Ice coats tree branches and power lines at the National Weather Service Forecast Office at the Greenville-Spartanburg International Airport on 23 January 2000. Picture taken by the NWS.

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

1.  The 22-23 January Winter Storm

The third time was the charm as the entire Greenville-Spartanburg County 
Warning Area (CWA) experienced winter storm conditions on 22-23 January 2000.  
This was the third winter event to affect the CWA in a week, the first two 
affecting the northern most counties in our area of responsibility.  Not 
this time.  The entire CWA experienced a significant amount of snow or ice 
for the event (Fig. 1).  Amounts listed generally reflect what was on the 
ground the morning of Sunday, 23 January.  Snowfall was locally higher 
during the night of 22 January, with less on the ground due to melting by 
the afternoon of 23 January.

Figure 1.  Storm total snow and ice accumulation for the 22-23 January 2000 
winter storm.  Click on image to enlarge.

The ice storm most significantly affected counties from Franklin County,
Georgia, and Laurens and Anderson counties in South Carolina, and to the
south with numerous reports of downed trees and several power outages.  
Most of the precipitation occurred as a very moist airmass overran a cold 
and initially dry arctic airmass in the lower levels of the atmosphere.  
Temperatures along the Gulf Coast rose into the 60s and 70s on 22 January, 
while only 100 miles to the north highs only rose into the middle 30s.  
It was this extremely warm and moist airmass which allowed for such 
significant amounts of snow and ice despite the lack of a very organized 
surface area of low pressure.

2.  The 17-18 January 2000 Winter Storm

During the late evening and early morning hours of January 17th and 18th 
a significant snow event occurred over parts of the North Carolina Piedmont, 
Foothills and Mountains (Fig. 2).  Most of the precipitation fell in about 
a six hour period during the early morning hours of the 18th.  Further to 
the south, the precipitation fell as a mix of snow, sleet, rain and freezing 
rain.  The area affected and precipitation type were very well predicted by 
our office; however, the forecasted amounts were about half of what actually 
fell.

Figure 2.  Storm total snow and ice accumulation for the 17-18 January 2000 
winter storm.  Click on image to enlarge.

The most challenging aspect to forecasting this event was the lack of a 
significant moisture source.  The numerical models that forecasters use to 
help in their determination of precipitation amounts generated only around 
a tenth of an inch of liquid equivalent precipitation (roughly equal to an 
inch of snow).  In actuality, liquid equivalent precipitation amounts 
exceeded a half inch, with snowfall amounts of around six inches in many 
locations from Avery County, North Carolina, east to Davie County, North
Carolina, and as far south as northern Lincoln County, North Carolina.

Also lacking was a significant feature in the upper atmosphere to help 
generate precipitation.  The Eta model showed only a very weak short wave 
in the fast northwesterly flow at 500 mb at about the time the snow began 
(Fig. 3).  From 200 to 300 mb, a weak convergent flow was observed.  This 
often equates to downward motion in the atmosphere.

Figure 3.  Eta model forecast of 500 mb geopotential height (dm; purple
contours) and vorticity (1 x 10-5 s-1; white contours) valid at 0600 UTC 
on 18 January from the 0000 UTC 18 January forecast cycle.  Click on 
image to enlarge.

The single most important contributing factor to this event was warm 
advection in the lowest 3 km of the atmosphere.  The 290K isentropic
surface (Fig. 4) showed significant warm advection (wind barbs orthogonal 
to the pressure lines, pointing from higher to lower pressure).  The 
contoured fields in yellow represent pressure advection.  Also known as 
isentropic omega, this field represents lift on isentropic surfaces.  
Note the finger of lift into western North Carolina.

Figure 4.  Eta model forecast of pressure (mb; purple contours), pressure 
advection (s-1; white contours), and wind (kt; barbs) valid at 0600 UTC on 
18 January from the 0000 UTC 18 January forecast cycle.  Click on image 
to enlarge.

Despite having no "Gulf Connection," the system did tap significant in-situ 
moisture.  At 700 mb, a pool of 4 to 4.5 g kg-1 mixing ratio was observed 
just upstream of the region at 0600 UTC (about 1 AM local time) on 
18 January (Fig. 5).  This pool of moisture crossed the region over the 
next six hours, helping to increase the efficiency of the precipitation 
(i.e. increased snowfall rates.)

Figure 5.  Eta model forecast of 700 mb geopotential height (dm; purple 
contours), mixing ratio (g kg-1; white contours), and wind (kt; barbs) 
valid at 0600 UTC on 18 January from the 0000 UTC 18 January forecast cycle.  
Click on image to enlarge.

The final challenge with the event was determining the type of precipitation. 
Model soundings and local thickness schemes (not available at this time) 
indicated that the atmosphere would be cold enough to support snow over the 
central North Carolina Foothills and Piedmont, while further to the south the 
strong warm advection would cause the precipitation to be in form of freezing 
rain and snow.  This was almost exactly how the event played out.

Acknowledgements

The images of the Eta model forecast were made using PC-GRIDDS.  The original
document has been reformatted and edited by Pat Moore.