National Weather Service United States Department of Commerce

Increasing Risk for Flash and River Flooding Through the Weekend

A storm system will track across the central U.S. and set the stage for severe thunderstorms and widespread heavy rainfall through the weekend. There will be a threat for flash and river flooding for portions of the Southern Plains into the Upper Midwest. Read More >

Severe Thunderstorms Across the

Lakelands and Lower Piedmont on

28 September 2006

Patrick D. Moore and Justin D. Lane
NOAA/National Weather Service
Greer, SC

Thunderstorm wind damage in Greenwood, SC, on 28 Sept 2006

Severe thunderstorms moved across Greenwood, South Carolina, late in the afternoon of Thursday, 28 September 2006. Several reports of wind damage were received, including this destroyed shed at a business along the Hwy 72 Bypass on the northwest side of Greenwood. Image courtesy of Lisa Boggs, Greenwood Index-Journal. Used by permission.

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

1. Introduction

An outbreak of severe thunderstorms happened in the afternoon and evening of Thursday, 28 September 2006, across parts of north Georgia, the Carolinas, and Virginia (Fig. 1). In all, the National Weather Service Office at the Greenville-Spartanburg Airport (GSP) issued 18 Severe Thunderstorm Warnings on that day, of which 14 were verified with the occurrence of wind damage or large hail. One severe thunderstorm developed over Elbert County, Georgia, and first produced damaging wind gusts near Elberton. This severe storm produced more wind damage as it moved rapidly across Abbeville County, South Carolina, and eventually evolved into a small mesoscale convective system across Greenwood and Laurens counties. In particular, wind damage was most widespread across the area from Elberton, Georgia, to Calhoun Falls, South Carolina.

(Click here to view a summary of severe weather reports for 28 September 2006.

Severe thunderstorm and tornado reports for 28 September 2006

Figure 1. Wind damage, large hail, and tornado reports for 28 September 2006. Click on image to enlarge.

The outbreak was unusual due to the time of year. Climatologically speaking, the period from mid-September through late October tends to have the least occurrence of severe weather across this part of the United States. Most outbreaks of severe weather during this time of the year are associated with the passage of tropical cyclone remnants.

2. Synoptic Features and Pre-Storm Environment

The potential for severe weather was recognized early in the morning as much of the Foothills and Piedmont of the Carolinas was placed in a slight risk of severe thunderstorms by the Storm Prediction Center (SPC) at 1300 UTC (all times are referenced to Universal Time Coordinated [UTC], which is Eastern Daylight Time plus four hours). The axis of a deep upper trough with a strong embedded short wave, located over the Mississippi Valley at 1200 UTC (Fig. 2), approached the area during the mid and late afternoon of 28 September. Most notable at 1200 UTC was the band of strong winds greater than 70 kts diving into the bottom of the trough over Missouri and Arkansas. The nose of the stronger winds moved around the bottom of the trough and over the western Carolinas by the time of the next 500 mb analysis at 0000 UTC 29 September.

500 mb analysis at 1200 UTC 28 September

Figure 2. SPC objective analysis of 500 mb geopotential height, temperature, and wind at 1200 UTC 28 September. Click on image to enlarge.

By the middle part of the day, a cold front stretched from Louisiana, across the Tennessee Valley, to the upper Ohio Valley , as shown by the surface analysis from the Hydrometeorological Prediction Center (HPC) at 1500 UTC (Fig. 3). Although surface moisture ahead of the system was quite meager, even by early autumn standards (lower to mid 50s dewpoints on Fig. 4), mid-level cooling associated with the approaching short wave resulted in mid level lapse rates on the order of 6 degrees Celsius per kilometer. This combined with the fact that timing was almost optimal to coincide with maximum heating (most locations east of the mountains saw full sun through early afternoon) allowed the atmosphere to become moderately unstable by 1800 UTC, with convective available potential energy (CAPE) generally in the 1000-1500 J/kg range. In addition, the highly dynamic nature of the system resulted in unseasonably strong shear values, with 0-6 km shear of 50-60 kts observed during the event. The 1200 UTC upper air sounding taken at Nashville, Tennessee (KBNA, Fig. 5), was generally representative of the pre-convective environment expected over the western Carolinas in the afternoon. The sounding shows the potential for a CAPE of only 800 J/kg, but strong shear on the order of 60 kts.

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

Surface analysis at 1500 UTC 28 September

Figure 3. HPC surface pressure and fronts analysis at 1500 UTC 28 September. Click on image to enlarge.

Regional surface plot at 1843 UTC 28 September

Figure 4. Regional surface plot at 1843 UTC 28 September. Observations are plotted according to the standard station model. Click on image to enlarge.

Upper air sounding at BNA at 1200 UTC 28 September 2006

Figure 5. Skew-T log P diagram (upper left) and hodograph (upper right) for upper air sounding at BNA at 1200 UTC 28 September. The tables at the bottom summarize several objective parameters used by the SPC to determine severe weather potential. Click on image to enlarge.

Considering the generally unidirectional profiles and large surface dewpoint depressions, damaging straight line winds were thought to be the main threat in terms of severe weather, as indicated in the updated Day 1 Severe Weather Outlook issued by the SPC at 1630 UTC. The development of a broken line of strong to severe thunderstorms over east Tennessee in the early afternoon prompted the SPC to issue a Severe Thunderstorm Watch for much of the southern Appalachians at 1745 UTC.

3. Radar observations

Several thunderstorms developed in a broken band across north Georgia, the western tip of South Carolina, and the North Carolina Foothills during the early part of the afternoon, out ahead of the cold front which was located at 1800 UTC over the Great Valley of east Tennessee. One thunderstorm produced wind damage across the area near Morganton, North Carolina, around 1900 UTC. However, it was increasingly apparent by 1900 UTC that the potential for wind damage extended to the south of Watch #802, as indicated in Mesoscale Discussion #2045 issued by the SPC at 1919 UTC and the updated Day 1 Convective Outlook at 1956 UTC. Another thunderstorm produced a 47 mph wind gust as it moved across the airport west of Anderson, South Carolina, around 1946 UTC, which gave forecasters a clear indication of the wind damage potential (Fig. 6).

KGSP 0.5 deg Reflectivity at 1946 UTC 28 September

Figure 6. Radar reflectivity on 0.5 degree scan from the KGSP WSR88-D radar at 1946 UTC. The radar is located at the upper right edge of the image. Click on image to enlarge.

Although there were several marginally severe discrete cells and line segments that developed during this event, by far the most interesting was the discrete cell that produced widespread wind damage from Elberton to Abbeville, before evolving into more of a linear Mesoscale Convective System (MCS). This particular thunderstorm had its genesis over northeast Georgia prior to 2000 UTC. The cell strengthened rapidly as it moved east from Madison County, Georgia, prompting the issuance of a Severe Thunderstorm Warning for Elbert County at 2027 UTC. Additional strengthening took place as the leading edge of the severe thunderstorm bowed outward toward the city of Elberton at 2041 UTC (Fig. 7). The weaker reflectivity behind the bowing part of the storm, west of Elberton, is known as a weak echo channel and is indicative of very strong winds. Based on this scan, a Severe Thunderstorm Warning was also issued for Abbeville County at 2045 UTC.

Click here to view a 24 frame java loop of 0.5 degree reflectivity from the KGSP radar from 2020 UTC to 2158 UTC.

KGSP 0.5 deg Reflectivity at 2041 UTC 28 September

Figure 7. As in Fig. 6, except for 2041 UTC. Note the leading edge of the higher reflectivity bowing outward over Elberton and the relatively weaker reflectivity immediately behind (west of) the bow west of Elberton. Click on image to enlarge.

The Elbert County storm developed into a classic "spearhead echo" (Fujita 1976, Fujita and Byers 1977) as it moved across the area between Elberton and Calhoun Falls, South Carolina. (For other examples of spearhead echoes, see these examples from Huntingdon County, Pennsylvania, and Perry County, Missouri.) These rare, compact, fast-moving cells (this one was moving at about 47 knots) usually pack quite a wallop, and this one was certainly no exception. Although no direct measurements of wind speed were available, numerous reports were received of large limbs or trees blown down from Elberton to Calhoun Falls.

KGSP 0.5 deg Reflectivity at 2054 UTC 28 September

Figure 8. As in Fig. 6, except for 2054 UTC. Note the "spearhead" appearance of the higher reflectivity protruding eastward between Elberton and Calhoun Falls. Click on image to enlarge.

As the Elberton storm moved almost due east into Abbeville County, the path of storm became increasingly perpendicular to the beam from the KGSP radar. As a result, the Doppler velocity was most likely sampling a smaller and smaller component of the true wind velocity in the storm. However, the WSR-88D radar at the NWS office in Columbia (KCAE) was in a more favorable location to view the rear-to-front wind velocity in the spearhead echo, as the storm was moving toward that radar (Fig. 9). The velocity along the radial from the KCAE radar was greater than 60 kts at the level of the radar beam, which at that location was approximately 2.4 km above ground level (Fig. 10).

KCAE 0.5 deg Reflectivity at 2056 UTC 28 September

Figure 9. Radar reflectivity on 0.5 degree scan from the KCAE WSR88-D at 2056 UTC, showing the spearhead echo moving into Abbeville County. The radar is located off the right side of the image. Click on image to enlarge.

KCAE 0.5 deg Radial Velocity at 2056 UTC 28 September

Figure 10. As in Fig. 9, except for radial velocity. Compare the location of the higher Doppler velocities with the location of the spearhead echo in Figure 9. Click on image to enlarge.

The Elberton/Abbeville storm eventually merged with other cells to produce more of an elongated (although still quite compact), "typical" MCS over the Lower Piedmont of South Carolina. Additional Severe Thunderstorm Warnings were issued for Greenwood County at 2120 UTC and for Laurens County at 2122 UTC in advance of the MCS. Even though the "spearhead echo" was no longer apparent, the MCS continued to produce fairly widespread tree damage across Greenwood County and the southern half of Laurens County.

4. Discussion

Although meteorologists often look for mid-level dry air when attempting to assess the potential for damaging winds, either due to initiation of negatively bouyant air or enhancement of pre-existing downdrafts through entrainment of dry air, it is important to remember that organized convection can develop intense subsiding air currents through various pressure gradient forcings. In the 28 September case, there wasn't much mid-level dry air, but organized convection was a given, as a consequence of the very strong shear. Subsiding air streams generated by the Elberton storm were probably enhanced by the dry air in the sub-cloud layer. Surface dewpoint depressions of ~30 deg. F were observed in the pre- convective environment across the southern Upstate. Combine these factors with the fact that the cell in question was moving at 47 kts, and you have a serious wind damage threat on your hands.

The Elberton/Abbeville storm was a good example of how storm history plays a role in the warning process when it comes to organized convection. Systems like these produce deep, nasty cold pools that are being continuously reinforced, and they don't easily give up the ghost. Often, when the lead- line convection appears to be weakening, it probably just means that the cold pool is becoming too deep and strong to force air parcels to their level of free convection. In these cases, wind damage is still quite possible, even likely due to the strength of the cold pool.

5. Summary

In the late afternoon of Thursday, 28 September 2006, an organized severe thunderstorm produced a swath of wind damage across the area roughly bounded by Elberton, Georgia, and Iva, South Carolina, across much of Abbeville County and Greenwood County, to the southern half of Laurens County, South Carolina. The most widespread wind damage occurred in the area between Elberton and Calhoun Falls, coincident with a spearhead echo observed on the KGSP radar. All reports of wind damage were preceded by a Severe Thunderstorm Warning. For the entire event, the NWS office in Greer, South Carolina, issued 18 Severe Thunderstorm Warnings. Of those warnings, 14 were verified by reports of wind damage or large hail. There was one missed event. Verification scores are as follows: Probability of Detection: 94% False Alarm Ratio: 22% Critical Success Index: 0.74 Average Lead Time: 8.7 minutes

Large tree limbs knocked down onto power lines in uptown Calhoun Falls, SC, 28 September 2006Large tree snapped mid-trunk in Calhoun Falls, SC, 28 September 2006

Tree damage around Calhoun Falls, South Carolina, courtesy of Darlene Cox. Click on images to enlarge.

Large tree limbs knocked down near Greenwood, SC, 28 September 2006Large tree uprooted in Greenwood, SC, 28 September 2006

Additional images of tree damage near Greenwood, South Carolina, courtesy of Lisa Boggs, Greenwood Index-Journal. Click on images to enlarge.


Chris Horne contributed additional insight as to the warning process during the event. The plot of storm reports in Figure 1, upper air charts, and the upper air sounding in Figure 5 were obtained from the Storm Prediction Center. Surface analyses were obtained from the Hydrometeorological Prediction Center. The regional surface plots and satellite imagery were obtained from the University Corporation for Atmospheric Research. All radar imagery graphics were created using the Java NEXRAD viewer obtained from the National Climatic Data Center. Our gratitude is given to Lisa Boggs of the Greenwood Index-Journal and to Darlene Cox for the use of their storm damage images.



Fujita, T. T., 1976:  Spearhead echo and downburst near the approach end of 
     a John F. Kennedy Airport runway, New York City. SMRP Res. Paper No. 137 
     [NTIS No. N76-2184/1GI], Univ. of Chicago, 51 pp.

Fujita, T. T., and H. R. Byers, 1977:  Spearhead echo and downburst in the 
     crash of an airliner.  Mon. Wea. Rev., 105, 129-146.