I think it's a really nice example of how you can combine very diverse datasets, which at first might not seem to have much to do with each other, to get a deep understanding of a key process. First, @agrinsted used a published database of crevasses based on a satellite DEM over the Arctic (#ArcticDEM), combined this with temperatures from the new very high resolution regional #Climate reanalysis (#CARRA), plus ice thickness data from #BedMachine + #IceVelocity data from many many #satellites