Snow Accumulations from Northwest Upslope Flow

by Steve Keighton, Science and Operations Officer


If you have read the previous article, then you know that we can get significant snow accumulations from several types of large scale weather patterns and storm systems, which all generally produce deep upward motion with ample moisture (and of course temperatures cold enough for snow). Horizontal wind flow near the surface forced upward by the mountains can add to this vertical motion, or create enough of it even in the absence of deeper vertical motion to generate precipitation on its own. One of the most favored flow directions for bringing in additional low-level moisture and causing enhanced upslope snowfall is from the northwest. In the wake of almost every storm system (behind the cold front and to the west or south of the low pressure center at the surface) west or northwest flow pushes in at the low levels, sometimes very dramatically with howling winds that usher in an arctic air mass from Canada. If there is sufficient low-level moisture with this flow, the upward motion created by the winds encountering the northwestern slopes of the Appalachians can squeeze out additional snowfall, sometimes for a day or two after the main storm has exited the region, and occasionally adding up to a foot or more of new snow in favored areas. At the same time, locations very nearby but on the east or southeast side of major ridges see just a few snow showers and flurries, or perhaps only passing low clouds. Figure 1 is a shaded image that shows favored upslope regions in northwest flow (windward slopes are indicated by the darker shading), compared to the downslope areas on the leeward side (indicated by lighter shading) which would not be favored for enhanced snowfall in northwest flow.



Map of the Southern Appalachians topography
Figure 1



Figure 2 shows a famous example of a major snow storm as viewed from satellite imagery beginning to move up the East Coast. This was the Blizzard of March 1993, and at the time of the image the surface low pressure center is beginning to move up through eastern Virginia , and strong northwest flow has already moved into southeast West Virginia , southwestern Virginia , and northwest North Carolina . Often these large scale storm systems have a comma-shaped appearance to the cloud shield as seen from space, and the “comma-head” region can still produce quite a bit of snowfall, even in the absence of mountainous terrain. [This is sometimes referred to as “wrap-around” precipitation and moisture, since it wraps around behind the low pressure in a counter-clockwise fashion.] on the west side of the Appalachians , the upslope flow that would be occurring in this situation (which would not be present in the Midwest for example) would enhance that snowfall even more. Furthermore, even well after the deeper moisture associated with the comma-head of the winter storm system has moved well away from the area, as long as moist northwest flow continues with cold air continually pushing into the area, accumulating snowfall often continues in favored areas.



Satellite image showing low pressure over Virginia

Figure 2



Recently, several folks in the academic community, as well as in National Weather Service offices impacted by northwest flow snowfall, have been looking at this forecast issue more closely. Some aspects of this research have been completed, while other aspects are still continuing. We’ve learned a little more about the climatology of favored areas, and are beginning to learn more about the specific impacts of air trajectories coming off of the Great Lakes , which are a few hundred miles upstream!


Climatological work by Chip Konrad at the University of Chapel Hill and Baker Perry at Appalachian State University have shown that some locations on the western edge of the Appalachians receive a significant portion of their seasonal snowfall during northwest flow, specifically from 30-40% or even a little higher in a few spots. Places like Bluefield WV , Burkes Garden VA, and Boone NC are a few examples of specific sites where a long history of data showed this. Locations such as Flat Top WV (Winterplace ski resort), Quinwood WV , Mount Rogers VA , and the ski resorts outside of Boone likely receive a similar if not higher percentage from northwest flow regimes, but they also receive more total annual snow because they are much higher in elevation too.


Past and ongoing studies strongly suggest that when air flow trajectories have come off the Great Lakes , and especially a long fetch such as north-to-south over Lake Michigan , this tends to enhance snowfall even more in northwest upslope situations. In addition, even when the more widespread northwest snowfall has begun to decrease, very narrow bands of light to moderate snow can sometimes form if the trajectory has come off the Lakes. It is believed that this trajectory supplies the lower levels of the atmosphere just enough moisture and perhaps just enough instability, that when it reaches the mountains and begins to lift upward the instability is released (in some ways like the formation of a thunderstorm but not nearly as deep or intense) and snow can form where it wouldn’t have otherwise. Figure 3 is a radar image of one of these bands of snow, which in this case actually reached east of the favored upslope areas. In some cases these bands can produce an inch or two of snow, but rarely more than that. This is in contrast to the more infamous “Lake Effect” snow bands that often form immediately downwind of the Great Lakes in western New York , Pennsylvania , and other areas around these lakes, sometimes producing local snowfall measured in feet! While the snow bands that form over the Appalachians in northwest flow would never produce those kinds of extreme accumulations they get immediately adjacent to the Great Lakes , it appears that these phenomena are closely related.


Radar image showing snow band setup

Figure 3

Reference our Forecast Area



A good example of a significant northwest flow snowfall event occurred on December 19-20, 2003 . Well after a storm system had moved up over New England, cold northwest low-level flow persisted for a couple of days behind this system, although an upper level disturbance help to enhance the depth of the moisture for a while and turn the low level flow more westerly, the resultant snowfall in Figure 4 clearly shows the dominant influences of the northwest and western ridges on the final totals. Ten to twelve inches of snow fell in far northwestern Greenbrier County West Virginia, as well as around Bluefield and Tazewell, and also in western Watauga and Ashe Counties in North Carolina, while locations just to the east received only a trace to an inch of snow (such as the Alleghany Highlands and New River Valley of Virginia).



Snowfall Map from Dec 2003

Figure 4

Reference our Forecast Area


The snowfall map from the December 2003 event is a classic example of the areas typically favored for significant snowfall in northwest flow regimes. Compare this to the areas highlighted in Figure 1 as the windward slopes. While there are certainly other ridges within the Appalachian region with northwest-facing slopes, and they may at times see some degree of enhanced snowfall or lingering snow showers during cold northwest flow, they do not typically get the significant accumulations that the areas on the far western edge of the Appalachians receive. This is because the air parcels coming from the northwest lose most of their moisture when they first experience upslope forcing. It’s like wringing out a sponge as hard as you can the first time, leaving little moisture to come out when you squeeze it a second time. The ridges are also generally higher in this western part of the Appalachians than they are along the Blue Ridge in northwest North Carolina and southwest Virginia,, and the deeper lift helps too. The folks who live near these western ridges of the Appalachians probably already know they are under that big sponge during the winter time!


Ongoing collaborated studies at our office, surrounding NWS offices, and the academic community in this region are focused on the use of fine resolution weather prediction models to simulate past events in order to determine what the primary mechanisms that produce the most significant northwest flow snowfall are. We are also planning to run a fine resolution local model operationally (hopefully but this winter) to try and generate better forecast guidance on the local variability of these events. Finally, future climatological studies will try to determine subtle differences in environmental parameters such as depth of moisture, exact wind direction, and other parameters on northwest flow snow amounts and areal distributions.