The GDFL heading that far north was worrisome because it has been one of the best models. The jumps are about the end of ridge and the movement of a low into the Gulf. These are providing the steering for the storm.

That low is a good news/bad news situation that Jeff Masters talks about. The interaction not only provides steering, but if they touch, Dean could be seriously weakened or STRENGTHENED on the exchange.

… taking a short break before going back to work.
my new favorite description of eyewall replacement cycles. thanks.

my new favorite hurricane page: hurricane hunter path through the storm

i see that the models have all moved the track further south than last night’s predictions. generally they’re all in agreement through about the first three days, and diverge quite a bit by five days out. there was a time [starting with ivan] when i was trying to keep track of each individual model’s predictions to see which came closest to being right, but that turned out to be more than i wanted to keep up with.