91L is looking better and better this afternoon. Since the upgrade last month, the 'new and improved' GFS seems to be missing developments of systems that do develop, and over-hyping others that do not develop at all. I tend to think 91L is probably going to be another miss for the GFS, despite it being historically one of the best genesis models.

When it comes to models so far this season, if a system has the DNA, I've paid more attention to the almost always bullish Canadian, because it is at least picking up on a cyclone, where others, including the GFS, may have not.

12z Runs

The most recent run of the Canadian expects a TD by the time 91L crosses the Lesser Antilles, becoming a mid-grade trop storm by time 91L has traveled half-way across the Caribbean.

Of the hurricane specific models that tend to just run on the assumption that a feature has already, or will develop, the new and improved HWRF follows a similar track and intensity trend, while the new and improved HMON keeps 91L a sloppy Trop Storm through the Caribbean, at best.