What is the speed of gravity?
Dear Straight Dope:
If the Earth's sun was to lose enough mass to affect the gravity field of the solar system, would the new gravitational effect on Earth be felt instantaneously or would it take 8 minutes or so (the light travel time from Earth-sun) for the new equilibrium to occur? --Scott C., West Virginia
A few months ago I read a short article about how some scientists were able to measure the "speed" of gravity, but the article did not go into much detail and I never saw anything else about it. So tell me, how did they measure gravity's speed, and does speed vary the same way force does, i.e., would the speed be different on different planets?
Now, this is a topic I can really sink my teeth into. Is everyone comfortable?
First of all, there's no such thing as "instantaneous" transmission of anything. Presumably, that would mean that at the exact same time that the sun disappeared, the planets would feel it (or rather, stop feeling it). The problem is that you can't define "same time," unless it's also the same place. If, according to an observer at rest relative to the sun, the two events happened at the same time, then an observer moving towards the sun from the Earth would say that the sun disappeared before the Earth felt the effects, and an observer moving in the opposite direction would say the reverse: that the Earth actually felt the effects before anything happened to the sun, which most folks would agree is ludicrous. This may strike you as one of those silly relativistic paradoxes--perhaps we can't detect instantaneous events, but surely they can sure occur, can't they? Sorry, but no, they can't. In the Einsteinian universe, instantaneity--that is, event X instantaneously causes effect Y at distance Z--is impossible in a meaningful sense, since for reasons we needn't delve into it would involve communication at faster-than-light speeds. As you suspected, gravitational effects travel at exactly c, just like light, so it would take eight minutes for us to know.
This inspires two obvious questions: (a) How do we know gravity travels exactly at c? Answer: We don't, but it seems consistent with other things we know. To put it in technical terms, general relativity is only self-consistent if the speed of gravity is Lorentz invariant (that is, the same for all observers), and the only speed for which that works is c. (b) If gravity travels at the same speed as light, is it propagated by waves and particles, as light is? Answer: It seems likely. In fact, the gravitational particle has already been given a name: the graviton. But even though Einstein's theory of general relativity, which posited gravitational waves, was published in 1916, no unambiguous instance of gravitational waves or particles has ever been directly detected. Indirect evidence for gravitational waves is very strong, however, such as the observations of the binary pulsar, which won the 1974 Nobel prize.
The search for the gravitational medium continues. While it's true that matter can't just disappear (detection of gravitational waves might be simpler if it did), it can wiggle around. Picture a pair of parallel wires, with a bead on each one. Since the beads are attracted to one another by gravity, they'll settle into a position such that they're as close together as they can get, right next to each other on the wires. Now suppose that you grab bead A and move it down the wire. Bead B will be attracted to it, and move closer to it again, but it won't do so immediately, because of the speed limitation mentioned above. If you wiggled bead A, you could detect the effect on bead B, transmitted gravitationally. This is all very well theoretically, of course, but it's not practical because gravity is an extremely weak force, and variations of the magnitude necessary to prove the existence of gravitational waves are undetectable except at planetary (or larger) scale.
On the other hand, what if instead of wiggling bead A, you have a pair of stars orbiting each other, and instead of bead B, you've got three masses in a huge vacuum chamber on the Earth, or in orbit around the sun? Then you've got a gravitational wave observatory, like LIGO or LISA, which promise to give us some very real, very usable information about the cosmos. The LIGO detector may well have detected gravitational waves directly by now, but the results are still somewhat ambiguous. LISA is almost guaranteed to have unambiguous observations within a few seconds of it being turned on, but that's about a decade in the future.