The Ice Storm of 15 December 2005
in the Western Carolinas and Extreme Northeast Georgia
Laurence G. Lee and Patrick D. Moore
NOAA/National Weather Service
Photo by John Cunningham
Photo by David P. Smith
Author's Note: The following report has not been subjected to the scientific peer review process.
A damaging ice storm occurred across portions of the western Carolinas and
extreme northeast Georgia on Thursday, 15 December 2005. Rain and freezing
rain began in the western portion of the WFO Greenville-Spartanburg (GSP)
County Warning Area (CWA) around midnight and spread across the entire area
during the morning. The precipitation continued into the afternoon, but it
moved steadily from west to east so that almost all precipitation ended by
sunset. Ice accumulations between one quarter and three quarters of an inch,
with locally greater amounts, caused widespread damage to trees and power
lines, resulting in numerous power outages. One quarter to one half inch
thick ice occurred in a band extending from extreme northeast Georgia across
Upstate South Carolina into the Foothills and northwest Piedmont of North
Carolina (Fig. 1). Several locations experienced more than three quarters of
an inch of ice, with the most significant ice accretion occurring across the
area bounded approximately by Hendersonville and Tryon, North Carolina, and
Greenville, South Carolina. No ice accumulation was observed in the far
western counties of North Carolina along the Tennessee border.
Figure 1. Total ice accumulation (inches) for 15 December 2005. Note that
sharp gradients in accumulation may not be indicated at the scale of the
graphic. Click on image to enlarge.
(Click here to view a list of ice accumulation reports for 15 December 2005.
The icing event was caused by a low pressure system that moved across the
southeastern United States while a high pressure system centered over New
England and the Canadian Maritime Provinces extended southward along the
Eastern Seaboard. Cold air damming east of the Appalachians developed nearly
coincidently with the onset of precipitation. Even though rapid warming
occurred aloft, the subfreezing surface wet bulb temperatures caused the
rain to freeze on contact. Highway and road conditions became hazardous
only in a few areas. The significant ice accretion was limited to trees,
power lines, signs, and other objects above the ground. The highest
elevations in western North Carolina extended above the warm layer. Several
locations above 5,000 feet received as much as ten inches of snow.
2. Synoptic Features and Associated Weather
A broad area of precipitation spread across the southeastern United States
between 0000 UTC and 1200 UTC on Thursday, 15 December 2005. (Universal Time
Coordinated [UTC] is Eastern Standard Time plus five hours, during the winter
months.) The precipitation and a surface low pressure system developed in
the isentropic lift region of the coupled jet structure depicted in the Storm
Prediction Center (SPC) 0000 UTC 300 mb analysis (Fig. 2). The dominant
characteristic of the 500 mb analysis was a low height center over the north
central United States (Fig. 3), but the key feature with regard to the
precipitation spreading toward the southeastern states was a short wave
trough over Texas. The short wave trough and associated surface low moved
from the Mississippi Delta to South Carolina during the next 24 hours.
Figure 2. SPC objective analysis of 300 mb wind isotachs, streamlines,
and divergence at 0000 UTC 15 December 2005. Click on image to enlarge.
Figure 3. SPC objective analysis of 500 mb geopotential height and
temperature at 0000 UTC 15 December. Click on image to enlarge.
In the lower troposphere, the isotherm pattern at 850 and 925 mb (Figs. 4 and 5)
revealed the presence of the surface based cold air east of the mountains.
However, significant warm advection was occurring at both levels across the
Carolinas and Georgia as indicated by the southeast and south winds blowing
nearly perpendicular to the isotherms. The southerly wind flow caused rapid
warming aloft and served to transport moisture into the region. The 0000 UTC
surface analysis identified a well-defined axis of high pressure in a cold air
damming configuration east of the Appalachians. Surface temperatures across
the Piedmont of North Carolina were generally between 30 and 35 deg F, and
dewpoints were generally between 10 and 15 deg F. In the North Carolina
Mountains around Asheville, temperatures were in the upper 20s. Temperatures
in the far western mountains were in the lower and middle 30s. [Note: During
cold air damming events, it is quite common for locations west of the
Asheville-Hendersonville area to be several degrees warmer because of the
inability of the shallow cold air to penetrate beyond the mountains on the
west side of the French Broad Valley. This relationship between topography
and low-level temperature advection explains in large part the smaller number
of freezing rain occurrences in the far western mountain counties during cold
air damming. -Ed] Temperatures in Upstate South Carolina were in the lower 30s,
and dewpoints were in the teens. Wet bulb temperatures prior to the onset of
precipitation were generally in the 25 deg F to 30 deg F range. The 0600 UTC
GSP surface analysis (Fig. 6) shows the surface pressure pattern at the time
when the precipitation was spreading across the southwestern portion of the
Figure 4. SPC objective analysis of 850 mb geopotential height, temperature,
and dew point at 0000 UTC 15 December. Click on image to enlarge.
Figure 5. SPC objective analysis of 925 mb geopotential height, temperature,
and dew point at 0000 UTC 15 December. Click on image to enlarge.
Figure 6. Sea level pressure contours (mb) and surface fronts analysis
for 0600 UTC 15 December. Click on image to enlarge.
The progression of upper level features can be seen in Figs. 7 and 8. By
1200 UTC the region of upper level ascent moved into the southeastern states
resulting in widespread precipitation. Even though warming occurred aloft,
surface temperatures east and southeast of the mountains remained below
freezing due in large part to cold air damming, as seen in the 1200 UTC
surface analysis (Fig. 9). Evaporational cooling at the onset of the
precipitation contributed to the lowering of surface temperatures which were
already near or below freezing at many locations. The near surface cooling
also increased the static stability of the air mass thus strengthening the
damming. Following the onset of the precipitation across the western
Piedmont and Foothills, temperatures warmed very little during the remainder
of the morning because of the cloud cover, precipitation, and cold advection.
These effects were sufficient to overcome - or at least balance - the warming
that accompanied the release of latent heat during the freezing process.
Figure 7. SPC objective analysis of 300 mb wind isotachs, streamlines, and
divergence at 1200 UTC 15 December. Click on image to enlarge.
Figure 8. SPC objective analysis of 850 mb geopotential height, temperature,
and dew point at 1200 UTC 15 December. Click on image to enlarge.
Figure 9. Sea level pressure contours (mb) and surface fronts analysis
for 1200 UTC 15 December. Click on image to enlarge.
By mid afternoon on 15 December, the parent high had moved far enough to the
east so that further cold air advection in the axis of the damming high was
greatly diminished. The eastern edge of the high eroded as a coastal warm
front moved inland. For example, the temperature at Raleigh-Durham increased
from 31 deg F at 1200 UTC to 35 deg F at 1500 UTC and to 41 deg F at 1800 UTC.
Hourly data from Asheville, Greer, Hickory, and Charlotte show nearly steady
temperatures in the cold air damming region (Table 1).
Table 1. Hourly precipitation type at AVL (Asheville), GSP (Greer), HKY (Hickory),
and CLT (Charlotte), for the period from 0600 UTC 15 December to 0000 UTC 16 December.
3. Thermal Structure
Freezing rain was the predominant precipitation type because the temperature
profile in the lower atmosphere consisted of below freezing surface
temperatures and above freezing temperatures aloft. Figure 10 displays the
North American Mesoscale Model (NAM) 00-hour vertical temperature profile
for GSP at 1800 UTC on 14 December. The panel on the right shows the
vertical dewpoint, wet bulb, and dry bulb temperature structure. A slight
warm nose existed at about 7,000 feet Above Ground Level (AGL), but the wet
bulb temperature profile indicated the entire troposphere could be cooled
to subfreezing if precipitation were introduced and no warm advection
occurred. However, the wind barbs on the right clearly indicated a well-
defined southwest flow above approximately 4,000 feet AGL with a veering
profile indicative of warm air advection. From the surface to approximately
3,500 feet the wind was blowing from the northeast. The panel on the left
side of Fig. 10 shows the universal precipitation type nomogram with a red
dot marking the 1000-850 mb and 850-700 mb values. The nomogram indicated
the most likely precipitation type, if it occurred, during the subsequent
six hours would be snow mixed with freezing rain and sleet.
Figure 10. Bufkit display for GSP from 00-Hour NAM at 1800 UTC 14 December.
The precipitation type nomogram is shown on the left. The dewpoint, wet bulb
temperature, dry bulb temperature, and wind profile is shown on the right.
Click on image to enlarge.
The eastward progression of the upper level features seen in Figs. 7, 8, and
9 resulted in continued warming just above the cold surface layer. As
precipitation spread into the southern Appalachian region, warming aloft
continued, but the cold air damming resulted in virtually no change in the
surface temperatures which were freezing or below from northeast Georgia
through upstate South Carolina into the western Piedmont of North Carolina.
Figure 11 shows the NAM 1200 UTC 15 December GSP initial hour profiles and
precipitation type nomogram. The most noteworthy feature of this thermal
structure was the subfreezing layer extending from the surface to about
2,000 ft AGL. Above the surface-based cold layer, the temperature increased
rapidly to approximately 5 deg C at 3,000 ft then cooled to zero deg C near
7,000 ft AGL. A top-down evaluation indicated snow descended into the warm
layer where complete melting occurred then the liquid precipitation froze on
impact in the colder air at the surface.
Figure 11. As in Fig. 10 except 1200 UTC on 15 December. Click on image
Figure 12 is the 18-hour NAM GSP forecast valid at 1200 UTC on 15 December
2005. The temperature and moisture profiles on the right were quite similar
to the 1200 UTC initial analysis in Fig. 11. A subfreezing layer with
northeast winds was capped by a warm layer and southwest winds veering with
height. The universal nomogram values on the left indicated that freezing
rain would be the predominant precipitation type during the next six hours.
The path depicted by the increasingly large dots on the nomogram provided a
guide regarding the trend of the precipitation type forecasts. The
significant warming of the 850-700 mb layer during the three most recent
hours was compatible with the veering wind and warm advection that occurred
just above the surface.
Figure 12. As in Fig. 10, except 18-hour forecast for GSP verifying at 1200
UTC on 15 December. Click on image to enlarge.
Other numerical guidance provided an indication that a freezing rain event
was likely. Figure 13 is an example from the National Centers for Environmental
Prediction (NCEP) Short Range Ensemble Forecast (SREF) model run at 0900 UTC
on 13 December 2005. The forecast panels are three-hour forecasts valid at
1200 UTC on 15 December. The panel in the lower right hand corner indicated
much of the Greenville-Spartanburg CWA was susceptible to freezing rain.
Figure 13. NCEP SREF three hour mean probability of precipitation (shaded)
and mean three hour QPF (contour) from the 0900 UTC 13 December 2005 run valid
at 1200 UTC 15 December. Click on image to enlarge.
Another indicator of the significance of the warm air advection was the rapid
temperature increase at Wayah Bald (elevation 5469 ft MSL) in Macon County,
North Carolina. Figure 14 is a temperature trace from the North Carolina
Environment and Climate Observing Network (ECONET) station on the top of the
mountain. The temperature climbed from 18 deg F around sunrise on 14 December
to the lower 30s by midnight on 15 December. The temperature peaked at 35 deg
F during the morning of 15 December, but it dropped into the teens later that
night. An interesting feature of this event is the shallow nature of the warm
air layer depicted in the GSP Bufkit profiles and the Wayah Bald temperature
record. The warmest portion of the layer remained below some of the highest
elevations in the southern Appalachians. The National Weather Service
cooperative observer at Mt. Mitchell, NC measured ten inches of snow. Three
to ten inches of snow fell along the Blue Ridge Parkway near Mt. Mitchell
State Park. NWS cooperative observers at Newfound Gap and Mt. LeConte in the
Great Smoky Mountain National Park measured 10 and 11 inches of snow,
respectively. Sleet mixed with snow produced local accumulations ranging
from one to four inches in a narrow band extending from southeastern Buncombe
County through Yancey, Mitchell, and Avery counties.
Figure 14. Wayah Bald, NC 2 meter temperature during the week of 12 December
2005. Vertical bars identify 15 December (EST). Click on image to enlarge.
While the warming was occurring just above the surface, the low-level pool
of cold air that formed as a result of cold air damming persisted throughout
the day on 15 December. The rising temperatures in the warm layer just above
the surface contributed to maintenance of the cold air resting against the
eastern slopes of the Appalachians. The strong inversion resulting from the
differential temperature advection provided a buffer resisting the incursion
of low level warm air from the south and east accompanying the eastward
motion of the surface low pressure and the inland progression of the coastal
warm front. By mid to late afternoon, however, surface temperatures were
above 32 deg F in all areas except some of the mountain valleys and in a very
narrow zone along the foot of the Blue Ridge escarpment. The precipitation
ended across the entire area by 0000 UTC on 16 December.
Precipitation totals (liquid equivalent) for the event included the following:
Anderson, 1.52 inches; Asheville, 1.48 inches; Charlotte, 1.76 inches; and
Greenville-Spartanburg, 1.55 inches. Some of the rain fell while temperatures
were above freezing, and the rate of precipitation during the period of
subfreezing temperatures was great enough so that much of the rain dripped
to the ground before freezing occurred.
A low pressure system traveling across the Southeast during a period of cold
air damming on 15 December 2005 caused a significant icing event for portions
of the western Carolinas and extreme northeast Georgia. Ice accumulations
of at least one quarter inch were common across Upstate South Carolina, a
portion of the North Carolina Mountains, and the Foothills and western
Piedmont of North Carolina. Local ice thicknesses of approximately one inch
were reported in the area bounded roughly by Greenville, South Carolina, and
Tryon and Hendersonville, North Carolina. Freezing rain was the predominant
precipitation type because a layer of warm air between approximately 2,000 ft
and 6,000 ft AGL spread across the surface based cold air. The precipitation
formed as snow then melted as it fell into the warm air. The liquid
precipitation froze on contact with objects in the surface based cold air.
Snow accumulated on some of the high elevations in western North Carolina
that extended above the warm layer.
The authors wish to thank Neil Dixon for assistance with converting the
original document to html. Figures 2, 3, 4, 5, 7, and 8 are courtesy of the
NOAA/NWS/NCEP Storm Prediction Center. Figure 13 is courtesy of the
NOAA/NWS/NCEP Environmental Modeling Center. Figure 14 was obtained from the
State Climate Office of North Carolina.