Model Mayhem: Major forecast differences on weekend snow potential

Model Madness:

So you want to be a meteorologist huh?

All a weather dude can ask for is a little consistency in the forecast models we look at each day. Give us a reasonably accurate idea of the position of high & low pressure centers, temperature profiles, moisture plumes… and we can credibly tell you when to expect rain or snow…and probably how much.

Apparently that’s asking way too much this week.

The main suite of numerical forecast models we look at every day is quite literally… all over the map on this weekend’s storm.

In this edition of Updraft I’ll try and tell you what we know about our incoming weekend storm, what we don’t know, and when we might know more.

73 lorenz_attractor.png

Chaos Theory

Chaos theory studies the behavior of dynamical systems that are highly sensitive to initial conditions, an effect which is popularly referred to as the butterfly effect. Small differences in initial conditions (such as those due to rounding errors in numerical computation) yield widely diverging outcomes for such dynamical systems, rendering long-term prediction impossible in general.Wikipedia

Overnight forecast model output for MSP:

.78″ liquid from European (ECMWF) model for MSP this weekend

(mostly snow)

.90″ from NOAA’s GFS (mostly rain or mixed precip)

292 ql2.PNG

Wanted: Clarity

This might be the widest variety of solutions for an incoming weather system I have seen in many years.

The various models we look at are presenting differing solutions to this weekend’s incoming weather system. The Euro, GFS and NAM among others have different storm tracks and thermal profiles that if correct will produce widely different weather conditions for the Twin Cities this weekend.

Even the same models are signifcantly changing tracks in subsequent runs.

292 500 mb sref.PNG

Short range ensemble forecasts (SREF) upper air plots show varied solutions this weekend


Our weekend weather system is still spinning off the Pacific Northwest Coast today, which means it will not work into the CONUS (Continental U.S.) surface and upper air sampling network until tonight or early Friday. That’s when we’ll get a much better read (more hard data) on the incoming system to feed into the models.

As of now, here’s what we know, and don’t know about the incoming system, and how the models are handling it.

What we know:

-There is a very high likelihood (95%+) of significant rain or snow in Minnesota & the Upper Midwest this weekend.

-Many locations will likely (80%) pick up at least .50″ precipitation by Monday night.

-Some locations will possibly (50/50) pick up to 1″ precipitation this by Monday night.

-The bulk of the precip will come in between Saturday night and Monday morning.

What we don’t know:

-Where will the surface low ultimately track?

-Where will the rain/snow line (850 millibar 0C isotherm) set up Sunday PM?

-How much of the precip will be mixed rain/sleet/ice vs. snow?

-Where will the heaviest snow band ultimately set up?

How the models are handling the system:

Overnight Euro: (0Z run) The “Euro” was the 1st model to pick up on the system Monday. It has been the most consistent with the surface low track with this storm, and maintains the “southernmost” storm track. The overnight Euro takes the surface low from western KS to northern Iowa by 6pm Sunday…clipping southeast Minnesota on the way to Green Bay by 6pm Monday.

292 euro.PNG

Image: College of DuPage

The Euro also maintains the coldest temperature profile, keeping precip mostly as snow for the Twin Cities area.

293 Euro.PNG

Image: Norwegian Met Institute

If the .78″ liquid output from the Euro falls mostly as snow with temps hovering near freezing as forecast, that would produce “several” inches of heavy wet snow for the metro(especially the NW metro) with an even heavier snow band likely setting up from southwest through central and northern Minnesota.

The GFS: NOAA’s GFS model was slow to identify and “resolve” a possible low pressure system for this weekend.

Strike one.

It finally caught on Monday night, but has constantly shifted the storm track westward since then. The overnight (0Z) solution tracks the surface low further NW than Euro…from Denver to Omaha then right over MSP to Lake Superior.

292 gfs.PNG


The NW track and “warmer” solution would mean a large percentage of the system’s precip would fall as rain or a “Wintery Cocktail” Sunday which would greatly reduce snowfall totals in the Twin Cities and southern Minnesota. The mix would change to all snow Sunday night, but by then the dreaded “dry slot” could reduce precipitation. Some wrap around snows could linger through Monday under the GFS scenario.

There would still likely be a band of heavy (6″+) snow from the eastern Dakotas into northern Minnesota.

292 earl.PNG

GFS snowfall output for Minnesota


The NAM: The 84 hour NAM’s latest runs only go to about noon Sunday. But the latest trends (06Z) show an even greater westward shift…taking the surface low over Sioux Falls then almost straight north into the Red River Valley!

292 nam.PNG

That presents yet another radically different solution that the Euro.

Ironically, the major Nor’Easter slamming the northeast with up to 18″ of snow Friday may steal the headlines form our weekend system.

292 us snow.PNG

Bottom line?

It’s still too early to present any credible solutions for this weekend’s incoming system.

The best bet for heavy snow still appears to be from the eastern Dakotas into central and northern Minnesota.

If you believe NOAA’s NAM and GFS models, warmer air will mean mixed precip in the metro and greatly reduced snowfall. If you believe the Euro, we could be shoveling by Sunday night.

With the way the Euro was (again) the 1st to grab hold of this system, and with at least 10-times the computing power of NOAA’s models, I’m not ready to write off the Euro solution and the possibility for heavy snow creeping into the metro (especially the NW metro) Sunday & Monday.

As the system comes ashore into the USA surface and upper air (weather balloon) network by Friday, we should get more consistency from the models on storm track, temp profile and thus precip type.

In “theory” anyway.

Stay tuned.


  • planhart

    I choose the Euro model. Yup, it’s mostly because I got snowshoes for Christmas!

  • Tyler

    Paul, I hope when this storm gets done (despite whatever form it takes…) you can do a followup. I’m finding these models (and their disparity more and more interesting.

  • Chris B. Critter

    I applaud your use of the 850 mb 0 °C isotherm as your rain/snow line. I never did like using the 500mb 540-decameter line.

    I’ve had quite a bit of respect and awe for people who can correctly forecast winter weather in events like these. The precipitation phase is so important and quite the challenge because a difference of 1 or 2 degrees in temperature or dewpoint can be the difference between rain or snow. Combine surface or near-surface temperatures, low-level temperature advection, condensation or deposition, evaporation or sublimation, and other adiabatic effects, and it’s amazing that we get as close as we do now.

  • Snow Miser

    Paul, you did an excellent job of laying out the different scenarios and putting forth an honest assessment of the situation as it appears at this time. Your explanations were very helpful and informative. As a snow lover, I am undoubtedly rooting for the Euro!!!

  • Angela

    Thank you for writing a clear and informative article for those of us who know nothing about weather forecasting!

  • Lindsay

    Oh so confusing. Thank you for explaining. I’d rather hear this type of forecast than a solid forecast that ends up being totally wrong in the end. You’re my PHavorite local meteorologist.