Weather Blog

Weather satellite pictures prove looks can be deceiving...

Weather satellite pictures prove looks can be deceiving...
Big storm brewing for Seattle? Not quite...

Satellite images are one of the most important tools a meteorologist uses in making a forecast -- right up there with the computer forecast models we run. But sometimes its looks can be deceiving.

Sometimes you can see what looks like a huge storm approaching, only to find it's barely got enough rain to wet the ground. Other times what looks just like a harmless batch of clouds rapidly turns into a monster-sized storm.

With two great examples of this just in the past few weeks, I thought it'd be neat to showcase how the computer models and satellite images work hand-in-hand. You might think they are two independent systems -- and in a sense, they are -- but their effectiveness is greatly reduced without one being hand-held by the other.

Satellite data is used extensively in initializing the forecast models -- basically telling the models what's going on right now as a base point for the model's calculations to be run. (It's just like you can't file a effective flight plan without knowing what airport you're leaving from!)

That data is especially crucial over the oceans where we don't have a whole lot of weather instruments to give actual conditions. It used to be, as Steve Pool would say, a satellite image was like looking at the roof of someone's house: you know it's there but you don't know what's going on inside. But recent advances in satellite technology now have better tools and sensors to decipher what is going on inside those clouds.

But then once those forecast models calculate and spew out their forecasts, satellite data is used extensively to verify the calculations -- making sure as time moves forward that the storms are where the models think they are and should be going. If not, then you know your model is going to be pretty useless.

Proof is in the pudding

Here are some actual examples of what I'm talking about. First up, our wind storm that hit on Sept. 30.

Here is a raw infrared satellite image of the "storm" 12 hours before it would make landfall. Doesn't look very ominous, does it?



Now, here is what the storm looked like just a few hours later -- a well-defined low center off Vancouver Island:



But we knew even though the satellite image didn't show the storm that morning, that a windy day was in the offing because the forecast models had the storm rapidly strengthening. Only problem was models were having issues where it was going to go:

One model had it going north into Vancouver Island:



The other had it coming in closer to Washington which would have been a windier day.



But then watching the satellite image as the storm developed, help guide forecasters as to which forecast model was going to have a better handle on the storm (the first model.)

Second example, our windstorm on Saturday, Nov. 2 that injured two people and shut down the 520 Bridge for much of the afternoon.

Here is the satellite image from noon on the Friday before. At face value, doesn't look too intimidating:



But here is the model animation showing that storm rapidly intensifying:



And sure enough, here was the satellite image, pretty much verifying the model's track as the storm swirled up, showing an intensifying low pressure center, and moved ashore.



Now, the other way around:

But it was the front that moved ashore on Tuesday morning that got me to thinking about how satellite looks can really be deceiving, because here is a satellite pic of the storm as it sat offshore on Monday:



Looks pretty intense! A big swirling 984 mb low offshore-- looking somewhat like the lows that moved ashore with the two wind events earlier this season. On top of that, a big, bad cold front that looks like it'll bring a drenching rain and maybe some Flood Watches and Wind Advisories, right?

Eh, not so much...

Peak wind gust at Seattle overnight as the storm came ashore? 8 mph. Drenching rain? Try 0.16". But there were no wind advisories up, no flood watches, no Steve Pool having seven weather hits during the first 10 minutes of the newscasts. Just a forecast of light rain in the morning, decreasing by midday and dare I say there are sunbreaks out at midday as I type this.

So no big storm, despite the ominous pictures, and it wasn't a surprise.

Again, the models help show that this system's bark was worse than its bite. First, it shows the mature low moving well off to the north and weakening:



And here is the rainfall forecast -- an organized front offshore rapidly fizzling as it moved inland with just a few showers in its wake.


And sure enough here is what the satellite looked like Tuesday morning -- pretty mundane stuff:



So just goes to prove that sometimes looks can be deceiving, and the complexities that go into trying to solve what Mother Nature has up her sleeve!