I will be quite surprised if this is not a depression in the 5pm and it may develop substantially overnight tonight.

The last 3 frames here: http://www.ssd.noaa.gov/goes/flt/t1/loop-vis.html Clearly show an accelerating cyclonic torque centered very close to 11.5N/58W -- of course this may vary from official.

Rapid improvement in sat presentation has commenced.

Deep layer wind analysis shows very little shear. As the convection potency has increased the transfer of mass to the upper levels, that injection enhances the outlfow in an already fairly favorable 200mb to begin with.. This means that the system has an above average evacuation being established. (There is also a thermodynamic process involved in the latent heat release of convection but that may be over folk's heads).

Oceanic heat content is quite high where 94L is and will only increase along its extrapolated path. There was some question of turbulent overturning left in the wake of Dean, but given to the depthy thermoclines of the Caribbean, and observing SSTs, this does not seem to be a problem.

The dry air that was likely contributory in preventing more robust development heretofore is still prevalent. However, belated but not denied is the key here... The dry air can be overcome given time via storm-centric production of increasing theta-e (moisture in atmosphere). In other words, the developing TC is producing a bit of its own core from which to fester within.

The reason there is only a vague presentation in the models may have to do with the models trying to initialize for a pressure perturbation that is all of 2mb lower than standard sea-level; and also relative to the ambient observed pressures in that area. As the system develops further, however, eventually the models will sniff out its presents more clearly and we will likely see some different solutions emanating rather abruptly. Fwiw, the ECM's 12z solution is a bit more robust now after having lost it for several runs, so perhaps we are beginning this process now.

John