Is it just me, or do you also see a recurring pattern here.

I get the sense that cloud features are patterned by cross wind bands radiating like arcs with the downwind evolution of the trades. Here for instance Barbados is in a dry fold, and there is an elongated cloud patch about 250 km to the east, followed by another clearing, then a patch that goes through the black space where the MODIS swaths don’t overlap, and so on. I like to imagine these as large-scale regions of enhanced low-level convergence and divergence, but even if this interpretation is correct, what causes them and what might they tell us about cloud controlling factors? 

Supposing that the bands have a wave-length of about 700 km; then this is about one day at the mean advection speed. There seems to be some suggestion of a slackening of the wind and a drying of the boundary layer on these time/space scales. Here for instance look at the time-series as we enter the dry period at around 14 UTC. Dew point temperatures are 4ºC cooler than they were at 0 UTC and winds drop off to about 6 m/s as compared to 9-10 m/s earlier in the day and yesterday.

Well to look at this a bit more I averaged the winds over two hour intervals plotted all of the 2018 data (A3), and just the February data (A4). Note the funny x-axis, as day 1 ends at 1. Maybe there is a hint over the last two or three days of a pulsing of the winds on the right time-scale. These features also remind me of folds we saw when flying over the Atlantic in Dec 2013, where the WV lidar RH profiles are attached (A5, upper most panel). Look at the feature at about 18:37, and also the dip in the dry air at the end of the flight, near 19:30.  

Something to look at more closely with the fullness of the data.