Thunderstorms, some severe, with very heavy rain have moved into the Mid South and Tennessee Valley today. Flash flooding and severe thunderstorms will be possible from the central Gulf Coast to the Upper Great Lakes. The greatest threat of excessive rains will be in portions of the Lower and Middle Mississippi Valleys. This storm system has a history of producing life threatening flash floods. Read More >
The 2-3 January, 2002 Snow Event
over the Western Carolinas and Northeast Georgia
Bryan P. McAvoy
Author's Note: The following report has not been subjected to the scientific peer review process.
This was an unusual event for the Greenville-Spartanburg (GSP) County Warning and Forecast Area (CWFA) in that the precipitation fell as all snow. Local research shows that only about 10 percent of our winter precipitation events are of this type (Moyer 2001). Every county that was included in the initial winter storm watch and warning verified with significant lead time. However, snow began earlier than expected and farther to the northwest. While the southeast part of the CWFA was almost perfectly forecast (see Fig. 1 for snowfall totals) the surprise snow in the southeast slopes of the mountains the afternoon of 2 January will be a large part of this paper's focus. As it became evident that the area of warning criteria snow would extend farther to the north and west, additional warnings were issued shortly after midnight on 3 January. Most counties in this second area of warnings received a very short lead time but they did verify.
Figure 1. Snowfall map, in inches, from the 2-3 January 2002 winter storm.
2. Synoptic Overview
Snowfall amounts ranged from an inch or less over most of the North Carolina Mountains, to as much as 10 inches in the eastern part of the CWFA (Fig. 1). Typical amounts for the storm were from 4 to 6 inches. Amounts exceeded 1 foot over a large part of central and eastern North Carolina. Much of the precipitation in eastern North Carolina fell late on 3 January and early on 4 January, after a nearly twelve hour break in significant precipitation between the warm advection snow and the intensification of the upper level deformation zone, as the upper level energy with the system crossed the area. Figure 3 shows how far upstream the 500 mb upper low was when heavy snow was falling over parts of the GSP CWFA during the late afternoon hours of 2 January. Accumulations associated with the passage of the upper low, and attendant intensification of the deformation zone, were only around one inch in the GSP CWFA. The significant secondary snowfall was mainly along and east of Raleigh, North Carolina. As a result, this review will not focus much on the contribution of the upper low and intensification of the deformation zone, but the interested reader may view this summary from WFO Raleigh (RAH).
Figure 2. Eta model 500 mb geopotential heights and vorticity valid at 2100 UTC 2 January from the 0000 UTC 2 January 2002 cycle. Moderate to heavy snow was occurring over the GSP CWFA at this time. Click on the image for a loop of the 500 MB heights from this run.
While the primary contribution to the heavy snow over the eastern part of the GSP CWFA was from low level warm advection, the precipitation which fell over the mountains of northeast Georgia,South Carolina, and southern North Carolina was not in a region of strong low level isentropic upglide seen on the 292K surface at 2100 UTC from the 0000 UTC 2 January run of the Eta model. While this run of the Eta model did not handle upper forcing well for the afternoon hours of 2 January, all indications are that its lower to middle tropospheric upglide was fairly accurate if a little too far east. However, even though there was little adiabatic omega, the atmosphere over the entire CWFA was saturated, and not particularly stable, so it should not be surprising that the forcing which developed resulted in a snow band.
Up to three inches of snow accumulation had occurred by late afternoon with the narrow snow band, quite a problem as the warnings did not go into effect until the second period (Wednesday night) for areas farther to the east. While higher terrain was certainly a factor in these snow totals (it was anticipated that a dry low level airmass would delay precipitation onset in all but the southernmost zones), this base reflectivity loop (valid from 1200 UTC 2 January through 0700 UTC 3 January) clearly indicates a northeast to southwest oriented band of precipitation over this area in the afternoon from 1700-2100 UTC. Thus, the precipitation in the mountains was likely the result of upper level forcing, as well as the higher terrain. Even as significant snow was falling in the mountains, the central part of the Upstate had nothing but flurries and few heavier snow showers until the early evening hours. This can be seen in these hourly decoded METARS from KGSP for 2 January 2002. This part of the forecast was exactly as expected. Snowfall totals did reach 2 to 3 inches in the southeast-most zones by later afternoon as well. It was anticipated that this might happen, and is not considered a significant bust.
Figure 3. Eta model cross section of relative humidity over Greenville-Spartanburg, SC (KGSP) from the 0000 UTC 02 January 2002 cycle.
The easiest way to compare the differences between the 2 January 0000 UTC run of the Eta and the 1200 UTC run, is through the use of BUFKIT time sections, centered over KGSP. Figure 3 is from the 0000 UTC run on 2 January. This was the run available to the mid shift, when the warnings were issued. Notice how dry the low levels were. Precipitation was not expected to make significant "landfall" until the evening hours over most of the CWFA. Not too far to the south, however, the Eta model projected a much earlier onset time as can be seen on the 292K surface at 1500 UTC. Figure 4 is from the 1200 UTC run of the Eta on 2 January. Notice that the low levels are now completely saturated over KGSP through the afternoon hours. Time sections of omega over KGSP illustrate why the moistening occurred so much more quickly. The 0000 UTC Eta omega time section showed weak upward vertical velocity (UVV) over the region from 1800 UTC to 0000 UTC. Strong UVV was not expected until overnight on 2 January and early on 3 January. The 1200 UTC Eta omega time section was completely different, showing the strongest UVV during the afternoon hours of 2 January.
Figure 4. As in Figure 3, except for the 1200 UTC 2 January 2002 model cycle.
A couple interesting features highlight why the upper-level forcing developed. The EDAS 400 mb winds valid at 1800 UTC on 2 January show a distinct double jet structure, right above the level of the strongest UVV on the 1200 UTC time sections. At 300 mb the jet was more consolidated, though the region was in the entrance region of the jet. Considering the slight anticyclonic curvature of the jet, the deep forcing is fairly easy to intuit.
3. Forecast Performance
The first watches were issued on the afternoon forecast shift of 1 January. Figure 5 shows the total model QPF from Eta model runs from 0000 UTC and 1200 UTC on 1 and 2 January. It was the 1200 UTC run on 1 January which moved the precipitation track farther to the west. Per the Meet-Me conference call held that afternoon, a decision was made to issue a watch for parts of the South Carolina Upstate and the southern North Carolina Piedmont.
Figure 5. Eta model total QPF from the model cycle at 0000 UTC 1 January (upper left), 1200 UTC 1 January (upper right), 0000 UTC 2 January (lower left), and 1200 UTC 2 January (lower right).
A second period warning was issued for the southernmost counties in the CWFA by the midnight shift early on the morning of 2 January (Greenwood, Abbeville and Elbert). The reason for issuing the second period warning was that only around an inch of snow was anticipated by late in the first period, but there was expected to be 4 inches of snow fairly early in the second period. The reasons for this are detailed in the previous section. The previous watch was continued and also expanded to include the rest of the Upstate (the so-called I-85 corridor), but not the mountains, and a significant portion of the Piedmont of North Carolina. Even though the 2 January 0000 UTC Eta run had backed off a little in its QPF, the Canadian GEM continued to advertise heavier precipitation for the eastern half of the CWFA. The forecasters on the mid shift had a great deal of confidence in the Canadian, a model which had outperformed (in the estimate of the forecasters) the Eta in significant winter events over the past two years.
The day shift on 2 January issued warnings for an area that included the original watch area (the one from the previous day shift), and issued advisories for the counties included in the watch by the midnight shift.
During the early morning hours on 3 January, the mid shift upgraded the advisory to a warning for the I-85 corridor of the Upstate as well as the South Carolina and northeast Georgia mountains. As near-winter storm criteria snow had fallen in the mountains, these zones were upgraded to account for light to moderate snow which continued to fall much of the night. In the North Carolina Foothills and western Piedmont, warnings were issued as the deformation zone developed farther west than expected, spreading moderate to heavy snow into this area during the late evening hours on 2 January and early on 3 January. This rather weak deformation zone diminished considerably during the early morning hours. However, the warnings were kept in effect for the entire day of 3 January as some precipitation was expected along and north of the track of the 500 mb low. Around an inch of light snow fell over the southern part of the CWFA during the day associated with this feature.
4. Discussion and Conclusion
A winter storm affected the GSP CWFA starting the afternoon hours of 2 January, and lasting until the late afternoon hours of 3 January. Precipitation fell in association with several different forcing mechanisms during the event. A band of snow associated with an upper level jet streak affected the southern mountains during the afternoon hours of 2 January. This feature was poorly handled by both the Eta model and the forecasters at GSP. Lower to middle tropospheric isentropic upglide developed over the eastern part of the CWFA during the late afternoon hours of 2 January. This was fairly well handled by the Eta model and forecasters. Snow did develop farther west then forecast during the late evening of 2 January. This was the result of a weak deformation zone in the nose of another 300 mb jet streak ejecting out of the trough, coupled with the western most edge of the strong low level upglide. Finally, light some persisted most of the day on 3 January as a strong 500 mb upper low crossed the southern CWFA. This low eventually lead to another bout of heavy snow over eastern North Carolina late on 3 January.
While the event was handled well on the whole, snow consistently developed farther west then expected. Mechanisms responsible for this were upper level forcing caused by a series of jet streaks that crossed the region in southwest flow ahead of a strong 500 mb trough. The Eta model had a poor handle on the ascent caused by these features almost up until the time they developed. In the future, it may benefit the office to issue watches and warnings father west than what low level upglide would suggest, perhaps as far west as the core of the 300 mb jet.
Moyer, Benjamin W., 2001: A climatological analysis of winter precipitation events at Greenville-Spartanburg, SC. Eastern Region Technical Attachment 2001-01. 19 pp.
Patrick Moore reconfigured the page with the GSP web page template.