An Examination of Lake Enhanced Snow across Central NC on 22 January 2014

by Michael Strickler

A band of moderate snow developed over the far northeast Piedmont and northern and central Coastal Plain of central NC on the morning of 1/22/2014. The intensity of the snow band increased such that the visibility at KRWI was reduced to between 1SM and 2SM between 13Z and 16Z (Table 1). The snow band was particularly interesting in that it developed after the parent cyclone and associated moisture and forcing for ascent had moved well to the northeast of NC (Fig. 1), and consequently was a surprising and rather unexpected development for forecasters. In fact, the accumulations between 12Z and 18Z ranged from around a half inch to as much as one inch of snow, locally satisfying winter weather advisory criteria.

1Figure 1. 1500Z WPC surface analysis.

It appeared that the band may have been influenced by Kerr Lake, located near the VA/NC border, since there was a plume-like feature in both satellite and radar data, which seemed to emanate from the lake (Fig. 2).

2Figure 2. METAR and dual-imaged visibile satellite and regional 0.5 degree radar imagery, valid at 1345Z – 1400Z.

We remembered at least a couple of similar events in the past, where it appeared that the lakes in southern VA and northern NC were contributing factors to the development of small snow bands downwind of the lakes, well after the parent cyclone had passed.  See the following links for details for those cases: March 2, 2009 and January 23, 2003. An additional horizontal convective roll cloud band originated over, and developed downwind of, Falls Lake in northern Wake County, evident in visible satellite imagery in Fig. 3 and the photo in Fig. 4.

fig.3Figure 3. METAR and dual-imaged visibile satellite and regional 0.5 degree radar imagery, valid at 1515Z – 1518Z.

photo 2Figure 4. Photograph of the lake band looking north-northeast from NWS Raleigh, located in central Wake County.

The development of the bands seemed to be the result of a unique junction of processes and properties on a range of scales in both time and space, such that without any one of them, the bands likely would not have existed. The first of these processes, and perhaps the most important one, was the meso-gamma scale sensible and latent surface heat and moisture flux from the respective lakes, whose temperatures were in the middle 40s, into the overlying arctic air above.  The temperature differential between the lakes and the overlying 850 hPa temperatures was 18-20 C, which contributed to significant low level destabilization, on the order of several hundred J/kg of CAPE, which can be seen in the RAP BUFR soundings shown in Fig 5 and 6. Since the lake band development preceded any additional cloud street HCR development that occurred with diabatic diurnal heating of the boundary layer during the late morning to midday hours (Fig. 7), it is apparent that the presence of the lakes and their associated influence were vital to the band development.

lake11ZFigure 5. KRWI RAP BUFR sounding, 00 hour forecast valid at 1100Z.

lake16ZFigure 6. KRWI RAP BUFR sounding, 05 hour forecast valid at 1100Z.

streets2Figure 7. METAR and visible satellite imagery, valid at 1500Z – 1515Z.

Another contributing influence for the band development was residual low level, meso-alpha to meso-gamma-scale moisture on the southwest flank of the departing parent cyclone, as seen in Fig 8. This nearly saturated air at 925 hPa, possibly a combination of upstream Great Lakes moisture and an orphaned portion of the deformation head attendant to the departing cyclone, produced a suitable environment/primed the atmosphere for condensation necessary to produce the cloud/snow bands.

low_clouds_RHFigure 8. RAP-analyzed 925 hPa RH and 11-3.9 micron satellite overlay, valid at 12Z.

Processes on the progressively larger scale also favored vertical motion necessary for the development of the bands. A series of low amplitude impulses/initially shear vorticity-dominated shortwave troughs on the back/west side of the mean trough axis aloft (Fig. 9), migrated through and briefly amplified into the trough base/across the Southern-Central Appalachians (Fig. 10), before the mean trough axis lifted up and away from the Middle Atlantic coast. Associated QG forcing for ascent/differential cyclonic vorticity advection shown in Fig. 11, was the larger scale process that acted upon and deepened the underlying moisture/instability, and which consequently resulted in the broader area of snow that developed to the south of the lake-induced band, shown in Fig. 12.

fig9Figure 9. Water vapor satellite imagery and RAP-analyzed 400-200 hPa vorticity, valid at 0900Z – 0915Z.

fig10Figure 10. Water vapor satellite imagery and RAP-analyzed 400-200 hPa vorticity, valid at 14Z.

vortFigure 11. 500 hPa height and vorticity, 700-500 hPa differential vorticity advection, and radar, valid at 14Z.

12Figure 12. METAR, 850 hPa VWP data, and RAP-analyzed 850 hPa temperatures, valid at 13Z.

Table 1. KRWI METAR data valid 1153-1853Z. 

METAR KRWI 221153Z AUTO 01006KT 9SM -SN OVC025 M06/M08 A2995 RMK AO2
     SLP143 P0000 60000 70012 T10561083 11039 21056 53024 $
METAR KRWI 221253Z AUTO 35005KT 8SM -SN BKN028 OVC033 M06/M08 A2998   RMK AO2 SLP152 P0000 T10561083
SPECI KRWI 221310Z AUTO 34005KT 5SM -SN SCT026 OVC032 M06/M08 A2998 RMK AO2 P0000 T10561083
SPECI KRWI 221337Z AUTO 35007KT 2SM -SN OVC032 M06/M08 A2999 RMK AO2
     P0000 T10561083
SPECI KRWI 221339Z AUTO 34007KT 1 3/4SM -SN BKN028 OVC034 M06/M08 A2999
     RMK AO2 P0000 T10561083
METAR KRWI 221353Z AUTO 34007KT 1 1/4SM -SN BKN024 OVC031 M06/M08 A3000 RMK AO2 SLP159 P0000 T10561083
METAR KRWI 221453Z AUTO 34005KT 1 3/4SM -SN FEW013 SCT017 OVC027 M06/M08 A3003 RMK AO2 SLP171 P0000 60000 T10561083 53018
SPECI KRWI 221505Z AUTO 01008KT 1 3/4SM -SN BKN011 OVC030 M06/M08 A3003
     RMK AO2 P0000 T10561083
SPECI KRWI 221516Z AUTO 01007KT 2SM -SN BKN013 OVC021 M05/M08 A3004 RMK
     AO2 P0000 T10501083
SPECI KRWI 221524Z AUTO 35006KT 1 3/4SM -SN SCT013 OVC021 M05/M08 A3004
     RMK AO2 P0000 T10501078
METAR KRWI 221553Z AUTO 34007KT 1SM -SN FEW013 OVC019 M06/M08 A3004 RMK AO2 SLP176 P0001 T10561083
SPECI KRWI 221612Z AUTO 33005KT 1 1/4SM -SN FEW009 BKN013 OVC019 M06/M08 A3005 RMK AO2 P0000 T10561083
SPECI KRWI 221617Z AUTO 33006KT 2 1/2SM -SN SCT013 BKN019 M05/M08 A3004
     RMK AO2 P0000 T10501078
SPECI KRWI 221619Z AUTO 33005KT 4SM -SN FEW007 SCT015 BKN022 M05/M08
     A3004 RMK AO2 P0000 T10501083
SPECI KRWI 221628Z AUTO 33004KT 9SM -SN SCT015 SCT022 M05/M08 A3004 RMK
     AO2 P0000 T10501078
METAR KRWI 221653Z AUTO 29005KT 9SM -SN FEW025 BKN036 M04/M09 A3003 RMK
     AO2 SLP172 P0000 T10441094
METAR KRWI 221753Z AUTO 33010KT 9SM -SN SCT040 M03/M12 A3002 RMK AO2
     SLP166 P0000 60001 T10331117 11033 21056 58004
     SLP159 P0000 T10281117

This entry was posted in Winter Weather. Bookmark the permalink.

5 Responses to An Examination of Lake Enhanced Snow across Central NC on 22 January 2014

  1. Matt Parker says:

    Very interesting case! Given that the scale of the lake here is quite small, this seems to break some of the rules about “fetch” that one learns for more traditional Great Lakes LES. I wonder if there is some way to account for the magnitude of the expected surface fluxes in order to determine the minimal lake size needed to produce such a snow band on a given day?

  2. Jonathan Blaes @ WFO RAH says:

    That’s a good comment and question. We have had some lively debate in our office about the importance of the lakes on the snow that occurred that morning. But several of us feel that while the environment that day was nearly supportive of snow showers (low level instability, residual low-level moisture, some broader Q/G forcing) the lakes provided the little kick that was needed.

    The Schultz et al. (2004) article noted below mentions other locations with lake effect snow that have fetches far less than the Bluestein value of 80+km including an example in section 3 of lake enhanced snow originating from Lake Kentucky.

    The complex shoreline and layout of Kerr would appear to make any calculations of fluxes difficult. We’d be interested in learning more about finding some lower size limits as well.

    Schultz, David M., Derek S. Arndt, David J. Stensrud, Jay W. Hanna, 2004: Snowbands during the Cold-Air Outbreak of 23 January 2003. Mon. Wea. Rev., 132, 827–842.;2


  3. bvincentnws says:

    Here’s some of the lively debate JB was talking about:)

    I have serious doubts with regard to whether or not Kerr lake played any role in the development of the aforementioned snow band, let alone that the lake was one of several necessary components. The mean low-level flow on the morning of Jan 22 was from the NW and would dictate a NW-SE orientation regardless of any influence from the lake. Similarly, the fact that the band appeared to develop just downstream of Kerr lake can be explained by conventional (though perhaps less interesting) factors.

    It seems more likely that steep low-level lapse rates and shallow instability that morning were related to:

    1) the presence of a higher theta-e environment downstream of the lake (per surface observations in figures 2,3,7) i.e. the arctic airmass had yet to fully advect into central/eastern NC at the lowest levels
    2) strengthening surface insolation between 12-15Z
    3) strong cold advection aloft at 2500-5000 ft AGL (925-850 mb)

    Given a favorable thermodynamic environment, low-level FGEN (per SPC mesoanalysis) observed near/in that region at the time could have driven the development of the band. Additionally, is it reasonable to assume that a narrow lake with a fetch of less than 10 miles can significantly alter the thermodynamic environment 25-75 miles downstream in the presence of strong CAA on a much larger scale? What about entrainment/mixing? Why is it a good assertion that the lake altered the thermodynamic environment, especially when differential thermal advection and diurnal heating can reasonably explain the presence of steep low-level lapse rates and shallow instability?

    It could be argued that a similar effect has been observed with other small bodies of water such as the Finger Lakes or Lake Tahoe. However, it’s difficult to make a direct comparison between those lakes and Kerr lake.

    The finger lakes have a maximum fetch of 30-40 miles and are far deeper/have a far greater volume than Kerr lake, in addition to being located at a much higher latitude where the overlying airmass can be significantly colder. Similar issues arise when trying to compare with Lake Tahoe, which is twice as long (~20 miles) and situated in complex topography at an elevation of ~6200 ft agl.

    I guess I’m thinking along the lines of ‘occam’s razor’ in that all things being equal, the simplest explanation tends to be the right one. The burden of proof increases the more atypical/exotic the explanation. That being the case, l would like to see more convincing evidence.

    -Brandon V.

  4. This was definitely an interesting case. We noticed it here and had some discussion as well, but I doubt it was as lively since it did not directly impact our area. Visible satellite data and radar data seemed to indicate Kerr Lake as a point source. As noted in the original post, there were several synoptic scale and mesoscale features that were supportive of snow showers. However, upstream moisture was very limited due to the arctic airmass overspreading the region.

    I formerly was a forecaster at the Glasgow, Montana WFO. Fort Peck Lake, a long narrow reservoir on the Missouri River, was centrally located in our forecast area. We would typically get ‘lake effects’ in the form of fog during the late fall and early winter (before the lake froze) when arctic airmasses would begin to overspread northeast Montana. This had to occur in light flow regimes given the limited fetch of the lake. The greatest operational impact would occur when surface high pressure would move to the east over North Dakota. This would maximize fetch in a southeast flow regime, and in some cases advect the fog nearly 20 miles into the town of Glasgow.

    On rare occasions, lake effect snow showers would develop. The best case I experienced was in October of 2009. An early season arctic airmass arrived October 9 and persisted over the region for several days. The high settled over the area on October 11 and moved to the east on October 12. During the early morning hours of October 12 (if I recall primarily 09-12z) a lake induced breeze developed as evidenced by a mesonet observation placed on an island by the NWS office. A narrow localized lake effect plume developed as a result of this and produced 7 inches of snow immediately downstream of the dam. It is difficult to estimate the fetch from this case, but it could be as low 20 miles and as high as 40 miles.

    The October 12 12z sounding from GGW revealed a well mixed layer from 850-700mb with nearly dry adiabatic lapse rates underneath a pronounced subsidence inversion. The USACE estimated the lake temperate around 10C, and the observed temperature at 850mb was -11C with -23C at 700mb. It is safe to assume very steep lapse rates (exceeding dry adiabatic) were present immediately above the lake surface resulting in significant low-level instability.

    This case is not directly analogous to the Kerr Dam case, as the Fort Peck case was more pure lake effect versus lake enhanced. However, it does highlight that smaller bodies of water with limited fetch do contribute moisture on the mesoscale. Based on my experience with Fort Peck Lake and observed data from January 22 I do think Kerr Lake had some responsibility with the band of snow, although it is apparent several factors need to come together for this to happen. Of note, Fort Peck Lake is considerably larger then Kerr Lake, but does have a similar shape (long and narrow with a complex shoreline).

    Andrew Zimmerman
    NWS Wakefield, VA

    • bvincentnws says:

      Andrew et al,

      The approach I often take in debates can run the risk of sounding blunt, defensive, condescending, rude, or even ‘like a lecture’, though that’s certainly not the way I intend it to sound, nor the manner in which I want it to come across. I’ve heard all of these at some point, though most frequently (and most recently) from my wife:) Though she may be my harshest critic, her sharp intellect and unique insight have been (and continue to be) invaluable with regard to accepting and understanding such criticism and in improving my attempts at discourse, scientific or otherwise. Simply stating that I sincerely value her opinion and suggestions would be an understatement. I bring this up only because I feel that it may aid the reader(s) in interpreting the following comments solely in the context of scientific discourse, the type of discourse that counsels us to carry alternative hypotheses to see which best fit the facts, that urges on us a delicate balance between no-holds-barred openness to new ideas, however heretical, and the most rigorous skeptical scrutiny of everything from new ideas to established wisdom.

      I’m not familiar with the Montana case that you described in your response, Andrew, though I’m certainly curious about it and interested in taking a more detailed look. Can you point me to any case studies or peer-reviewed journal articles related to the case you mentioned?

      Though interesting, your reply does not address any of the specific questions or concerns I noted in my initial reply. Could you respond to the alternative hypothesis I posited in my previous post, the one which attempts to explain the existence of the snow band in terms of more conventional factors? For example, is there a flaw in my reasoning? If so, why? Am I omitting any pertinent information or misinterpreting the data? Is there anything I need to clarify with regard to my reasoning?

      Let me be clear, I am not suggesting that there was absolutely no flux of moisture from the lake into the surrounding environment, it seems totally logical that there would be. However, the same could be said with regard to a river, stream, retention pond, pool, swamp, puddle, etc.

      Additionally, as far as I can tell, I see no evidence (other than anecdotal experience) to support the vague assertion that:

      “smaller bodies of water with limited fetch contribute moisture to the mesoscale”

      * how much moisture is ‘contributed’? can you quantify the contribution in any way?
      * smaller bodies of water with limited fetch in comparison (or reference) to what?
      * under what circumstance(s)?
      * have you ruled out alternative explanations?

      Brandon Vincent
      NWS Raleigh, NC

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s