Tim Vasquez has posted a fantastic weather analysis for the AF744 flight here
Air France 447 - AFR447 - A detailed meteorological analysis - Satellite and weather data . At the same time some pilots have posted observations of temps within cells climbing with 30 degrees within seconds and on the R&N live thread this is currently being discussed.
I have asked Tim his opinion on the likelyhood of this and below is his reply. Unfortunately I am not able to post on R&N, perhaps pilots reading this thread may wish to report it there.
Sorry for the thread drift. I felt it was necessary given the circumstances.
I do not agree that a bubble of warmer air (that is, any warmer than about
5 degrees compared to the environmental air) would have made it up
to flight level. This requires exceptionally high equivalent potential temperatures at some
lower altitude. The atmosphere has a tendency to overturn bubbles of hot air as soon as they
start becoming significant because "absolutely unstable" lapse rates are unsustainable.
We do see thunderstorm heat burst phenomena on the Great Plains at night, but this occurs due to
the downward forcing of a low-level inversion, and I can't picture a mechanism for
this to occur at flight level given the conditions shown.
But in regard to the above mention of an aircraft's "coffin corner"
(<a href="http://en.wikipedia.org/wiki/Coffin_corner_(aviation)">Wikipedia</a>) it is
possible that wind values alone could greatly affected airspeed -- on Doppler radar we often
see anomalies of 40 to 80 kt at flight level within storms (Google "storm top divergence" for
some examples).
Tim