New GFDL model out taking it to a Cat 3. Given the track record on the GFDL intensity forcasts for the last major storms this year, the running average of intensity is generally a fairly good average:

00Z GFDL predicted a weak Cat 5
06Z GFDL predicted a strong Cat 3
12Z GFDL predicted a strong Cat 5
18Z GFDL predicts a weak Cat 3

Based on the average of these model runs, I'd expect a Cat 4 out of this system right now. Even more amazing, none of the GFDL model runs are trying to disipate the system, so it might be that the Cat 5 runs aren't so far fetched.

The runs now take her into western Cuba heading north. The Gulf coast and all of Florida needs to watch this system. I don't think the SSTs in the Gulf will support yet another Cat 5 system, but we said the same thing when Rita went through and look at what it did. If Wilma (yeah...it's almost certain she will be Wilma by morning...) heads toward the pan handle of FL, she will be heading over the worst oceanic conditions (in terms of SSTs). If she heads east, she'll hit land before having a chance to weaken much. If she heads west she will go up strait over the hot loop current. The track in the 5+ day period will be the determining factor for this storm's strength at landfall.

--RC

GFDL model runs: http://moe.met.fsu.edu/tcgengifs/

Hurricane Heat Potential: http://www7320.nrlssc.navy.mil/hhc/watl/hh3.watl.20051015.gif
SSTs: http://www7320.nrlssc.navy.mil/hhc/watl/sst2.watl.20051015.gif
(NRL SST/HHC Index for the above two links: http://www7320.nrlssc.navy.mil/hhc/all_watl.html)