Originally posted by David Wolfson
Bryan, a really big haha is the 18Z GFS just out. Ivan skirts the Florida east coast before looping off Hatteras, then slams straight up into New England.
The 12z GFS run is available in the 16-day full forecast at
http://www.nco.ncep.noaa.gov/pmb/nwprod/analysis/ .. That run has Ivan hitting FL, curving out off the NC/SC coast, spending a couple of days out at sea, and then recurving to come back on land in SC...
There has been a trend to shift track forecast positions west with time. Historically, the UKMET and GFDL have done well with the longer-range forecasts (~96hrs+), though the GFDL did not handle Frances very well at all (neither did the GFS, for what it's worth). The UKMET did do quite well with Frances, though its forecast for Ivan seems too far north/east.
SST maps do show the some cooler water over the southern half of the Caribbean, which may limit maximum strength... Then again, it's almost guarenteed that there'll be some unforeseen forecast complication (shear, dry air entrainment, etc), so I don't know if it's worth that much effort to try to get this thing down...
EDIT: For what it's worth, a 22z recon report shows central pressure down to 950mb. An eye has emerged in the past few sat pics, and there is some _very_ cold cloud tops symmetrically surrounding the eye. It looks as if the storm is undergoing rapid intesification given it's improved satellite representation and the 6mb pressure drop in 3-6 hours. Interestingly, the storm has made a relatively abrupt turn to the northeast it appears...