP at 2209 UTC for northern Gaston County, providing a lead time 
of 5 minutes.

(Click here to view a summary of severe weather reports for 26 May 2006.

Figure 1.  Wind damage, large hail, and tornado reports for 26 May 2006.  
Click on image to enlarge.  

The events of 26 May are interesting because the initial threat for 
tornadoes was perceived to be low.  Environmental clues that morning
suggested the most likely threat to be damaging straight line winds.  
After the first multicell thunderstorm formed in the middle part of
the afternoon, preferential development of new cells to the west of 
the initial cell acted to strengthen the boundary between the 
environmental air and the rain-cooled air at low levels, as the new 
cells moved to the east.  The boundary further acted to strengthen 
the low level vorticity ingested by successive cells, culminating 
in the development of a low-level mesocyclone and the production of 
the weak tornado over extreme northwestern Gaston County.

2.  Synoptic Features and Pre-Storm Environment

A severe weather episode triggered in part by the passage of a strong 
short wave was a recurring theme during the Spring of 2006 and the 
events of 26 May were no exception.  The upper air analysis from the 
Storm Prediction Center (SPC) at 1200 UTC on that day showed the axis 
of a short wave trough at 500 mb over the Cumberland Plateau and 
Tennessee Valley (Fig. 2).  A belt of stronger winds at that level, 
shown by the wind barbs in the figure, was expected to provide 
favorable wind shear as it translated across the Carolinas later in 
the day with the passage of the upper trough.  The surface analysis 
from the Hydrometeorological Prediction Center (HPC) at 1200 UTC 
showed a cold front pushing east across the Ohio Valley and a lee 
trough extending down across the Piedmont of North Carolina (Fig. 3).  
Forecasters at GSP expected the passage of the short wave, combined 
with the lee trough and outflow from a decaying mesoscale convective 
system over northeast Tennessee, to trigger a round of severe 
thunderstorms in the afternoon given sufficient instability, which was 
highlighted in the early morning Area Forecast Discussion and Severe 
Weather Outlook.  

Figure 2.  SPC objective analysis of 500 mb geopotential height, 
temperature, and wind at 1200 UTC 26 May.  Click on image to enlarge.

Figure 3.  HPC surface pressure and fronts analysis at 1200 UTC 26 May.  
The lee trough over North Carolina is shown by a dashed orange line.  
Click on image to enlarge.

In spite of its closer proximity to the southern Piedmont, the upper 
air sounding at Greensboro, North Carolina (GSO), at 1200 UTC was not 
particularly supportive of severe weather, probably due to contamination 
from the remnants of earlier storms that moved out of northeast Tennessee 
and across the Piedmont Triad during the early morning hours.  Instead, 
the sounding at Peachtree City, Georgia (FFC), was more indicative of the 
expected environment across the western Carolinas (Fig. 4).  The modified 
sounding yielded a surface-based convective available potential energy 
(CAPE, a measure of the potential energy of rising air parcels related to 
buoyancy) of greater than 2000 J/kg (a moderate amount), while shear of 
21 kt in the surface to 6 km layer, though not strong, was expected to be 
sufficient to provide for some organization of storms.  The favorable
environment prompted the Storm Prediction Center (SPC) forecast a Slight 
Risk of severe thunderstorms for all of the Carolinas on the Convective 
Outlook issued at 1300 UTC.  At the time, the primary threat for severe
weather was expected to be damaging wind and large hail.

Figure 4.  Skew-T log P diagram (upper left) and hodograph (upper right) 
for upper air sounding at FFC at 1200 UTC 26 May. The tables at the bottom 
summarize several objective parameters used by the SPC to determine severe 
weather potential.  Click on image to enlarge.

Visible imagery from the GOES-12 satellite at 1245 UTC suggested the 
presence of a boundary moving east across the Piedmont, manifested as 
the sharp edge between the clouds over the eastern part of the Piedmont 
and the clear sky over the Midlands of South Carolina and Sandhills 
of North Carolina (Fig. 5).  The passage of the boundary across the 
western Piedmont earlier in the morning, followed by debris clouds 
from overnight convection, effectively delayed the destabilization 
of the air mass across the western Carolinas during the late morning.  
A Mesoscale Discussion issued by the SPC at 1542 UTC highlighted the 
cloudiness over the mountains as a mitigating factor in convective 
development through the middle part of the day, but mentioned the 
possibility of rotation in the stronger updrafts.  In spite of the
slow start, the potential for severe thunderstorms still existed later 
in the day over the Carolinas, as discussed in the updated Convective 
Outlook issued by the SPC at 1630 UTC.

Click here to view a 13 frame java loop of GOES-12 visible satellite imagery.

Figure 5.  Visible satellite image from GOES-12 at 1245 UTC 26 May.

The debris clouds moved east of the Foothills by 1745 UTC which allowed
for the boundary layer to warm quickly.  The 1800 UTC surface analysis 
continued to show the lee trough over the Piedmont, while a closer look 
at surface observations showed a pool of slightly higher dewpoints in 
the mid-60s from Anderson and Greenwood northeast to Charlotte (Fig. 6).  
By 1800 UTC, the last of the convective inhibition (CIN, a measure of 
resistance to convection at low levels) had been eliminated by boundary 
layer heating, which raised the surface-based CAPE above 1500 J/kg 
east of the mountains (Fig. 7).  Most signs continued to support the
conclusion that damaging wind gusts were the primary threat.  Water 
vapor imagery from GOES-12 at 1745 UTC showed a surge of mid-level dry 
air across the Tennessee Valley and southern Applachians (Fig. 8).  
An intermediate upper air sounding taken at FFC at 1800 UTC showed the 
mid-level drying along with moderate instability and a deep sub-cloud 
layer (the layer from the surface up to the level of free convection at 
1647 m above ground level). The SPC analysis of downdraft CAPE showed a 
large area of greater than 1000 J/kg across the Carolinas (Fig. 9).  
Downdraft CAPE over 1000 J/kg is considered significant to the production 
of damaging wind gusts.  The perceived tornado threat was comparatively 
low owing to weak values of storm relative helicity (SRH, in both the
surface - 1 km and surface - 3 km layers) and high lifting condensation 
level. 

Figure 6.  Surface observations plot at 1743 UTC 26 May.  The traditional 
station model is used.  Click on image to enlarge.

Figure 7.  SPC objective analysis of surface based CAPE and CIN for 
1800 UTC 26 May.  Click on image to enlarge.

Figure 8.  Water vapor image from GOES-12 at 1745 UTC 26 May. 
Click on image to view a 5 frame Java loop of water vapor imagery.

Figure 9.  SPC objective analysis of DCAPE at 1800 UTC 26 May.  
Click on image to enlarge.

Deep convection initiated just northeast of Asheville at 1745 UTC,
perhaps due to enhanced instability on the southern edge of the Black
Mountains.  The initial shower moved off the Blue Ridge after 1800 UTC, 
developing into a thunderstorm as it moved over southern McDowell 
County.  The first Severe Thunderstorm Warning was issued for northeast 
Rutherford County and northern Cleveland County at 1906 UTC as the 
thunderstorm gained strength.  The SPC acknowledged the threat for 
more severe thunderstorms along the lee trough in the updated Day 1 
Convective Outlook issued at 1943 UTC.  More strong thunderstorms 
developing over the western part of Upstate South Carolina led the SPC 
to surmise correctly that an outbreak was underway, and a Severe 
Thunderstorm Watch was issued at 1955 UTC for nearly all of western 
North Carolina, western South Carolina, and extreme northeast Georgia.

3.  Radar observations of the Southern Piedmont Storms

The initial severe thunderstorm developed into a multicell cluster of
storms (Cell 1) as it moved east across the northern part of Cleveland 
County through 1930 UTC (Fig. 10), producing large hail near Lawndale.  
Additional warnings were issued for southern Lincoln County at 1944 UTC 
and northern Gaston County at 1957 UTC as a new multicell cluster of 
thunderstorms (Cell 2) developed to the west (upshear) of the first 
cluster (Fig. 11).  Wind damage and one-half inch diameter hail were 
reported in Cherryville (northwest Gaston County) at 2000 UTC with the 
passage of the first two multicell clusters.  By that time, a third 
multicell (Cell 3) developed near Lawndale in northern Cleveland County 
on the western (upshear) flank of the original complex of multicells
(Fig. 12).  Movement and propagation of the multicell clusters continued 
along the Lincoln - Gaston county line to the area near the southern end 
of Lake Norman, prompting the issuance of a Severe Thunderstorm Warning 
for northern Mecklenburg County at 2010 UTC.  

Click here to view a 25 frame java loop of 0.5 degree reflectivity 
from the KGSP radar from 1901 UTC to 2103 UTC.

Figure 10.  Radar reflectivity on 0.5 degree scan from the KGSP WSR88-D 
at 1930 UTC.  The radar is located off the bottom left corner of the
image.  The TCLT radar location is shown in northern Mecklenburg County.  
Multicell clusters are annotated.  Click on image to enlarge.

Figure 11.  As in Figure 10, except for 1945 UTC.  Click on image to enlarge.

Figure 12.  As in Figure 10, except for 2000 UTC.  Click on image to enlarge.

Multicell 3 followed the same track as the previous multicell clusters, 
eventually producing large hail across the area from Dellview in extreme 
northwestern Gaston County to Crouse in far southwestern Lincoln County 
between 2035 UTC and 2045 UTC.  Around that time, a fourth and final 
multicell (Cell 4) developed once again over northern Cleveland County 
(Fig. 13), resulting in another Severe Thunderstorm Warning for southern 
Lincoln County at 2048 UTC.  Multicell 4 quickly grew to severe intensity 
as it moved east through 2113 UTC (Fig. 14), prompting another Severe 
Thunderstorm Warning for northern Gaston County at 2117 UTC.

Click here to view a 33 frame java loop of 0.5 degree reflectivity 
from the KGSP radar from 2049 UTC to 2326 UTC.

Figure 13.  As in Fig. 10, except for 2049 UTC.  Click on image to enlarge.

Figure 14.  As in Fig. 10, except for 2113 UTC.  Click on image to enlarge.

This final multicell complex was persistent with frequent new updraft 
development on the upshear flank.  The repeated movement of cells 
prompted the issuance of a Flash Flood Warning for northern Gaston 
County at 2140 UTC.  Additionally, the Severe Thunderstorm Warning for 
Lincoln County was reissued for a third time at 2145 UTC.  This cell 
quickly acquired severe characteristics as it moved north of Cherryville 
around 2202 UTC (Fig. 15).  

Click here to view a 15 frame java loop of 0.5 degree reflectivity 
from the KGSP radar from 2049 UTC to 2326 UTC.

Figure 15.  Radar reflectivity at 0.5 degree scan from the KGSP radar at 
2202 UTC.  Click on image to enlarge.

Rotation within the main updraft of multicell 4 developed quickly 
at 2202 UTC, resulting in both a radar-defined and operator-defined 
mesocyclone (not shown).  The lowest level base velocity image from 
the KGSP radar showed evidence of strong rotation on the 2207 UTC 
scan (Fig. 16), indicated by the red-green couplet just east of 
Cherryville. 

Figure 16.  Doppler velocity at 0.5 degree scan from the KGSP radar 
at 2207 UTC.  The red colors show motion of the target away from the 
radar, while the green colors show motion of the target toward the radar.  
Click on image to enlarge.

The rapid spin-up of the mesocyclone and the low level rotation 
signature, combined with corroborating evidence from the Terminal 
Doppler Weather Radar located north of the Charlotte - Douglas 
Airport, led to the issuance of a Tornado Warning for northern
Gaston County at 2209 UTC.  Subsequent radar images showed low level 
rotation weakening as quickly as it had developed as the storm moved 
across northern Gaston County.

Click here to view a 15 frame java loop of 0.5 degree base velocity 
from the KGSP radar from 2049 UTC to 2326 UTC.

4.  Discussion 

The initial convective cell appeared to have strengthened in a zone 
of low level convergence to the lee of the mountains, perhaps as a 
result of confluence in the westerly flow downwind of the mountain
barrier.  The convergence zone was manifested by an east-to-west line 
of enhanced cumulus seen on the visible satellite imagery at 2000 UTC,
stretching from northern Cleveland County, across northern Rutherford 
County, to northern Buncombe County, in the wake of the initial 
multicell (Fig. 17).  The surface observations at that time clearly 
show the convergence in the flow between the stronger southwest winds 
at Rutherfordton (KFQD) and Shelby (KEHO), and the weaker west-southwest 
winds at Hickory (KHKY).

Figure 17.  Visible satellite imagery from GOES-12 at 2000 UTC.  Surface 
observations are indicated with the traditional station model.  Note the 
satellite presentation of the expanding cloud anvil associated with 
multicell cluster (cell 1) in vicinity of KIPJ.  Also note the line of 
enhanced cumulus extending westward from cell 3 back to the north of KAVL.  
Click on image to enlarge.

New cells appeared to develop preferentially over northern Cleveland 
County on the upshear (in this case, upwind) side of the multicell 
storms over Lincoln and Gaston counties.  The speculation is that 
interaction between the westward-directed cold pool from the mature
cells over Lincoln and Gaston counties (Fig. 18), the westerly wind
shear observed in the upper air sounding at GSO, and the low level 
convergent zone in the previous figure resulted in a region of favored 
updraft growth over northern Cleveland County.  Cumulus growing in the 
convergence zone and moving east over northern Rutherford County 
developed rapidly when they encountered this region over northern
Cleveland County.

Figure 18.  Composite reflectivity from KGSP radar at 2058 UTC.  Surface 
observations are indicated by the traditional station model.  The dashed 
grey line indicates the outflow boundary from the multicell clusters over 
the southern Piedmont.  Note the low level convergence over northern 
Rutherford County as suggested by surface observations.  Click on image 
to enlarge.

In reality, the shear was stronger than expected from the morning upper 
air soundings.  The SPC objective analysis of 0-6 km shear at 2200 UTC 
showed moderate amounts of shear (greater than 30 kt), which resulted 
in environmental storm relative helicity in the 0-1 km layer of greater 
than 150 m2/s2.  The storm relative helicity of parcels lifted over the 
cold pool in the vicinity of new cell development over northern 
Cleveland County was likely much greater.  The moderate environmental 
shear combined with the greater potential for generating vorticity north 
of the cold pool boundary in northwestern Gaston County may have allowed 
multicell 4 to develop enough rotation to support the development of the 
tornado.

5.  Summary 

A tornado of F1 intensity touched down in extreme northwern Gaston County, 
North Carolina, to the east of Cherryville, at 2214 UTC on 26 May 2006.  
The tornado produced an intermittent damage path approximately 2.9 miles 
long and 40 yards wide, beginning near the intersection of Hepzibeth Road 
and St. Marks Church Road (Fig. 19).  Winds were estimated at 75 mph.  
The damage included a carport that had been blown into the top of a tree 
and twisted pieces of a barn roof 0.6 miles away from the property.  A 
Tornado Warning was issued by the National Weather Service at 2209 UTC 
(609 pm EDT) for northern Gaston County, providing a lead time of 5 
minutes.

Figure 19.  Approximate path of damage from the tornado that 
moved across northern Gaston County on 26 May 2006.  The purple
colored line is the damage path.  Click on image to enlarge.

The rapid growth of new cells in the presence of moderate shear,
maximizing low level storm relative helicity as they moved to the 
north of an outflow boundary, may have contributed significantly to 
the production of the tornado.

Acknowledgements

The graphics for the storm reports (Fig. 1), 500 mb analyses (Fig. 2), 
upper air soundings (Fig. 4), and mesoscale analysis graphics (Figs. 7
and 9) were obtained from the Storm Prediction Center.  The surface 
fronts analysis (Fig. 3) was obtained from the Hydrometeorological 
Prediction Center.  The surface observations plot (Fig. 6) and all
satellite imagery were obtained from the University Corporation for 
Atmospheric Research.  Radar images were created using the Java NEXRAD 
Viewer and Data Exporter, obtained from the National Climatic Data 
Center.  The background map in Figure 19 was obtained from the U.S.
Geological Survey.