Friday, February 21, 2014
Heads or tails?
With the European model saying essentially no snow, the NAM (one American model) giving snow to the foothills and places to the far north, and the GFS (the premier American model) giving 3-8 inches of snow everywhere, you might as well flip a coin at this point to decide whether there will be snow or not. I’ve had to make some tough forecasts, but this may just about be the toughest one yet.
What’s interesting, however, is that the GFS has been showing snow for our area since February 17, and it has been doing so on every single run. Some runs have had greater amounts than others, and the range has been different, but all of the runs have had sea-level snow extending southward to northern King County. The newest WRF-GFS from the UW this morning brings 3-4 inches of snow over all of Western Washington this weekend. I was EXTREMELY skeptical of these runs, as the temperatures did simply not look remotely cold enough for it to snow. But, it’s Friday, and the GFS is still showing snow. Here’s this morning’s run.
|Valid 04:00 am PST, Mon 24 Feb 2014 – 72hr Fcst: http://www.atmos.washington.edu/~ovens/wxloop.cgi?mm5d2_x_wa_snow72+///3|
You think that’s crazy? Check out the run the morning before. Tacoma southward gets left out, but look at how much snow Bellingham gets! We’re talking a foot in places! That would certainly make up for a humdrum winter.
|Valid 04:00 pm PST, Mon 24 Feb 2014 – 108hr Fcst: http://www.atmos.washington.edu/~ovens/wxloop.cgi?mm5d2_x_wa_snow72+///3|
The ECMWF or Euro (European model), on the other hand, is much less optimistic. No snow anywhere remotely close to sea level in Western Washington save the furthest northern regions near the Canadian border. Again, the NAM is somewhat in the middle, with low snow levels providing some accumulation in the foothills and places up north but not quite down south or low enough to Seattle.
The bottom line is that this is one of, if not THE, hardest forecasts I’ve ever had the pleasure of making. If you asked me if it was going to snow over the weekend a couple days ago, I would have said no and that the GFS was just going bonkers. But we are quickly approaching the weekend, and the GFS is still showing snow; one pulse Saturday night and another, larger one Sunday night. If I had to make a preliminary forecast, I would guess that much of Western Washington would see a rain-snow mix or non-sticking snow, but places above 500 feet could certainly see accumulation. We’ll get a better idea with each model run.
The point of this post is to highlight the ridiculously large uncertainty in the forecast. I’ll get down to the details as the models converge on a solution (assuming they even converge on a solution at all). Forecasting snow in Seattle is hard enough, but when you have models that are varying wildly this close to the event with temperatures right on the fringe, the task becomes impossible. One thing’s for sure, though. If it does end up snowing, the GFS is going to gain mad respect from the local weather buffs across the area. Maybe in the future, people won’t have to resort to flipping coins. 😉
Do those dances!