It appears the Google DeepMind model will be one of the winners again. Several other models kept wanting to bring Melissa farther north, in between Haiti and Jamaica or into eastern Jamaica from the east/southeast. The Google ensembles never budged from the wsw/west track to the south of Jamaica.
So what about the previous 10 recon passes today that did not confirm the NHC intensity? Did those not count? The NHC even said in their discussion their estimate might be generous. It's no big deal. It was likely going to strengthen more at some point today/tonight.
Melissa absolutely looks better now than it did this morning. Much sharper edges of the eye and CDO.
3rd recon pass this morning continues to find a steady-state storm. The pressure is 953 mb, actually up a hair, and 120 kts flight level in the right (strongest) quad, which equates to 125 mph at the surface using the NHC's 0.9 method.
The new recon plane suggests this is actually not a cat 4. The latest NHC estimate is too low by 8 mb. The pressure has only dropped from 959 to 952 over the last several hours. Also, the new plane did not find any increase in wind on its first pass, whereas the NHC increased the wind by 20 mph.