The Mountain Snow Event
of 11-13 February 2006
Patrick D. Moore
National Weather Service
Photo by L. G. Sheets
Author's Note: The following report has not been subjected to the scientific peer review process.
A major east coast winter storm produced record-setting snowfall across the population centers of the northeastern United States during the weekend of 11-12 February, 2006 (Grumm 2006) pdf. (Click on these links to view more information concerning impacts elsewhere along the East Coast, including Albany NY). Before the low pressure system responsible for this high-impact event deepened explosively off the Mid-Atlantic coast, it moved across the midlands of the Carolinas, producing a significant amount of snow mainly across the mountains of North Carolina after its passage (Fig. 1). The effect of this winter storm across the western Carolinas and northeast Georgia can be split into two phases: An initial phase associated with the passage of the surface low (the synoptic phase), and a second phase associated with a northwest flow at low and middle levels of the atmosphere in the wake of the low (the northwest flow phase). The northwest flow phase presented its own set of challenges, including the eastward extent to which snow showers would reach and the potential contribution of shallow convection. These facets are explored in subsequent sections.
Figure 1. Total snow accumulation in inches for the period 0000 UTC 11 February through 1800 UTC 13 February 2006. Note that sharp gradients in accumulation across the higher terrain of western North Carolina may not be indicated at the scale of the graphic. Click on image to enlarge.
(Click here to view a list of snow accumulation reports for the period 11-13 February 2006
An upper-level pattern shift about one week prior to the development of the winter storm resulted in a deep upper trough at the 500 millibar (mb) level over the eastern half of North America, which is a pattern typical of mid winter in the eastern United States. The days leading up to the event were characterized by below normal temperatures across the western Carolinas. The upper trough provided entry for a high pressure air mass of arctic origin into the northern plains on Wednesday, 8 February. The arctic high moved down across the mid-Mississippi Valley on Thursday, 9 February, and pushed a surface cold front across the Southeast and over the northern Gulf of Mexico. However, the high continued to weaken as it moved across the Southeast Thursday night and off the coast Friday morning, 10 February, as a secondary cold front approached from the northwest. By 1200 UTC on 10 February, the surface analysis from the Hydrometeorological Prediction Center (HPC) showed a cold front stretching in an arc from central Illinois, across the Ozark Plateau, to the Big Bend region of southwest Texas, with a weak wave of low pressure on the front over north central Texas (Fig. 2, left). Meanwhile, another weak low pressure area remained over south Texas along the first cold front. A strong short wave at 500 mb was shown by the Storm Prediction Center (SPC) objective analysis, diving across the northern Plains and upper Mississippi Valley to reinforce the upper trough (Fig. 2, right).
Figure 2. Sea level pressure contours (mb) with HPC surface front analysis (left) and 500 mb SPC objective analysis of height contours (hPa) (right) for 1200 UTC 10 February. Click on each image to enlarge.
The Synoptic Phase The low pressure system responsible for this significant event had its origin in the two weak lows over Texas on the morning of Friday, 10 February. Throughout the day, increasing baroclinicity ahead of the short wave dropping down from the northern Plains, combined with upward vertical motion in the right entrance region of a jet streak stretching across the Ohio Valley and Mid-Atlantic regions seen at the 250 mb level, provided an environment favorable for the development of low pressure over the northwestern Gulf Coast region. Infrared imagery from the GOES-12 satellite (Fig. 3) shows the development of a baroclinic leaf structure across the Arklatex and lower Mississippi Valley regions, indicating that cyclogenesis was occurring, although it was weak. By 0000 UTC 11 February, the two lows were organizing into one low center over the Mississippi Delta region (Fig. 4).
Figure 3. GOES-12 enhanced infrared imagery at 0015 UTC 11 February. Click on image to enlarge.
Figure 4. Sea level pressure contours (mb) and HPC surface fronts analysis for 0000 UTC 11 February. Click on image to enlarge.
Forecasters expected a Miller Type-A surface low (Miller 1946) to track to the south across central Georgia and central South Carolina before moving up the East Coast. Forecast soundings and partial thickness values suggested the main precipitation type issue would be a determination between rain and snow. The Quantitative Precipitation Forecast (QPF) from the operational Global Forecast System (GFS) and North American Mesoscale (NAM) forecast models suggested between one-quarter and one-half inch of liquid could be expected (Fig. 5), which would be enough to support a forecast of 3 to 5 inches of snow across the higher terrain, generally above 2000 feet. However, periods of moderate precipitation were expected that would provide enough cooling of low levels through melting and evaporation to bring the snow level down closer to 1,000 feet during the early morning hours on 11 February.
Figure 5. Quantitative Precipitation Forecast (inches) from the 1200 UTC 10 February model cycle for the 6-hour period ending (a) 1200 UTC 11 February from the NAM-80 model, (b) 1800 UTC 11 February from the NAM-80 model, (c) 1200 UTC 11 February from the GFS-80 model, and (d) 1800 UTC 11 February from the GFS-80 model. Click on each image to enlarge.
Precipitation developing in the increasingly moist southwest flow at low levels across Alabama and Georgia during the early part of the evening on 11 February was aided by a low level jet at 850 mb stretching from the northern Gulf of Mexico to the Tennessee Valley region. Light precipitation reached the southwestern corner of North Carolina and extreme northeast Georgia around 0300 UTC. The warm moist upglide associated with the low level jet translated eastward over the western Carolinas by 0600 UTC and allowed for precipitation to spread northeast across the upstate of South Carolina and the mountains of North Carolina.
(Click here to view a 19 frame java loop of radar reflectivity centered on the Greer (KGSP) WSR-88D radar, depicting the development and movement of the precipitation across the western Carolinas.)
Figure 6. Surface low track from the HPC Surface Analysis Branch for 0300 - 2100 UTC 11 February. Sea level pressure (mb) at the low center is underlined.
At the surface, the low pressure system remained relatively weak across the Deep South during the early morning hours of Saturday 11 February, as the center of the low moved to a position near Mobile, Alabama, at 0600 UTC and a position near Columbus, Georgia, at 1200 UTC (Fig. 6). However, the upper level system continued to slowly gain strength with a 531 decameter low closing off over northern Illinois by 1200 UTC on the 500 mb analysis. Light precipitation, forced by moist ascent in the developing warm conveyor belt (Carlson, 1980) ahead of the deepening upper low, spread across the Piedmont of the Carolinas between 0600 UTC and 0900 UTC. In the same time period, radar and infrared satellite imagery showed the emergence of bands of light precipitation over the Tennessee Valley and Cumberland Plateau, forced by weak upward motion in a developing deformation zone to the northwest of the surface low.
The digging upper system helped to strengthen the subtropical branch of the jet stream with a 120 knot jet streak developing over south Texas and the northern Gulf of Mexico by 1200 UTC. The interaction between the developing subtropical jet and the existing strong polar jet streak over the Appalachians and the Mid-Atlantic coast contributed to increased vertical motion ahead of the surface low over southern Alabama and southern Georgia between 0600 UTC and 1200 UTC. The coupling of jet streaks, combined with weak convective instability along the Gulf Coast, provided a favorable environment for deep convection to develop on the leading edge of the precipitation shield along the Gulf Coast. In fact, a small linear mesoscale convective system (MCS) developed over south Alabama at 0600 UTC and 0900 UTC (Fig. 7), which proceeded to move quickly east across southwest Georgia and northwest Florida by 1200 UTC, well ahead of the surface cold front which lagged across southeast Alabama and the western part of the Florida Panhandle at that time.
(Click here to view a 16 frame java loop of radar reflectivity centered on the Maxwell Air Force Base (KMXX) WSR-88D radar, which shows the progression of the MCS across Alabama, Georgia, and north Florida.)
Figure 7. Radar reflectivity (dBZ) mosaic centered on Robins AFB (KJGX) WSR-88D at (a) 0600 UTC, (b) 0900 UTC, and (c) 1200 UTC, on 11 February. Click on each image to enlarge.
The center of surface low pressure moved over northeast Georgia in the morning to a position near Athens by 1500 UTC (Fig. 6). The back edge of a weakly organized warm conveyor belt reached the western tip of North Carolina between 1200 UTC and 1500 UTC, nearly coincident with the cold front at 850 mb and the leading edge of the dry slot seen on water vapor satellite imagery (Fig. 8). The eastward movement of this feature brought an end to precipitation to the west of a line from Morganton, North Carolina, to Greenville, South Carolina, including all of northeast Georgia, the western part of the upstate of South Carolina, and most of the North Carolina mountains, albeit temporarily.
Figure 8. GOES-12 enhanced water vapor at 1515 UTC 11 February. Click on image to enlarge.
The dry slot continued to move east across the foothills and Piedmont of the Carolinas through 1800 UTC as the center of low pressure moved to a position near Columbia, South Carolina. Precipitation ended across most of the area to the west of a line from Greensboro, North Carolina, to Wadesboro, North Carolina, and Lexington, South Carolina by that time. Meanwhile, the deformation zone to the northwest of the surface low slowly reorganized as it translated east across eastern Tennessee, with the leading edge of light snow associated with this feature reaching the western edge of the North Carolina mountains seen on radar at 1800 UTC. The development of the linear MCS along the Gulf Coast area may have contributed to the reduction of precipitation across the western Carolinas and northeast Georgia by preventing moisture transport northward from the Gulf of Mexico (Mahoney and Lackmann 2005). In fact, the models overestimated the amount of precipitation by nearly a factor of two. Although the majority of the precipitation at Asheville fell as snow through 1800 UTC, the rate at which the snow fell was not fast enough for an accumulation of more than one inch due to melting. Outside the mountains, the precipitation rate was not great enough to allow cooling effects to suppress the melting level, thus the snow level remained around 2000 feet and the precipitation fell as all rain at the Greenville-Spartanburg Airport. Only a trace of snow fell at Hickory.
Observed 12-hour Precipitation (liquid equivalent) ending 1800 UTC 11 February
Greenville - Spartanburg (KGSP)
Snowfall amounts across the western Carolinas generally reflect a snow level which remained between 2000 feet and 2500 feet during the first part of the event (Fig. 9). Although a large part of the North Carolina Mountains received greater than 4 inches, much of it was limited to elevations above 3000 feet, especially the Balsams. Most of the population centers, in particular the French Broad Valley and locations such as Bryson City and Waynesville, failed to accumulate more than one inch.
Figure 9. Total snow accumulation (inches) for the Synoptic Phase of the event (0000 ? 2100 UTC 11 February). Click on image to enlarge.
The center of surface low pressure moved to a position near Fayetteville, North Carolina, at 2100 UTC. As the low began to move away, the deformation zone precipitation area skirted along the Tennessee border and moved across the northern mountains of North Carolina. After 2100 UTC, the back edge of the deformation zone lifted north of Avery County, North Carolina, and the mechanism responsible for snow falling across the mountains began to change.
(Click here to view a 9 frame java loop of radar reflectivity centered on the Morristown (KMRX) WSR-88D radar, depicting the precipitation transition across east Tennessee and western North Carolina.)
The Northwest Flow Phase
Radar imagery from the KMRX WSR-88D clearly showed a transition across eastern Tennessee and western North Carolina between 1900 UTC and 2200 UTC, as the back edge of light precipitation associated with the deformation zone lifted northeast and precipitation redeveloped over eastern Tennessee (Fig. 10). Nearly coincident with the transition of precipitation echoes, the winds across the mountains of North Carolina at 850 mb veered from southwest to northwest between 1700 UTC and 2100 UTC, after which a northwest flow continued unabated. Observations across the mountains of North Carolina showed the wind shift and coincident temperature drop during the early part of the afternoon. Click the links to see meteograms at Wayah Bald (temp, wind direction) and Bearwallow Mountain (temp, wind direction). The 850 mb wind speed strengthened to 35 kts as the surface center of low pressure continued to move away to the northeast across the coastal plain of North Carolina and Tidewater Virginia at 0000 UTC during the evening of 11 February, and eventually off the Mid-Atlantic coast by 0600 UTC, Sunday, 12 February. The ensuing cold advection flow dropped the temperature at 850mb from -5 deg to -9 deg Celsius between 2100 UTC 11 February and 1200 UTC 12 February along the Tennessee border. The mechanical forcing from the northwest winds impinging upon the higher terrain along the Tennessee border resulted in an area of light to moderate snow, particularly over the Great Smoky Mountains National Park, that persisted through about 1200 UTC.
Figure 10. Radar reflectivity (dBZ) mosaic centered on KMRX WSR-88D at (a) 1856 UTC, (b) 1956 UTC, (c) 2100 UTC, and (d) 2200 UTC, on 11 February. Click on each image to enlarge.
Cyclonic flow around the rapidly deepening surface low moving up the eastern seaboard at 1200 UTC on Sunday, 12 February, maintained the northwest winds at low levels across the mountains through the daytime hours, as seen on the 925 mb analysis. The character of the northwest flow precipitation changed again during the morning as the first area of light snow weakened and moved over the northern mountains, and new precipitation developed over eastern Tennessee. Instead of the layered appearance of radar echoes along the western slopes of the mountains that is typical of many northwest flow events, the precipitation echoes that developed over northeast Tennessee after 1200 UTC had a more cellular appearance. The upper air observation taken at 1200 UTC at Nashville, Tennessee (KBNA), showed a nearly dry adiabatic lapse rate from the surface to 925 mb that suggested the potential for convective instability if surface moisture remained sufficient (Fig. 11). In fact, the unmodified temperature sounding showed very weak amounts of Convective Available Potential Energy (CAPE), suggesting that small changes to the profile such as cold advection aloft or moisture advection near the surface would quickly increase the instability of surface air parcels.
(Click here to view a 32 frame java loop of radar reflectivity from the Morristown (KMRX) WSR-88D radar, which shows the evolution of the Northwest Flow snow across the mountains of North Carolina.)
Figure 11. Skew-T, log P diagram for the upper air observation at Nashville (KBNA) 1200 UTC 12 February. The thick red line is the temperature sounding and the dashed black line is the dewpoint sounding. Wind barbs (knots) are shown on the right. Click on image to enlarge.
The 850 mb objective analysis from the Storm Prediction Center (SPC) at 1200 UTC (Fig. 12) showed strong cold advection, indicated by the wind barbs oriented perpendicular to the isotherms across the Appalachians and Cumberland Plateau. When compared to the KBNA sounding, the continued cold advection at 850 mb suggested that boundary layer convective processes could dominate the development of clouds and precipitation across eastern Tennessee, even with minimal amounts of surface heating. It is hypothesized that strong cold air advection over relatively warm ground led to thermal instability, allowing horizontal convective rolls to develop upstream of the upslope areas along the North Carolina-Tennessee border which contributed to snow production, as in Schultz et al. (2004). The banded nature of precipitation echoes across northeast Tennessee by 1602 UTC (Fig. 13, northeast of the KMRX radar site), aligned in the direction of the northwest wind at 850 mb, agreed with the appearance of horizontal convective rolls seen on Moderate Resolution Imaging Spectroradiometer (MODIS) satellite imagery at 1602 UTC (Fig. 14). Farther south of the KMRX radar site, less distinct echoes agreed with the appearance of open cellular convection over southeast Tennessee and northwest Georgia.
Figure 12. SPC Objective Analysis at 850 mb for 1200 UTC 12 February. Wind barbs (knots) are blue, isotherms (deg. C) are shown as dashed blue lines, and geopotential heights (dm) are shown as solid dark gray lines. Click on image to enlarge.
Figure 13. Composite Reflectivity (dBZ) from the Morristown, Tennesseee (KMRX), WSR-88D at 1602 UTC (11:02 AM) 12 February. The radar site is indicated by the red plus sign. The solid yellow lines are state boundaries and the thin gray lines are county boundaries. Click on image to enlarge.
Figure 14. Terra MODIS image taken from 1602-1613 UTC 12 February scan. Brighter white shades indicate more reflective (thicker) cloud cover. Brown shades indicate bare ground. Click on image to enlarge. In fact, by the late morning hours, the Local Analysis and Prediction System (LAPS) analysis of CAPE showed values greater than 100 J kg-1 across much of eastern Tennessee (Fig. 15). The weak downslope flow off the Cumberland Plateau provided additional convective instability as solar heating kept the boundary layer relatively warm across the Great Valley of east Tennessee.
Figure 15. LAPS analysis at 1600 UTC 12 February. The left side depicts the analysis of CAPE with contours every 30 J kg-1 in yellow. The right side shows a Skew-T, log P diagram at KTYS, with the temperature and dewpoint profiles in green. Note the table of computed indices at the lower right. Click on each image to enlarge.
By the time of the MODIS image at 1919 UTC (Fig. 16), horizontal convective rolls are apparent across northeast Tennessee, nearly aligned in the direction of the northwest flow at 850 mb. The radar imagery from the KMRX WSR-88D around the time of the MODIS image also showed the appearance of horizontal convective rolls stretching northwest to southeast across northeastern Tennessee and intersecting the southern Appalachians (Fig. 17). The organization of the precipitation elements may have played an important role in the variable nature of snow accumulation across the mountains during the northwest flow phase of the event, both in terms of providing an enhancement to precipitation in locations where convective rolls intersected the mountains and providing a mechanism for convective elements to persist downstream of the initial rise of terrain on the west side of the mountains.
(Click here to view a 35 frame java loop of GOES-12 visible satellite imagery, which shows the development of horizontal convective rolls over northeast Tennessee and the evolution of low clouds moving up the west side of the mountains.)
Figure 16. As in Fig. 14, but for Aqua MODIS at the 1919-1930 UTC (2:19 - 2:30 PM) 12
February scan. Click on image to enlarge.
Figure 17. As in Fig. 13, at 1918 UTC (2:18 PM) 12 February. Click on image to enlarge.
The loss of daytime heating after 2200 UTC spelled an end to the convective organization of precipitation echoes seen on the KMRX radar. Coverage of precipitation decreased significantly by 0100 UTC 13 February and was limited mainly to a persistent band across Haywood County, North Carolina, which itself weakened by 0500 UTC. The production of light snow gradually waned during the early morning hours of Monday, 13 February, as low level moisture dwindled and northwest winds diminished, with the event essentially ending around sunrise.
The extent of the snow accumulation was revealed on satellite imagery as cloudiness decreased across the southern Appalachians on Monday morning. A Terra MODIS image from the 1644 UTC scan showed the eastern edge of the snow fall as the transition between white and brown shades near the Blue Ridge Escarpment, although clouds still obscured some of the snow pack across east Tennessee, north Georgia, and the mountains of North Carolina (Fig. 18). A later image from the 1824 UTC Aqua MODIS scan reveals the true extent of the snow cover, as clouds have completely dissipated across the southern Appalachians (Fig. 19). Note how accumulating snow was limited almost entirely to elevations above 2000 feet, with very little snow cover present across the upper Little Tennessee River valley and the middle and upper French Broad River valley.
Figure 18. As in Fig. 14, but for Terra MODIS image from the 1644-1657 UTC (11:44 AM- 1157 AM) 13 February scan. Click on image to enlarge.
Figure 19. Aqua MODIS image from the 1824-1837 UTC (1:24 - 1:37 PM) scan. Note the contrast between the white shades indicating snow cover and the brown shades near the Blue Ridge Escarpment. Click on image to enlarge.
The greatest snow accumulations from the Northwest Flow phase of the event were observed along the northwest facing slopes of the higher elevations along the Tennessee border, including reports of over three feet along the Cherohala Skyway in western Graham County and drifts of five to six feet in the parking lot at Newfound Gap (Fig. 20). Although there is a definite elevation dependency seen in the snow accumulation, many valley locations in the shadow of the high peaks near the Tennessee border, such as Robbinsville, Cherokee, Sylva, and Burnsville, still received significant amounts. The organization of precipitation noted by the horizontal convective rolls seen on satellite and radar imagery may have contributed to the downstream transport of snow from the main production area where the terrain rises quickly along the Tennessee border.
Figure 20 Snow accumulation (inches) for the Northwest Flow Phase (2100 UTC 11 February to 1200 UTC 13 February). The graphic may not indicate sharp gradients across the higher terrain. Click on image to enlarge.
Photo courtesy of Ron and Nancy Johnson at www.Tailof theDragon.com
The author wishes to thank Larry Lee (Science and Operations Officer, NWS Greer) for his assistance with locating references and providing a critical review of the manuscript. Jonathan Blaes (Information Technology Officer, NWS Raleigh) provided a list of web pages where archived weather data could be located. Rick Neal (Information Technology Officer, NWS Greer) archived the event and assisted with loading the data on the Weather Event Simulator. Last, but not least, this page would not have been possible without the help of Neil Dixon (webmaster, NWS Greer), who set up the html framework for the page and provided numerous pointers on how to accomplish most of the html coding.
Carlson, T. N., 1980: Airflow through midlatitude cyclones and the
comma cloud pattern. Mon. Wea. Rev., 108, 1498-1509.
Grumm, R. H., 2006: The Megalopolitan snowstorm of 11-12 February 2006:
Problems with uncertainty. Unpublished manuscript. 17 pp.
Mahoney, K. M., and G. M. Lackmann, 2005: The effects of organized
upstream convection on downstream precipitation. Preprints,
21st Conf. on Weather Analysis and Forecasting/17th
Conf. on Numerical Weather Prediction, Washington, D. C.,
Amer. Meteor. Soc., CD-ROM, 3.1.
Miller, J. E., 1946: Cyclogenesis in the Atlantic coastal region
of the United States. J. Meteor., 3, 31-44.
Schultz D. M., D. S. Arndt, D. J. Stensrud and J. W. Hanna, 2004:
Snowbands during the cold-air outbreak of 23 January 2003.
Mon. Wea. Rev., 132, 827-842.