A Comparison of Two Northwest Flow Snowfall Events from the 2007-2008 Winter Season
NOAA/National Weather Service
Snowfall in Mars Hill, NC on 2 January 2008. Photo provided by Charles Newcity. Used by permission.
Author's Note: The following report has not been subjected to the scientific peer review process.
During the 2007-2008 winter season, two significant northwest flow
snowfall (NWFS) events occurred across the North Carolina mountains.
The first event took place 1-2 January 2008 and the other 27-28
February 2008. Both events produced warning criteria snowfall
accumulations and were anticipated well by forecasters. This event
review will provide a brief summary and comparison of these two events.
2. Summary of the 1-2 January 2008 event
The NWFS event that took place 1-2 January 2008 occurred as a cold and
broad 500 hPa trough moved across the eastern third of the country
(Fig. 1). At the surface, a complex low pressure system developed
along the Mid-Atlantic and eastern Great Lakes by 1200 UTC on the 1st,
and moved into the Canadian Maritimes by 1200 UTC on the 2nd (Fig. 2).
During this time, the North Carolina Mountains were caught between the
surface low to the northeast and strong high pressure over the central
and southern Plains, setting up a tight pressure gradient and strong
low-level northwest flow. Snow showers began developing across the
North Carolina mountains and areas upstream during the afternoon and
evening on the 1st, and continued through the early afternoon on the
2nd (CLICK HERE FOR REGIONAL RADAR LOOP). A vertically pointing
MicroRainRadar (MRR) positioned in northern Avery County revealed even
more details about the snowfall during this event, showing that most
of the snow shower activity was confined to the time period between
1800 UTC on the 1st and 0800 UTC on the 2nd (Fig. 3).
Overall for the event, total snow accumulations ranged from a trace to
6 inches across the western Carolinas and northeast Georgia Mountains
(Fig. 4). Trace accumulations stretched as far east as Anderson and
Greenville, South Carolina as snow showers broke containment from the
North Carolina Mountains and moved downstream of the immediate upslope
areas. This event was also well anticipated by forecasters as the
potential for warning criteria snowfall accumulations was highlighted
in the Area Forecast Discussion four days prior to the beginning of
the event. In fact, a Winter Storm Watch was issued for the North
Carolina mountain counties bordering the Tennessee state line on the
afternoon of 30 December 2007. All total, the Winter Storm Watch had
a lead time of 61 hours and the subsequent Winter Storm Warning had a
lead time of 48 hours.
Fig. 1. Storm Prediction Center (SPC) objective analysis of 500 mb geopotential height, temperature, and wind at 0000 UTC 2 January 2008 (left) and 1200 UTC 2 January 2008 (right). Click on images to enlarge.
Fig. 2. Hydrometeorological Prediction Center (HPC) surface fronts and pressure analysis at 1200 UTC 1 January 2008 (left) and 1200 UTC 2 January 2008
(right). Click on images to enlarge.
Fig. 3. MicroRainRadar (MRR) data image for the time period 1800 UTC 1 January 2008 and 1759 UTC 2 January 2008. Image from Dr. Sandra Yuter's Cloud and Precipitation Processes and Patterns Group, North Carolina State University. Made available by Dr. Baker Perry. Click on image to enlarge.
Fig. 4. Map of event snowfall accumulations as reported by spotters, cooperative observers, CoCoRaHS observers, and county officials.
3. Summary of the 26-28 February 2008 event
The second NWFS event discussed here occurred during 26-28 February
2008 and developed as a cold front moved across the western Carolinas
late on the 26th (Fig. 5). The cold front and low pressure system
were associated with a sharp 500 hPa trough that moved across the
Mississippi Valley at 0000 UTC on the 27th (Fig. 6). This event was an
example of a long duration NWFS event as low-level northwesterly winds
continued across the western Carolinas through the early morning hours
of the 28th as weak high pressure built into the region from the
southwest (Fig. 7). Snow shower activity began late on the 26th and
continued over the next 24 hours, with even light snow showers continuing
through the morning of the 28th (Fig. 8).
CLICK HERE FOR A REGIONAL RADAR LOOP.
Total snowfall accumulations ranged from trace amounts to 16 inches in
western Graham County along the Cherohala Skyway (Fig. 9). Much like
the January event, this was another case where the event was well
anticipated by forecasters. A Winter Storm Watch was issued early in
the morning on the 25th with 48 hours of lead time and a Winter Storm
Warning was issued early in the morning on the 26th with 23 hours of
Fig. 5. Hydrometeorological Prediction Center (HPC) surface fronts and pressure analysis at 0000 UTC 27 February 2008 (left) and 1200 UTC 27 February 2008 (right). Click on images to enlarge.
Fig. 6. Storm Prediction Center (SPC) objective analysis of 500 mb geopotential height, temperature, and wind at 0000 UTC 27 February 2008 (left) and 1200 UTC 27 February 2008 (right). Click on images to enlarge.
Fig. 7. Hydrometeorological Prediction Center (HPC) surface fronts and pressure analysis at 0600 UTC 28 February 2008. Click on image to enlarge.
Fig. 8. MicroRainRadar data image for the time period from 0000 UTC 27 February 2008 to 2359 UTC 27 February 2008 (left) and from 0000 UTC 28 February 2008 to 2359 UTC 28 February 2008 (right). Image from Dr. Sandra Yuter's Cloud and Precipitation Processes and Patterns Group, North Carolina State University. Made available by Dr. Baker Perry. Click on images to enlarge.
Fig. 9. Map of event snowfall accumulations as reported by spotters, cooperative observers, CoCoRaHS observers, and county officials.
4. Event differences
Even though both of these winter storms were NWFS events and resulted
in significant snowfall accumulations, there were some important
differences in the way they unfolded. This section will highlight
a) Upper trough evolution
As noted above, both of these events occurred under the influence of a
deep eastern CONUS trough that contained an embedded shortwave trough
that passed across the southern Appalachians around the middle of each
event. A visual comparison of the two respective 500 hPa troughs (Figs
1, 6) revealed that the trough associated with the February event had a
much sharper trough axis as well as a much shorter wavelength. However,
the differing effects of these two troughs were not realized until local
upper air soundings from Poga Mountain in northern Avery County are
investigated. These soundings were launched every three hours during
both NWFS events.
During the 1-2 January event, the shortwave trough embedded in the
larger scale trough moved into the North Carolina Mountains between
0000 UTC and 1200 UTC (Fig. 1) on the 2nd. The effects on the lower
troposphere were readily seen comparing upper air soundings taken at
0300 UTC and 0600 UTC (Fig. 10). Between 0300 UTC and 0600 UTC, as
the trough moved through, the depth of the low-level moisture increased
and the height of the capping inversion also increased. This was also
the time period when the highest reflectivity appeared in the MRR data
(Fig. 3), indicative of the most vigorous snow shower activity.
A similar effect occured in the 26-28 February event, though to a greater
degree. In this event, the embedded shortwave trough moved into the
southern Appalachians between 0000 UTC and 1200 UTC (Fig. 6) on the 27th.
Again, the results were readily apparent in the sounding data between
0600 UTC and 0900 UTC (Fig. 11) where the depth of the moist layer and
the capping inversion was raised approximately 100 hPa. Looking at the
MRR data showed that the time period of the best snow shower activity
at Poga Mountain began around 0700 UTC which was between the times of
the two soundings.
Fig. 10. Skew-T log P diagram for upper air sounding from Flat Springs, NC at 0300 UTC 2 January 2008 (left), and 0600 UTC 2 January 2008 (right). Click on images to enlarge.
Fig. 11. Skew-T log P diagram for upper air sounding from Flat Springs, NC at 0600 UTC 27 February 2008 (left), and 0900 UTC 27 February 2008 (right). Click on images to enlarge.
b) Snow-to-liquid ratio and other event
Though these two NWFS events occurred under similar synoptic patterns,
they differed greatly with regards to specific event characteristics.
Identification of these differences was made possible through access to
a detailed set of observations gathered at Poga Mountain in the Flat
Springs area of Avery County. These data were collected as part of a
snow density study (Perry et al. 2008). The first main difference between
these two events dealt with event duration. Simply put, this was the
total time that upslope northwest flow and snow shower activity was
present. From the Poga Mountain MRR data for the January event, it was
clear that snowfall occurred during about a 12-18 hour time period.
Conversely, the February event took place over a longer time period of
approximately 36 hours. The main effect from duration was that snow
had more time to accumulate the longer an event went on.
Overall, during the 1-2 January 2008 NWFS event, storm total snowfall
at Poga Mountain was 4 inches with 0.13 inches of liquid-equivalent
precipitation. This equates to a snow-to-liquid ratio (SLR) of 30.8
for the event. Likewise, during the 26-28 February 2008 NWFS event,
storm total snowfall was 8.3 inches with 0.39 inches of liquid-equivalent
precipitation. Therefore, for the February event, the SLR was 21.3.
The higher SLR in the January event implied that the snow density was
much lower than during the February event. This was further supported
by snow density observations of 33 (kg/m^-3) for the January case and
47 (kg/m^-3) for the February case. A lesser snow density implied an
increased susceptibility of accumulating snow to blowing and drifting
which made accurate measurement of accumulations increasingly difficult.
This effect was magnified by the fact that 850 hPa winds from the
Blacksburg, Virginia upper air sounding were 37 kts during the January
event, compared to 21 knots (Fig. 12) during the February event. The
presence of stronger low-level winds in the January event certainly
could have prevented the accumulation of snowfall and the subsequent
Fig. 12. Skew-T log P diagram for upper air sounding from Blacksburg, VA (RNK) at 1200 UTC 2 January 2008 (left), and 1200 UTC 27 February 2008 (right). Images from University of Wyoming Department of Atmospheric Science (http://weather.uwyo.edu/upperair/
sounding.html). Click on images to enlarge.
5. Summary and conclusions
During the 2007-2008 winter season, two significant NWFS events affected
the North Carolina Mountains. Both events produced warning criteria
snowfall and were well anticipated by forecasters. Though these events
occurred under similar synoptic conditions, subtle atmospheric
differences caused significant variations in the storm total snowfall
accumulations realized in each event. The January event produced
widespread 3 inch amounts with isolated areas receiving 6 inches, while
the February event ended with widespread 6 inch accumulations with maximum
totals over a foot.
One of the main points that can be taken from this event review is the
importance of SLR in NWFS events. As seen in the data collected in Flat
Springs, North Carolina, SLR values of about 20:1 or higher are routinely
observed in NWFS events. Such SLR values are the result of very cold
air and a relatively shallow layer of moisture. SLR is important because
when values approach or exceed 20:1, even light amounts of precipitation
can result in multiple inches of snow. This is significant, especially
for forecasters who can use more accurate SLR guidance to portray more
representative snowfall forecasts for NWFS events.
Perry, L.B., D.K. Miller, S.E. Yuter, L.G. Lee, and S. Keighton, 2008:
Atmospheric influences on new snowfall density in the southern
Appalachians Mountains, USA. Proceedings of the 65th Eastern Snow
The upper air analysis were obtained from the Storm Prediction Center
and surface analyses were obtained from the Hydrometeorological
Prediction Center. Upper air sounding images from RNK were accessed
through the University of Wyoming Department of Atmospheric Science
web page. Regional radar mosaic loops were created using the Iowa
Environmental Mesonet web page. A special thanks to Dr. Baker Perry,
Appalachian State University, and Dr. Sandra Yuter and the Cloud and
Precipitation Processes and Patterns Group, North Carolina State University,
for the MRR data and observations from Flat Springs. Also, the upper air
soundings from Flat Springs were made availble by Dr. Doug Miller of the
University of North Carolina at Asheville (UNCA).