You be the forecaster: Let’s talk about next week’s possible storm

The weather rumor mill is churning. There's talk of big snow next week somewhere in the Upper Midwest.

But how much can we credibly say, and not say, about a possible snowstorm six days in advance? What is the state of the science in snowfall forecasting in the days leading up to winter storms?  Which forecast models do the best job overall? What model biases do we need to watch out for?

Model forecast skill has improved dramatically over the past 30 years. In spite of your uncle's bad jokes to the contrary, weather forecast skill is a major science success story.

In this post, I'll lay out my take on these questions and assess next week's system based on what we know today. And I know a lot of professional meteorologists and hobby forecasters read this space.

So let's use the comments on this post as a running update as each new model run comes in. I'll let you carry the discussion, but I'll chime in now and then with my latest read. Maybe together we can achieve the best consensus on next week's system?

Will it be a boom? Or a bust? Place your bets below.

Good pre-Christmas travel 

First things first.

Travel conditions leading up to Christmas look generally favorable. A weak low-pressure system brings light snow (generally less than 1 inch) to northern Minnesota Saturday. It's likely just flurries for the Twin Cities. Another minor system may bring a snowy dusting around Christmas but travel looks generally good.

Here's the National Oceanic and Atmospheric Administration's NAM 3-km resolution model for Saturday.

More seasonable temperatures 

Temperatures hover a few degrees above average the next few days. The average high and low for the Twin Cities by Sunday are 25 and 10 degrees.

Storm next week?

The forecast models continue to drive a strong low-pressure system in the Upper Midwest next Wednesday-Thursday time-frame. All major models show some kind of system. But what qualifies as actionable weather information 6-days ahead of an inbound storm?

I monitor but don't broadcast specific snowfall amounts or maps a week out from a snowfall event.

The storm in question for next week is churning over the Mongolia-China border today. That's a lot of territory to cover before it arrives in the Upper Midwest. The Twin Cities National Weather Service office does a nice job of reminding us how many variables lie ahead.

So what are the confidence levels we can communicate six days ahead of a possible snowstorm? What is actionable weather information this far ahead of a storm? Here's my messaging as of Friday.

• High confidence: There will likely be a major winter weather system somewhere in the Upper Midwest on Dec. 26-27. Travel will likely be impacted with areas of rain, ice, and the potential for heavy snow. Those planning post-Christmas travel should monitor forecasts closely.

• Low confidence: The precise storm track and temperature profile for the storm are still in question. It's way too early to pinpoint precipitation coverage and type with any confidence.

• Low confidence: Snowfall coverage and totals are impossible to pinpoint with any certainty this far out.

Forecast model spread 

It's typical to have strong model divergence  six days out. We often see model spread right up to the day before, and some models just plain miss system dynamics like track and temperature profile on a given system.

Right now the Canadian model is on the west end of the forecast envelope. This scenario would bring mostly rain to the Twin Cities and eastern Minnesota, with a transition to wet snow as the system moves by.

NOAA's Global Forecast System model and the European Centre for Medium-Range Weather Forecasts model favor a more southeasterly track. That would bring heavy snow from the Twin Cities north and west.

The rain-snow line in this scenario would set up pretty close to Minneapolis-St. Paul, so mixed precip would favor southeast Minnesota with rain across much of Iowa and Wisconsin.

The Twin Cities NWS has this insightful take on the different storm track scenarios.

So, there are still plenty of variables to watch for in the next 5 days. I issue snowfall forecasts within 24-48 hours ahead of a system. That's where the forecast model accuracy and most actionable information lead-time curves meet.

Looking at this system now I think it's likely that somebody in Minnesota is going to get heavy snow, possibly a foot or more.

So what's your take on next week's system at this point? What will you be watching for? Which models do you trust, or does performance vary on individual storms? I have my own experiences with and take on model-bias and my favorites in different situations.

What will you be watching? Will this be the biggest storm of winter? And who do you trust the most (and least) when it comes to snowfall forecasting in Minnesota? Why?