Let's say we had a probe in orbit around the moon mapping out mass concentrations (mascons), as we've done in the past. Unexpectedly one day, a massive object lands on the surface of the moon, an object as big as a mountain range.
How long would it take us to detect it using only the variations in the orbit of our probe?
(Basically I'm asking about the temporal resolution of gravitational mapping.)
It depends on where the masscon is in relation to the orbit, and just how big the masscon is, as well as how carefully the orbit is measured, and whether or not anyone is actually looking.
If it's a polar impact and an equatorial orbit, it would need to move the center of mass appreciably to be noticed at all.
If it's an equatorial impact and a polar orbit, it might be noticed in as few as a single orbital path as the path changes during a direct flyover.
Further, it requires someone be actually looking for the changes in orbit.
Technically, it also propagates the gravitational waves at C, but the distances involved in this case are small enough that it is (if you'll pardon the pun) immaterial.
You need to be close enough to it for long enough to alter the orbit enough that it's detectable. And it is particularly helpful to make multiple passes; the current systems need several orbits and look at the statistics to generate the topological mapping which is turned into the density maps.
