Anatomy of a Western Washington wind storm forecast
There might have been some disappointment from those who pine for large storms that Sunday's wind speeds were less than forecast but I wanted to pull the curtain back a bit to show you what goes into a forecast like this and why the forecasts were crafted as such.
For instance -- did you know that as of Sunday morning, the storm hadn't even formed yet? Here was a storm about 12 hours from landfall and was nearly invisible on the infrared satellite image. Storms like these undergo what meteorologists call "bombing" (or "Bombogenesis" for those following along with me on Twitter during last night's storm) where the storm undergoes intense development over a few hours.
Our computer forecast models have advanced to the point where they can see this coming, amazingly. If you were just watching the satellite image, you'd have no idea the storm was imminent until it was right upon us (although the water vapor image would have given a few early clues something was up.)
Case in point, here is the infrared satellite image from 5 a.m. Sunday when the morning forecast models came out. Can you spot the windstorm? (Answer to follow and no, it's not that swirl up near the Alaska panhandle)
But as you might imagine, trying to predict a storm seemingly developing out of nowhere and figuring out exactly how strong it'll be and where exactly it's going to go is a challenge.
Here is what the "GFS" model came out with -- a 973 mb low making landfall around central Vancouver Island:
The Euro and some of the other models had similar outcomes, which forecasted a fairly strong wind event -- especially for September.
But then came the NAM model, which painted this grim picture -- again, remember this is just 12-18 hours out, not days away:
That, my friends, is a catastrophic wind scenario for Western Washington. Note how packed together the isobars were along the coast and Puget Sound/Northern interior areas. This would have been easily hurricane force on the coast -- perhaps even nearing triple digit wind speeds and likely 70-80 mph if not a little more in the northern inland areas, and probably 60-70 mph in the Seattle area. Port Townsend, Whidbey Island and Forks would have likely had widespread damage.
If you were to go back and look at historical windstorm setups, you would see some similar patterns, although perhaps not even this strong. I think it would have easily matched the Hanukah Eve storm from Dec. 2006 although that storm was at 976mb when it made landfall near Forks and I know it's tough to read on this map but it I think it says 970.
Add in the fact that the ground was super saturated from Saturday's drenching rains and that all the trees around here had all their leaves still on them (as opposed to mid-December) and it was a set up for a massive number of tree falls and jaw-dropping power outage numbers.
Now to be fair, the NAM does not have as good of track record (no pun intended) as the GFS (and Euro) --- if we were batting the NAM in a lineup, we'd put it 6th or 7th behind cleanup hitters the Euro and GFS. And as it turned out, the NAM was indeed off on its own a bit and the GFS was the winning model. But considering the dangers the NAM presented, and knowing sometimes, the NAM does do it better (like the Jan. 17, 2011 ice storm, for example) you have to take that into consideration and I think that's why the High Wind Warnings were posted by the National Weather Service with higher-than-usual wind speed forecasts.
It's bad in the sense the wind speeds didn't match the forecast, but it was better than the reverse scenario.
Did you find the storm?
To finish up my point: Did you find the storm in that satellite image above?
Here is where it formed:
Look for the swirl that went into Vancouver Island -- right around 973 mb which, had it gone in a bit further south, would have indeed done a lot more damage around here. So I hope that better explains a bit the likely thinking behind those High Wind Warnings.