Maybe you can observe it over a long time, and you'll note that the data repeats every N hours. That must be the rotation time, and then you can find the corres

Maybe you can observe it over a long time, and you'll note that the data repeats every N hours. That must be the rotation time, and then you can find the corresponding "pixel" for each rotation and work with that data. Sounds like several near impossible problems on top of each other though.