Event Summary
     National Weather Service, Raleigh NC

February 13, 2008 Winter Storm
Updated 2008/09/11

Event Headlines

...Much of central and eastern North Carolina received between a half-inch and two inches of rain during the morning hours of February 13, 2008...
...An intense shortwave trough combined with an increasingly unstable atmosphere produced another area of precipitation and embedded convection across the Western Piedmont...
...The heavier precipitation rates allowed evaporation and the melting of snow aloft to cool the vertical column and eventually the near surface temperatures which allowed snow to reach the surface...
...Snowfall rates of up to an inch per hour helped overcome warm surface temperatures, allowing a rapid accumulation of snow with numerous reports of 4 inches of snow across portions of the Triad and the Northern Piedmont...
...The limitations and details regarding the use of AMDAR aircraft soundings are demonstrated when compared to the GSO RAOB...

Event Overview

On February 12th, there was a deep upper-level trough across the eastern U.S. A surface low was lifting northeast toward the Great Lakes. A second surface low developed along an air mass boundary over the Carolinas and tracked north along a coastal warm front. Heavy rain developed over the North Carolina Coastal Plain as the low moved northward, with some locations receiving up to 2 inches of rain. A shortwave trough moved into the area on the evening of the 13th sparking showers and a few thunderstorms over Upstate South Carolina and southwestern North Carolina. Much colder and drier air was transported into the Appalachian Mountains and Western Piedmont during the evening hours. As the atmosphere cooled, the rain showers changed to snow, and snow showers, with a few embedded stronger bands developed and moved across the Triad and the Northern Piedmont. The change in weather conditions was dramatic. At 400 pm on the 13th, Winston-Salem reported cloudy skies and 50 degrees, just four hours later, it was snowing and 33 degrees. The result was a notable accumulation of snow with many locations receiving at least two inches of snow and a few locations across Person and Caswell Counties receiving as much as 4 to 5 inches of snow.

Snow Accumulation Map

Snow accumulation map

Liquid Equivalent Precipitation Map

Liquid equivalent precipitation map

Meteorological Overview

The February 13 event was preceded by a progressive large amplitude upper-level wave over the central U.S. and a surface low lifting northeast toward the Great Lakes. A warm front extended from the low center south across the Appalachian Mountains. A sounding from Greensboro (GSO) on the evening of the 12th (00 UTC on February 13) revealed a cool, shallow layer of air near the surface, which impeded the progression of the warm front northward. Twelve hours later, the morning sounding from GSO (12 UTC on February 13) shows the cold, shallow air mass lingering over the area. The weather pattern in which a cold, shallow air mass lingers over the Piedmont region is often called cold air damming. The damming region is the area in which the cold, shallow, stable air mass is located. Cold air damming typically forms along the lee side of the Appalachians, either through the funneling of cold dense air down the East Coast by an area of high pressure to the north, or through diabatic processes that lead to the formation of a cold stable air mass in situ. This creates a scenario where relatively cooler, more stable air resides over western NC, while warmer, and often less stable air remains east of the damming region. Between these two differing air masses lies a front, often referred to as the wedge front. The location of the wedge front varies from event to event, and the front may migrate inland or retreat towards the coast. Such conditions often result in tremendous temperature gradients across NC, as was the case during the overnight hours on the 12th. Temperatures across much of western NC remained in the upper 30's to lower 40's, while temperatures just 80 miles east rose into the lower 60's.

Early on the morning of the 13th, a surface low developed along the wedge front over Upstate South Carolina and tracked north along the coastal warm front. As the low moved north, a band of heavy rain developed and tracked north over the North Carolina Coastal Plain, where up to 2 inches of rain fell. Warm air was eventually able to erode the cold near-surface layer and the wedge front. However, temperatures were only able to recover into the mid to upper 40's across the western and northern Piedmont. Meanwhile, the original cold front and drier air advecting into the region from the Midwest remained west of the Appalachians. The combination of cooler high temperatures on the 13th and dry air upstream would prove significant later that evening as the upper-level wave moved into the area and more precipitation developed.

While the main storm system exited the area to the northeast, a shortwave trough was slowly pivoting around the base of the longwave trough. The lighter colors over western NC in the water vapor image from 21 UTC on the 13th represent a small area of convection that developed just ahead of the shortwave.

This area of showers and thunderstorms formed in response to the increasing mid-level lapse rates associated with the shortwave. At around 22 UTC, the area of precipitation was fairly small and just southwest of the NWS Raleigh County Warning Area near Charlotte. Over the next couple of hours, the precipitation expanded and moved into the Western Piedmont region by 00 UTC on the 14th. There were around a half dozen lightning strikes during each hour in this period, primarily across the southern portion of the precipitation (one hour lightning strikes ending at 23 UTC | one hour lightning strikes ending at 00 UTC | one hour lightning strikes ending at 01 UTC).

At the same time, drier air was beginning to move into the Northwest Piedmont. Dew point temperatures began to plummet, and with temperatures only in the lower 40's, evaporation of the precipitation caused significant cooling of the low-level air. Such cooling is known as evaporational cooling, where the process of evaporating a rain drop removes heat from the surrounding air, and subsequently cools the surrounding air. This cooling, and moistening of the ambient air will continue until the air becomes saturated, at which point the precipitation ceases to evaporate. The temperature at which this occurs is known as the wet-bulb temperature. The 00 UTC sounding from GSO shows dry low-level air that would promote evaporation and cooling as the precipitation overspread the area. Given the temperatures and dew points within the dry layer, the evaporational cooling would bring most of the layer to wet-bulb temperatures near freezing. After the layer was saturated, additional cooling was generated from the melting of snow aloft. As snow melts, it removes heat from the atmosphere, although it has a less significant impact then evaporation, and with moderate to heavy precipitation rates, the role of melting becomes mores significant.

As the precipitation moved into the area, the change over to light snow occurred at around 23 UTC on the 13th. The first report of snow came from Hickory, NC (see observation below).
KHKY METAR 132347Z 36007KT 2 1/2SM -SN BR FEW005 SCT009 OVC050 01/00 A2969 RMK AO2 RAE02SNB02 P0006

The small area of precipitation expanded as it spread northwestward from Lexington, through the Triad area, to the North Carolina/Virginia border. A few bands of moderate snow developed at the leading edge of the precipitation, with a southwest to northeast orientation. One band in particular developed over northwest Person County, along the NC/VA border. With the entire system moving to the northeast, this band gave northwest Person County an hour of moderate snow even before the main area of precipitation arrived.

At around 01 UTC, the Satellite Analysis Branch at NESDIS, issued a Satellite Precipitation Estimates discussion (SPENES). The full discussion and imagery is available. The discussion noted that the vorticity center near Hickory had been strengthening. The precipitation producing clouds associated with the vorticity center had been rapidly expanding with increasing lift ahead of the system producing an increase in showers across central NC. The discussion noted some concern for the spread of the precipitation further east and northeast and that intensities will likely pick up with concern that moderate to heavy snow.

Snowfall rates of up to an inch per hour were reported over much of the Northern Piedmont over the next few hours. The enhanced snowfall rates were aided, in part, by strong forcing associated with the mid-level shortwave and a localized area of strong upper-level divergence. While surface temperatures were relatively warm and the snow initially melted as it reached the surface, the intense snowfall rates quickly overwhelmed the surface temperatures, allowing snow to first stick on grassy surfaces, and eventually on roads. Although no major traffic problems were reported during the event, some less traveled roads in portions of Forsyth, Guilford, and Person Counties were quickly covered with snow.

The precipitation dissipated fairly quickly after midnight with just a few lingering flurries persisting after 100 am.

A radar loop of KRAX reflectivity imagery during the critical portion of the event with images from every 15 minutes between 13/2228 UTC (528 PM EDT 02/13/2008) through 14/0755 UTC (255 AM EST 02/14/2008) is available.
Note - this loop includes 39 frames

Radar (KRAX) Loop of Snow in Western and Northern Piedmont - click to enlarge

Satellite Imagery

An impressive shortwave trough can be seen approaching the southern Appalachians in the water vapor satellite image below from 2215 UTC on February 13, 2008. A Java Loop water vapor satellite imagery from 0015 UTC on February 12 through 1215 UTC on February 14, 2008 is available. The loop captures the impressive trough as it moves across the lower Mississippi Valley and then northeast across the Carolinas.

Water vapor satellite image from 2215 UTC on February 13, 2008

Southeast Regional Radar Imagery

Regional radar imagery shows the evolution of the event across the Southeast. Showers developed from eastern NC all the way to southern Florida as the storm developed and lifted to the northeast. These showers can been seen moving quickly to the northeast throughout the radar loop. A second area of precipitation, which eventually transitioned to snow, developed over southwestern NC later in the event and moved much more slowly to the north, northeast. Two different types of forcing can be seen in the movement of the two precipitation regions:

  • First, the frontal forcing and flow of the southwesterly low-level jet along the coast.
  • Later the upper-level forcing of the slowly rotating shortwave trough over central North Carolina.

    A Java Loop of Southeast Regional Radar Imagery from 1558 UTC on February 13 through 808 UTC on February 14, 2008 is available.

    Southeast Regional Radar 
Reflectivity loop from 1558 UTC on February 13, 2008 through 808 UTC on February 14, 2008

  • Enhanced Snowfall Rates

    One of the most important aspects of this event was the formation of mesoscale bands and enhanced snowfall rates. Although upper-air observations from the evening of the 13th took place just before the event unfolded, there is evidence of upper-level forcing that helped induce mesoscale bands of snow and invigorate snowfall rates. As the shortwave pivoted through the state and deepened, the movement of the precipitation region transitioned from a northeastward track to more eastward.. The rotation seen in water vapor imagery suggests the shortwave deepened into a closed low, which also helped to steepen mid-level lapse rates and increase lift. Weak banding can be seen in radar reflectivities as the upper-level low as it moved across North Carolina. In particular, bands moved over the Greensboro area and along the NC/VA border.

    Radar image from 0213 UTC on February 13

    Upper-air sounding from Greensboro at 00 UTC on February 14 - click to enlarge Strong lift was also necessary to allow the microphysical processes necessary for snow crystal development to occur. The upper-air sounding from Greensboro at 00 UTC on the 14th, only an hour before precipitation began to overspread the area, is shown to the right.

    Although saturation of the low-levels would have caused cooling to subfreezing temperatures, the region between -10C and -20C was not completely saturated. The formation and growth of dendrites occurs within this layer, which is essential for snow. In order to saturate this layer, in the absence of moisture advection, significant lift was needed, and has been shown, occurred with the passage of the upper-level wave.

    The short table below provides an estimate of the potential of ice crystal initiation in clouds based on temperature.

    0 deg C - no initiation
    -4 deg C - no initiation
    -10 deg C - 60% chance of ice
    -12 deg C - 70% chance of ice
    -15 deg C - 90% chance of ice
    -20 deg C - 100% chance of ice

    AWIPS color curve used to estimate ice presence in clouds - click to enlarge AWIPS can be used to highlight the liquid-to-ice temperature range in cloud tops. The table to the right gives an example of an IR channel color scheme used to highlight the liquid to ice temperature range. Based on the cloud top temperature, and the color curve to the right, forecasters can asses whether ice is present in the cloud via cloud-top sampling. There are limitations with this approach, primarily in that high clouds can obscure the lower and mid level clouds in which the precipitation is being generated from. This technique can also assist in seeder-feeder assessment.

    Based on cloud-top temperature, and the color regimes in the image to the right, the presence of ice in cloud tops can be determined through the use of a special color curve in AWIPS. The infrared satellite image from 01 UTC on February 14 across North Carolina and Virginia is shown below. The image shows the cloud top temperatures with the color curve described above along with the 01 UTC METAR reports. The precipitation was changing to snow across the Triad region between 00 and 02 UTC as the precipitation intensified and the cloud tops cooled. The evolution of the cooling cloud tops and changing precipitation type at the surface can be seen in the IR satellite imagery with the AWIPS color curve to highlight the presence of ice in clouds: 21 UTC | 22 UTC | 23 UTC | 00 UTC | 01 UTC | 02 UTC | 03 UTC | 04 UTC

    AWIPS Sat IR image from 01Z on February 14, 2008

    Snow Cover and Rapid Snow Melt

    A relatively modest portion of northern North Carolina received measurable snow from this event with snow accumulations in excess of two inches reported across at least portions of nearly 10 counties. A few locations across Person and Caswell Counties received as much as 4 to 5 inches of snow.

    The image below from the CIMSS Satellite Blog is a MODIS true color satellite image viewed via Google Earth. The satellite image clearly shows the snow cover across northern North Carolina and southern Virginia. The snow cover stretches from the Triad region including Greensboro northeastward to just southwest of Richmond Virginia.

    MODIS true color satellite image of snow cover

    The 9 hour satellite loop below from the CIMSS Satellite Blog shows the remarkable, rapid snow melt and nearly complete disappearance of the snow cover across northern North Carolina and southern Virginia on February 14 (click on the image to open a larger picture in a new window).

    Morning temperatures on February 14 in the area where snow fell the previous night were generally in the lower to mid 20s while locations without snow cover had morning temperatures in the upper 20s to around 30 degrees. Late afternoon temperatures reached the upper 40s to lower 50s in the locations without morning snow cover while the snow cover locations had cooler afternoon temperatures, only in the mid 40s. Dew point temperatures in the snow cover area reached the lower to mid 20s during the afternoon hours.

    Much of the snow fell during the evening hours of the 13th with surface temperatures near freezing. This resulted in a rather wet snow with a relatively high snow to liquid ratio. Temperatures fell later during the night when the skies cleared but they recovered somewhat after sunrise. Soil temperatures on the 13th and 14th ranged from around 40 during the morning hours and reaching the mid to upper 40s during the late afternoon. This suggests that the snow pack temperatures were near freezing, in a "ripe" condition that would support rapid snow melt.

    Visible satellite image of snow cover with surface observations - click to enlarge

    Typically, rapid snow melt is favored during situations when there is a large turbulent transfer which occurs under windy conditions with temperatures and dew point temperature above freezing. The rate of snowmelt increases as humidity increases, due to the latent heat released when water vapor condenses on the snow when air temperature is above freezing. In comparison, the heat supplied by rainfall is usually minor, unless a warm rainfall of long duration occurs. An energy diagram from COMET shows the various contributions made to the atmosphere/snow pack energy exchange.

    Once the temperature of the snow pack reaches zero degrees Celsius throughout, the snow pack becomes ripe, and liquid water starts forming within the snow pack. When the liquid water exceeds a threshold (about 15 percent of total snow pack water equivalent), snowmelt begins. Even though the winds were generally light (around 10 MPH or so) and the dew points were subfreezing, the snow melted quickly as temperatures warmed into the 40s.

    AMDAR Aircraft Soundings

    There were just a few AMDAR aircraft soundings in Greensboro during the event. AMDAR is an acronym for Aircraft Meteorological Data and Reporting (AMDAR) which is an international effort within the World Meteorological Organization to coordinate the collection of environmental observations from commercial aircraft. In the United States, we often refer to the Meteorological Data Collection and Reporting System (MDCRS) which is a private/public partnership facilitating the collection of atmospheric measurements from commercial aircraft to improve aviation safety.

    AMDAR data is very useful for short term forecasting situations where conditions are changing rapidly. Regarding winter weather events, AMDAR data can provide forecasters with the height of the freezing level, the presence of elevated warm layers, indications of thermal advection and dry layers. All of these are necessary for accurate precipitation type forecasts. The availability of this upper air data at times and locations where RAOB data may be lacking is invaluable.

    An AMDAR sounding at GSO from 2334 UTC on the 13th shows a deep near freezing isothermal layer from around 900 to 800 mb. This isothermal layer suggests the presence of a melting layer where snow aloft melted and cooled a layer that was once above freezing to near freezing. Once the temperature in that layer is near freezing, no additional cooling will take place. In addition, note the surface based warm layer with a relatively steep lapse rate. The warm layer extends up to around 900 mb and is about 2,000 feet deep.

    About 20 minutes earlier, at 2315 UTC, the 00 UTC RAOB was released from GSO. Even though the RAOB is referred to as the 00 UTC RAOB, they are usually released a little after 23 UTC since it often takes more than 90 minutes for a complete sounding (the release time for this RAOB was confirmed with the text data provided by GSD). The 00 UTC RAOB actually shows more of a pronounced warm nose between 850 and 900 mb.

    A comparison image of the AMDAR (temperature data in light blue) and the RAOB (temperature data in red) taken about 20 minutes apart shows the differences between the two datasets. A larger and zoomed in version of this comparison is also available. It is difficult to determine the reason for the differences between the two datasets but there are several possibilities.
  • The instrumentation on the aircraft and the radiosonde are different
  • The aircraft was descending while the RAOB was ascending
  • The instruments were sampling different parts of the environment around the airport. Because of the lower level environmental winds, the ROAB was blown a mile south-southeastward from the release site by the time it reached 800 mb (139 degrees at 1 mile). While the aircraft is traveling at a much greater speed (typically over 100 kts) and from a different direction. The aircraft was nearly 20 miles west of GSO when it dropped below 800 mb (269 degrees at 19 miles).
  • The near airport environment varied greatly as showers were developing and moving across the region immediately around the airport (radar loop from 2246 UTC through 2350 UTC with the Piedmont-Triad International Airport (GSO) noted by the red dot west of Greensboro). As discussed earlier, the diabatic contributions of precipitation (evaporation and melting) would have a notable impact on temperatures in the sounding.

    Another AMDAR sounding at GSO from 0450 UTC on the 14th shows a much different sounding. The atmosphere had cooled considerably and was entirely subfreezing. The precipitation which was falling as snow across the central and eastern portions of the Piedmont had moved away from the Triad.

    Comparison of 2315 UTC RAOB and 2334 UTC AMDAR data and report locations relative to the KGSO airport

    Comparison of RAOB and AMDAR data and report locations relative to the KGSO airport - click to enlarge

  • CoCoRaHS Observer Network

    CoCoRaHS is a grassroots volunteer network of weather observers working together to measure and map precipitation (rain, snow, and hail) in their local communities. The snow accumulation maps from the CoCoRaHS web site (shown below) were a great resource for WFO RAH. For more information, visit the CoCoRaHS web site at www.cocorahs.org.

    Since the addition of CoCoRaHS to the observational network in Central North Carolina, it has become an invaluable source of observations, especially in areas where population and regular weather reports are sparse. The map below is an example of how the added reports provide greater detail of the snowfall on February 14, 2008. One can very easily separate areas of higher snowfall totals from areas of lesser totals.

    Snow Accumulation Totals

    click to enlarge

    Archived Text Data from the Winter Storm

    Select the desired product along with the date and click "Get Archive Data."
    Date and time should be selected based on issuance time in GMT (Greenwich Mean Time which equals EST time + 5 hours).

    Product ID information for the most frequently used products...

    RDUAFDRAH - Area Forecast Discussion
    RDUAFMRAH - Area Forecast Matrices
    RDUHWORAH - Hazardous Weather Outlook
    RDUNOWRAH - Short Term Forecast
    RDUPFMRAH - Point Forecast Matrices
    RDUPNSRAH - Public Information Statements (snow/ice reports among other items.)
    RDUWRKDRT - Soil Temperature Data from the NC State Climate Office
    RDUWSWRAH - Winter Storm Watch/Warning/Advisory
    RDUZFPRAH - Zone Forecast Products


    Selected Photographs of the Winter Weather Event

    Photos courtesy of Russell Bullock and Peggy Johnson.
    (Click the image to enlarge)

    Snowy scene in Efland, Orange County - photo courtesy of Russell Bullock - Click to enlarge           Wet snow sticks to trees in Efland, Orange County - photo courtesy of Russell Bullock - Click to enlarge           3 inches of snow in Efland, Orange County - photo courtesy of Russell Bullock - Click to enlarge

    Snow covered deck in Efland, Orange County - photo courtesy of Russell Bullock - Click to enlarge           Snow covered deck in Efland, Orange County - photo courtesy of Russell Bullock - Click to enlarge           Snow covered tress in Gibsonville, Guilford County - photo courtesy of Peggy Johnson - Click to enlarge

    Final Thoughts

    While model guidance in the days leading up to this event hinted at a chance of snow, anticipating snowfall rates of more than one inch per hour was difficult. Moisture is typically the limiting factor in events like this, either the absence of moisture in the snow growth layer, or the advection of dry air at the surface. In this case, however, vigorous lift was able to overcome a marginal amount of moisture. Instead of completely drying out the near surface air, the advection of dry air, combined with intense precipitation rates, helped to bring surface temperatures to near or below freezing.

    The significant precipitation rates also allowed the role of evaporation and then melting to be a significant factor in the eventual changeover of rain to snow. Contributions from diabatic processes are often key components of the processes driving precipitation type during winter storms.

    Forecasters used NCDOT and other highway traffic cameras to supplement traditional observations during the storm and become aware of the rapid accumulation of snow. NCDOT traffic camera pictures (Example 1 | Example 2 | Example 3)

    This event highlights the need for forecasters to closely examine and understand the details of AMDAR aircraft soundings. These soundings are dependent upon the aircraft flight track and significant differences in the atmosphere on the micro-scale can have a big impact on the data the AMDAR sounding provides.

    Forecasters during the evening shift had to overcome a data outage earlier in the day which reduced the continuity of the meteorological analysis.

    This event demonstrates the need to issue Winter Weather Advisories during rapidly changing events, even when conditions exceed the current forecast. When light snow is expected to continue for an hour or two before ending in areas where the advisory criteria has been met, it is still a good practice to issue the advisory even though the heaviest snow has ended.


    Many of the images and graphics used in this review were provided by parties outside of WFO RAH. The surface analysis graphic was obtained from the Hydrometeorological Prediction Center. The upper air analysis images were obtained from the College of DuPage and the University of Wyoming. GOES satellite data was obtained from National Environmental Satellite, Data, and Information Service. Precipitation maps were obtained from the Southeast River Forecast Center. Radar imagery was obtained from the National Weather Service web site. MODIS satellite imagery and the snow melt loop provided by the CIMSS Satellite Blog. Some of the ice micro physics information provided by Dan Baumgardt's Wintertime Cloud Microphysics Review. AMDAR aircraft sounding data was obtained from the Earth System Research Laboratory - Global Systems Division. CoCoRaHS maps were provided by the CoCoRaHS organization. Traffic camera pictures were obtained from the NCDOT. Photos are courtesy of Russell Bullock and Peggy Johnson.

    Case Study Team

    Barrett Smith
    Jonathan Blaes

    For questions regarding the web site, please contact Jonathan Blaes.

  • NWS Disclaimer.