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

1.  Introduction

Tornadoes have not been common over the western Carolinas and 
northeast Georgia the past year, and the few which have occurred 
have been in association with rather unusual radar signatures 
and parent thunderstorms.  The April 3rd storm in Pickens County 
was no exception.  Before we get into details, Figure 1 shows 
the damage path of the tornado as determined by our storm survey 
crew.

The public information statement provides information about the 
path as well.

Figure 1.  Damage path of the tornado east of Clemson, South Carolina, 
on 3 April 2000.  Click on the image to get a larger picture.

2.  Radar Observations

As for the storm itself, the radar operators that night issued 
a severe thunderstorm warning as the cell moved into our county 
warning and forecast area.  The first warning was for Stephens 
County in northeast Georgia.  As the storm raced northeast at 
nearly 50 mph, warnings were quickly issued into Pickens and 
Oconee counties in the early morning hours.

The rather unusual structure of the storm can be seen in the 
radar picture in Figure 2, taken a little less than 30 minutes 
before the tornado touched down.  The top two panels are 0.5 
and 1.5 degree reflectivity cuts respectively and the bottom 
two are storm relative winds for the same cuts as seen by the 
WSR-88D radar at the Greenville-Spartanburg office.  The storm 
of interest is the one nearly in the center of the frame.  
Notice the greens and yellows to the east (right) of the cell. 
It developed behind a previous line of storms.  Typically the 
air behind a line of storms is cooler and less conducive for 
convective development.  There is a sliver of almost no radar 
returns behind the storm.  This is very dry air being pulled 
in behind the storm.  The reason that the dry air is being 
entrained is that the cell itself is embedded in the circulation 
of a mesolow (a small area of low pressure only 100 or so miles
across).  When you look at the larger image, you can see that 
the storm relative winds are swirling into the storm.  The 
red colors to the north of the cell are outbound from the 
radar (located about 50 miles to the east northeast of the 
picture), while the greens just to the south are inbound.  If 
you look carefully you can see a smaller "red/green couplet" 
right at the very bottom of the storm.  This very small 
circulation was apparently the circulation which eventually 
gave rise to the tornado.  This is a very unusual position for 
a mesocyclone (rotating thunderstorm updraft), in a thunderstorm.

Figure 2.  Base reflectivity (top) and storm relative motion 
(bottom) from the KGSP radar at 0.5 degrees (left) and 1.5 degrees
(right) at 0830 UTC on 3 April 2000.  Click on the image to enlarge.

The base reflectivity and storm relative motion from the KGSP 
radar at 0850 UTC are shown in Figure 3, shortly before the 
tornado touched down.  It looks very similar to the image above, 
except that the dry air (blue reflectivities) has wrapped around 
to the south and east side of the cell.  This "occlusion" process 
is possibly what contributed to the weak mesocyclone actually 
forming a tornado just minutes later.  In looking at loop of
these radar data for the duration of the storm (not provided 
here) it was obvious that the cell interacted with a boundary 
left by the earlier thunderstorms at about the time it produced 
the tornado.  Such boundary interactions have been shown many 
times in the past to be associated with tornadogenesis.

Figure 3.  As in Figure 2, except for 0850 UTC.  Click on the 
image to enlarge.

Acknowledgements

Patrick Moore converted the html to the new template.  Vince 
DiCarlo provided the damage survey.