Ed and gang,

My feeling is that an Invest should be scheduled sooner rather than later for the TW that is flaring moderate to heavy convection in the NE Caribbean.

Currently, the looping IR with deep layer wind composite layering shows a strong divergent vector field in the 200mb level, expanding, and situated nearly colocated with densest masses of convection. It is likely this is instrumental in giving additional lift and continued propagation of intense shower activity.

I am having trouble finding wind and pressure obs from that part of the Atlantic Basin that are up to the hour, but TPC reports that upwards of 3mb falls have occurred in the last 24 hours, as well as winds veering south over the Leeward Islands; and all this, prior to this evening's preceived intensification of convection. With difluence so strong and on-going deep tropospheric convection intensified [perhaps in nocturnal cyclic forcing], I would not be surprised if surface convegence were getting better defined.

I believe it is worth it, in part, because the Canadian CMC had a rather intense system for several runs [its TC bias aside] and now we have a feature entering that same domain in an atmosphere increasingly more conducive to development. Granted, the 12z had back off a bitl; however, the 12z UKMET hints that it was sniffing something out, as did the GFS, both showing pressure perturbations riding up the eastern flanks of the EC shear axis in the mid level westerlies in roughly concerted timing with the CMC. The 18z GFS continued along this track with a subtle but yet clear pressure perturbation getting pulled north off the East Coast. And, it should be noted that the Global-based models will typically perform poorly until a tropical feature is better developed and in the sounding array at times of model initialization

Point being, given that there was a presence in the models at all, combined with observations this evening, I think it should peg for the GFDL and let her rip.

[Edit: I found this site and it is great! http://www.srh.noaa.gov/sju/caribm.html ]

John