The Trouble With Models

I have been faced this week with a classic dilemma that all meteorologist face from time to time; how to deal with vastly different model solutions.

I wrote my blog partially on this Monday. Back then there was as much as a 300 meter difference in the strength of the 500 mb trough over California next Tuesday with the European and Canadian much weaker than the deep closed low the GFS operational model had. At the time, it seemed like the GFS was out on its own limb, too far out for my comfort. With all my forecast, I pretty much followed the European and Canadian.

As soon as 24 hours later, the GFS proved me right. It started to back off a lot on the strength of the trough and was becoming more in line with other models. This transformation was complete by midweek when the models were now relatively close in their solutions.

Here again is the GFS 24-hour forecast from Monday for Tuesday April 22.

Now here is what the same model has today for Tuesday.

It doesn’t take a degree in meteorology to notice the difference in the forecast. Note also the tremendous disagreement in the Great Lakes and along the East Coast.

Studies have shown that more often than not, the European is out dueling the American Global Forecast System model. Many times, as in this case, it isn’t even much of a duel. It’s unfortunate, in my opinion, that we can’t get a better and more refined longer-range model than the American long-range model. Though the European is not always right, when there are discrepancy the percentage is high enough to at least go much more toward the European model. Also comparing both these two models with the Canadian and UKMET models may help choose the most likely pattern. This is exactly what tipped the scales for me on Monday for a weaker solution.

I am faced again today with the GFS forecasting a rather cold and wet pattern toward a week from this weekend. Unlike the previous case, other models are also showing a cool, unsettled weather pattern. They are some differences, especially in the orientation of the trough, and thus where it will or will not rain. While a week away, and there most likely are going to be changes, the chances of cooler and maybe even west weather well south into California is at least looking possible. More on this early next week.

Follow me on Twitter @Kenwxman