So out of those 2,700 days, if there's no nuclear test, there's a transient on 11% of those days.
But if there's been a nuclear test the day before, then it's almost 19% of those days have a transient.
And if we could get access to digitized plates from some of these other locations, I would think we'd be some ability to kind of triangulate on those days when you get a transient that coincides with the nuclear test and may be able to identify roughly what direction that was in.