Maybe they could be synced using RF over fiber. This has been proposed as candidate technology for 6g wireless networks, to enable cell free massive MIMO.

That would mean that you would need to run optical fiber to each of them, though we’ve already seen fiber drones spool out kilometers of the stuff as they fly.

EDIT: I just remembered this interesting article about doing radio interferometry over a fiber network using cheap quartz oscillators instead of atomic clocks. My (layman’s) understanding is that the quartz oscillators are good enough over a few milliseconds, but will fall out of sync with each other over longer time spans. Meanwhile the fiber optic reference signal (distributed from a central atomic clock) can be kept correct on average by reflecting the reference back down the fiber and doing active correction of the changing path length (caused by thermal fluctuations and vibrations along the fiber) but will be incorrect on a millisecond-to-miliscond basis because of light speed lag and the path length being a moving target. So they use the quartz oscillators over small time scales and use the fiber reference signal to keep them synced over long time scales. Surprisingly the article says they actually get a better sync this way than with using multiple atomic clocks.

So perhaps something like that is possible.