Why Did the Snowstorm Cross the Road?

In our news story about the mid-Atlantic snowstorm today, which set up farther south than originally forecast, Elliot Abrams was quoted as saying:

"A narrow swath of snow and ice reached for more than 1,200 miles, but rather than centering near the I-80 to I-70 corridor, the mess was centered on the I-70 to I-64 corridors."

In other words: We answer the question: Why Did the Snowstorm Cross the Road?

I think it's very important that forecasters say when, but more importantly, why they were wrong. In my opinion, this was the worst medium-range forecast bust (in general, for all meteorologists) since the one where Henry and I got death threats in 2009 (he has some additional comments today).

I contributed to our article, and one reason that I really wanted to hammer home: It had been a while since meteorologists had to deal with a Pacific storm (there's a big drought in California, remember). Those types of storms are notoriously hard for computer forecast models to predict until they hit land. This time, that included the European model (see Wednesday's vs. Sunday's forecast below), which many meteorologists hold in high regard because of its forecast for Hurricane Sandy. Hurricane Sandy was a much-better-measured Atlantic storm.

Why? There are no upper-air or surface weather sensors in the Pacific Ocean, so the storm can't be measured sufficiently for it to be forecast in a 3-D computer algorithm. This isn't the case when the storms come from Canada or Alaska, or across the Gulf states, as most winter storms this season have.

Forecasters might have been better off to wait on issuing accumulation forecasts until Saturday, when the models handled the storm better and placed it over the mid-Atlantic states. The problem with that is: the public demands it sooner, and if you wait too long, you're the last to get the news out and people complain about that too.