bildan wrote:
http://www.newscientist.com/blogs/on...photo-lens-spo
ts-clear-air.html

Crap - they almost stole my idea! ;-)

Seriously though, I had been thinking of using a normal video camera with
changing focus to check for dust sparkle at different ranges and do frame
comparisons. Take two (or more) snapshots at each focus range. Subtract out
image from previous focus range (trying to eliminate the roughly constant
out of focus "blur" of the background) to leave only spots in the new focus
range that are likely to be dust particles. Do this at least twice, waiting
some time T between frames. Use a heuristic to pair up pixels that appear
to represent the same particle in both frames but have moved. Compute speed
of movement of particles based on focus range, time T, and subtended arc
the particles appear to have moved. You now have an estimate for wind
speeds and directions normal to your view line at that distance.

The above doesn't appear to be what Boeing has in mind; they appear to be
using distortions of the distant background itself.