December 10-12 Potentially Significant Winter Storm

 There is increasing potential for accumulating snow across the north-central and Upper Midwest in the coming days, as a strong low-pressure system moves northeast through the Central Plains and Great Lakes. Note that there are two disturbances in this timeframe - one affecting the Midwest, and another affecting the East Coast. This post covers the first region.

Because of relatively large variance between model guidance, I will first examine the GFS model, followed by the ECMWF model. A summary of my own thoughts will conclude.

GFS Model Analysis

Attached above is a look at 3-hour accumulated precipitation (by type) and sea level pressure contours, per weathermodels.com and per the 12z GFS, valid at 12pm Central Time on December 10th. The GFS anticipates a strong surface low (989 millibars) to eject from the Rockies into the Kansas / Nebraska area around this time, tossing out a band of snow well ahead of the low stretching into South Dakota, Minnesota and northern Wisconsin. Snowfall intensifies in SD as the low approaches and allows for the deformation zone to form.

Moving ahead to 12:01 am on December 11th, we find the storm to be a quick-moving one, with the surface low now in the Dubuque, Iowa vicinity and developing into a mature cyclone (evidenced by ongoing rain and thunderstorm activity in the South and a continued band of moderate snows north and northwest of the low). Accumulating snow looks to overspread Minnesota and northern Wisconsin at this point in time.

The primary issue for those hoping for snow appears to be an aggressive warm front associated with the storm system. Indeed, a warm antecedent air mass (a.k.a. relatively warm weather in the Central U.S. in advance of this storm system exiting the Rockies in the first place) will allow the warm front to keep precipitation type as liquid well north and east of the surface low, stretching past Michigan and into south-central Canada!

Taking a look at projected accumulated snowfall for this event (remember, NOT including the East Coast disturbance around this same time) shows a relatively impressive snowstorm for parts of the Upper Midwest and Plains. Assuming a 10-to-1 snow-to-liquid conversion (meaning 10" of snow can be melted down to 1" of water), a swath of South Dakota is at risk for snows exceeding 6", while portions of Minnesota and especially northern Wisconsin could see a run at more than 12" of snow as the disturbance associated with this storm system matures and intensifies. Indeed, taking the GFS verbatim, some locations in northern Wisconsin may have a shot at 18" of snow!

It's also possible these totals are a little under what they may actually turn out to be, as surface temperatures look to be firmly in the mid- to lower-20s during this event, turning that 10-to-1 snow ratio more into a 15-to-1 ratio.

Now that we've gotten a look at what the GFS model wants, let's peek at the ECMWF's interpretation.

ECMWF Model Analysis

We'll be examining the same points in time here as we did in the GFS analysis: for instance, this first graph shows forecasted precipitation type and sea level pressure contours at noon on December 10th. The ECMWF places this surface low somewhat to the south of the GFS, resulting in a southward shift of the moderate snow band. Indeed, rather than South Dakota receiving the moderate snows from this system as in the GFS, we find much of Nebraska (but not Omaha!) under the gun for accumulating snow. The overall precipitation presentation in the eastern Rockies also looks more robust in the ECMWF, implying that the system may put down more snow than what the GFS showed us.

Moving ahead to 12:01am on December 11th, we find the ECMWF placing the surface low pretty close to where the GFS showed it, but the big difference is the marked increase in moisture available immediately north and west of the surface low. While the GFS has snowfall coming to an end not too long after the surface low passes a given line of longitude, the ECMWF allows moderate to heavy snowfall to continue in Nebraska even as the surface low is near Dubuque!

The result is a much snowier storm system:

As depicted above, via WeatherBell, the 12z ECMWF delivers over 12" of snow to the northern tier of Nebraska, even laying down several inches of snow on the back end of the storm system in parts of eastern Nebraska. A sharp gradient near the surface low likely results in a modest icing threat in the Omaha metro area, while continued snow accumulations exceeding 15" track north and east through northwest Iowa, southern Minnesota and northern Wisconsin.

My Thoughts

A look at how each model handles the mid-level pattern gives some insight as to why these two forecasts are markedly different. 

The first thing that is complicating this outlook is the fact that we're dealing with two separate pieces of energy in making this winter storm happen. I've highlighted these two pieces of energy in a 500-millibar relative vorticity image on the right, valid at noon on December 9th. That might all sound a bit complicated, but in a nutshell, the more colors you see, the more likely it is that there is a 'piece' of energy present that can develop into a storm system.

Both models agree that these two pieces of energy will make landfall on the West Coast in the Pacific Northwest and in California, respectively. It's how they mingle and interact when they meet up in the Rockies that is throwing models for a loop. 

The GFS model wants to hold the core of this energy back in the Rockies a little longer than the ECMWF, making the system positively-tilted at the mid-levels (a development that keeps the storm system relatively immature, and making it tougher to get those heavier snow totals). In contrast, the ECMWF allows a core of this energy to track north and east with the surface low and tilt negatively as it does so, amplifying the surface low and resulting in a more enthusiastic snowfall outlook across the Central.

The current strongly-positive state of the Arctic Oscillation (shown above) helps mold the broader weather pattern by making it less likely that we get these big storm systems with big snow totals, and does lead some credibility to the weaker, more progressive GFS model. At the same time, the Pacific-North American index (PNA) is poised to be strongly-negative at this point in time, which helps fire up a ridge of high pressure in the East U.S. and gives any storm systems in the Central a "ramp" in the mid-levels on which these pieces of energy can tilt negatively and mature into powerful storms.

Putting all these pieces together, I am in favor of the GFS' solution right now, but I do think it's pretty realistic we could see a southward adjustment like the ECMWF shows, in which case we would also see a stronger storm system and higher snow totals. The overwhelming negative state of the PNA drives that conclusion, with multiple GFS ensembles joining a majority of ECMWF ensembles in playing that southern shift out and making for a stronger storm.

To Summarize:

• A potentially significant winter storm is expected in the December 10th - 12th timeframe across the Central U.S.
• At this time, South Dakota, northwest Iowa, southern Minnesota and northern Wisconsin may see snowfall totals of 4-12", with those higher amounts much more likely in NW Iowa / S MN and especially northern WI
• Given regional teleconnections, it seems quite plausible that future forecasts shift this snow axis a bit south and increase snowfall totals, as the ECMWF model projects
• It is too early to tell for sure, but there may be a minimal icing threat along the freezing-temperature line. Material ice accretion is not expected from this storm.

Andrew