Weather Nerd

Fay keeps forecasters guessing

At 5:00 PM EDT, the National Hurricane Center adjusted Tropical Storm Fay’s track 70 miles to the west — which, because of the angle of approach to Florida’s coast, moved the predicted landfall location hundreds of miles north — reflecting what appeared to be the evolving computer-model consensus. Compare for yourself the 11am EDT and 5pm EDT forecast tracks for Florida:


Left: The 11am track. Right: The 5pm track.

As you can see, it was a big shift. Instead of Sarasota, the forecast landfall location was now near Cedar Key — and southeast Florida was apparently out of harm’s way entirely, outside the “cone.” Among other things, this abrupt track change convinced blogger Rand Simberg, in Boca Raton, to forget about boarding up his windows.

But wait! A few hours later, the computer models shifted back to the right. Compare the 2pm EDT model runs (below left) — on which the 5pm NHC forecast was based — to the 8pm EDT model runs (below right), courtesy of the Colorado State University “spaghetti” tracks:


Left: The 2pm models. Right: The 8pm models.

Whoa! This, ladies and gentlemen, is why the Hurricane Center so often waits for additional computer-model runs before adjusting its official forecast to fully reflect a sudden change in track guidance. What’s unusual in this case, though, is that the models are still ping-ponging so severely at this late date. Dr. Jeff Masters called it days ago: Fay is “The Joker.”

It will be interesting to see whether the NHC adjusts its track back to the right at 11:00 PM, or maintains the status quo for now. The current track is still justifiable given the new model data — indeed, it’s arguably moreso than it was at 5:00 PM, when it was concededly near the right-hand edge of the guidance envelope — so I suspect they’ll leave it as-is, pending further input from the models. That, of course, raises another big question: what will the 2:00 AM and 8:00 AM models say? It could be a very interesting night.