Hard to model as it will be dependent on temperature amongst other things. I would assume there is a FAA or whatever spec on required minimum lifetime, and one would hope that any manufacturer would add significant margin to that.
For our underwater stuff (sonarbuoys, ocean bottom recorders, + others, military and civilian) we had to characterise the battery life over various storage/transport temperature cycles, and then again with operations temperatures ranges. And non of this accelerated testing stuff, if it was a months battery life, we'd run groups of them for a month under the real loads.
Then on top of that you'd have vibration as well, and there are standards for for aircraft, trucks etc. Not to mention different suppliers, and batch differences.
So it's very complex business if you take it serious, remember that batteries like this get hot and cold cycles and vibrated during transport as well as operation. .e.g parts of the the back deck of a seismic boat in the hot sun could get to 60C+, and then the ocean can be
below freezing due to the salt water. And all that before they get used.