^{Weather model forecast for October 6 and October 17. Source: Weatherbell.com}

The last few days of September and the first few days of October have been snowy across the west (pictures). But now the pattern will reverse as the cold air moves to the eastern US.

With the upcoming dry and sunny period lasting at least 7-10 days for the west, of course the next question is "When will the snow return?"

Forecasting the general weather pattern is something that the computer weather models do pretty well inside of 5 days, but beyond about 7-10 days any single forecast from these same models is rarey accurate.

One way that we can get some value from the model's 10-15 day forecasts is to change each of the models just slightly, run them a bunch of times with these changes, and average the results together. This is called a model "ensemble", and while these ensembles aren't perfect, sometimes they can help to tease out the general long-range weather pattern.

When looking the American GFS model ensemble (above) and the Canadian GEM model ensemble (below), you'll notice a pattern of warm in the west and cool in the east around October 6th. Then, around the middle of October, the models hint that this pattern will reverse and the west and midwest will cool off (with the potential for snow).

^{Weather model forecast for October 6 and October 17. Source: Weatherbell.com}

When you look closely at these predictions, you'll notice that the forecast from the American GFS ensemble (top) for mid October is showing much colder air for the west than the Canadian model (above). The European model (not pictured due to licensing ... the Euros charge a LOT of money to use their graphics) goes halve-sies between the American and Canadian models as it shows cool and stormy weather over the west during the middle of October, but not as much as the American model.

So what does all this mean?

It means that if you're in the east, enjoy the fall-like weather for the next week.

If you're in the west, it means that you should enjoy bonus days of mountain biking and hiking, and look toward mid October for a hopeful pattern shift back toward cooler and stormier weather.

One additional note to this long-range mumbo jumbo is that pattern shifts like the one we're looking for in mid October often take longer than the models predict, so I wouldn't be surprised if the shift came later, perhaps during the 3rd or even 4th week of October. Joel vs the models ... who will win? Ach, it doesn't matter, because as long as it snows, we're all happy:-)

JOEL GRATZ
OpenSnow.com