Most of the videos that include the bang start after the flash because they were taken manually by people with phone cameras. And most of the security and dashboard cameras running at the time of the fireball stop before the bang. But there are a few that show both, so I got interested in trying to estimate the altitude of the explosion. With that I hoped to use my references on nuclear weapons effects to estimate the energy release needed to give the observed effects on the ground.
The sunlight on the underside of the cloud is one qualitative clue of its minimum altitude, but I haven't checked to see what the ground elevation of the sun (which appears to have been just rising) was at that time.
In one dashboard camera I measure a flash-bang delay of 2 minutes 25 seconds. It does not appear overhead in that one so the resulting straight-line distance estimate of 46 km (based on an average speed of sound of 315 m/s) probably overestimates the altitude. It's hard to estimate elevation angle in these cameras, and some of the security footage doesn't show the fireball at all.
I'm sure there are all sorts of refraction effects that I'm not equipped to analyze.
There were actually quite a few nuclear weapons tests at similar altitudes by both the US and USSR, but I can't find any discussion of their blast effects. The reports are all about thermal and EMP effects, presumably because the blast is nil (mere broken windows would be pretty mild for a nuclear war).
Comments?