Stolen from hurricantrack.com's facebook page from last night:

Here is what our colleague Mike Watkins had to say about it all on Storm2k.org this evening: Mark and I were talking today about that giant low that develops in the GFS that runs up the coast like a summer nor'easter later in the week. After thinking about this today, I think the GFS is simply having trouble dealing with all of the heat in this region, so it's creating a feedback low in the Gulf of Mexico.... See More Looking for confirmation on this thinking, I checked the Wilmington NWS AFD, since they would have to deal with this system, and here's what they said: Quote: GFS DEVELOPS A FEEDBACK LOW ALONG THE GULF COAST WHICH IT THEN MOVES EAST ACROSS FL AND UP THE SOUTHEAST COAST. PARALLEL GFS DOES NOT SHOW THIS LOW...FURTHER ADDING CONFIDENCE THAT THE FEATURE WILL NOT EXIST. The GFS has never handled heat like this very well, and I think the problem with the 18Z gfs and the other models that feed from it (HWRF and GFDL to name a couple) is this: The "feedback low" is busting out just enough of the Gulf ridge at 500mb. This allows steering north of Alex (in the model) to break down enough to let Alex catch the trough coming through. Oh, there's no doubt that trough is going to come through, but odds of it digging Alex out on it's own are slim, Alex is too far south and it's almost July. So...the GFS gives the trough help with it's famous create-a-storm feature. Seriously, watch this nonsense in the 500MB loop: http://www.nco.ncep.noaa.gov/pmb/nwprod ... loop.shtml If you really, really want to see if the GFS is on the right track, look for unusual development of some sort of deep-layer low pressure system in the NE Gulf in 2-3 days. If this feature shows up, then maybe GFS scores a coup. However, I think the GFS can't handle the heat properly, creates a phantom system, and pulls Alex up incorrectly.

Then as noted in a later discussion...there is an upper level low coming into play in East Texas as we speak.