Models On How Cold Late November Could Be
We published a news story yesterday entitled "November Arctic Cold Blast Forecast", where we explained our concern for a cold, stormy latter half of November, possibly disrupting travel during Thanksgiving week. Today I want to take a look at some of the computer forecast models and try to quantify just how cold they think it will be.
There are really only two forecast models that go far enough into the future to glimpse this cold snap. The GFS goes out two weeks, and its last low temperature prediction is shown above. That's right, below zero in the Plains and single digits into Pittsburgh. As a result of that extremely-cold forecast, the model is estimating that this is what the Snow Cover map will look like on Black Friday:
But before you get too excited, checking the GFS ensembles (multiple runs of the same model), the average temperature in Pittsburgh is near 0 C (32 F) that night, not single digits. Why the disagreement? Apparently the "real" GFS is an outlier when you take all the different perturbations into account -- although that average (mean) prediction is +/- 5 C at Pittsburgh and +/- 15 degrees (based on the other runs) in the Plains because the model gets less accurate as it forecasts out into time.
The ECMWF model only goes out 10 days, but shows single digits on the central Canadian border on November 20th. If both models continue to be in agreement about a major cold outbreak, it will likely happen, though it's also likely it won't be as bad as the Operational GFS says above.
