Science Questions

How fast does gravity propagate?

Tue, 7th Jan 2014

Listen Now    Download as mp3 from the show Why don't microwaves spark off themselves?


Peter Patay asked:



If you could instantly switch off the sun's gravity, how long would it take for the earth to revert to a galaxy centered orbit?


During the 8 miniutes that the zero gravity wave travels to earth does the effect of the sun's gravity gradually diminish or will it be instant on arrival?


Peter Patay



Dominic - So yes, gravity is this force that binds everything in the universe together. A pulsarThere's this quite fundamental principle in physics that nothing can travel faster than the speed of light. And so, physicists have rather suspected on the basis of that that gravity must propagate at a finite speed. At the speed of light, because if it happened instantaneously then wobbling something from side to side in the universe would essentially be propagating information about how that thing was moving faster than the speed of light and violating this very fundamental principle.

Itís been obviously quite hard to test because trying to find some experimental setup where you test whether gravity propagates faster than the speed of light, is really quite a challenge. But actually, in the last 10 years or so, we have done that with objects called pulsars which are very compact neutron stars. They are basically the mass of a star in the size of a mile or two across. Some of these things are very close to one another and spinning around each other very fast. Actually, Einsteinís theory of general relativity which is the best description of gravity we have, predicts that when these things are orbiting very fast, they should produce what's called gravitational waves which are ripples of gravity, that travel out at the speed of light. If they're doing that, they should be gradually losing energy through these gravitational waves. In fact, we have found pulsar binary pairs that seem to be gradually getting closer and closer together as if they're losing energy, at exactly the rate that Einstein predicts, if gravity travels at the speed of light.

Chris - Mark...

Mark Peplow - As the pulsars change their rate, you're sort of inferring the existence of gravity waves. What would it take to detect the gravity waves themselves?

Dominic - Well, there are a number of teams around the world who are trying to build detectors to detect these ripples of gravity moving through space. The sensitivity you need to that is absoloutely mind boggling. You're talking about distances of about a mile that you're sending light beams down and you're trying to see whether gravity is causing that distance to ripple by about the size of an atom. So, you've got an experimental setup a mile long and you're trying to detect something that's moving by the width of an atom.

No one has yet detected those gravitational waves. They are, I think getting quite close incredibly. I'm always quite incredulous when I hear about these experiments because they sound bonkers to me. But I think in the next decade or so, we might actually start to detect these ripples in space time.


Subscribe Free

Related Content


Make a comment

I see no reason why gravity would diminish gradually if you had switched it off instantly.

The fact that GR tells us that gravity is spacetime curvature caused by the presence of mass/energy, one would have to say good luck with your efforts to remove the sun instantly! Bill S, Sat, 3rd Aug 2013

Gravitational effects propagate at the speed of light. Pmb, Sun, 4th Aug 2013

I think Peter may have gone back to sleep now.

If the orbit instantly changes would there be some effects of inertia ?

I think the surface of this planet is travelling at over 60000 mph. :o 1stmoondog, Mon, 5th Aug 2013

I think it would take some time for the Earth to revert to a purely galaxy-centered orbit. We're going pretty fast around the sun and we'd continue to move in whatever direction the sun left us in regardless of what a stable galaxy-centered orbit would be.

I'd also imagine that nearby encounters would have more impact on our trajectory than the centre of the galaxy. I could be wrong.

Having said all that of course, we are in a galaxy-centered orbit at the moment. The same way the moon is going around the sun. bizerl, Tue, 6th Aug 2013

Having more thoughts - if the problem was shrunk down and it was the Earth that disappeared, I guess the moon would still orbit the sun, however a stable orbit may never quite be acheived. (?)

Sorry, I think I've just clogged this thread with two highly unscientific opinions. bizerl, Tue, 6th Aug 2013

heh, I like the spider net analogue myself, in where the net is there as long as energy mass exist, well hopefully :) But where the 'propagation' of change goes with lights speed. Then again, it's all observer dependent, what you see, even with a 'net'. The world of headache. yor_on, Tue, 6th Aug 2013

I believe that matter is captive energy, and its gravity is the gravity of that energy. Gravity propagates much faster than light, but you can't make matter instantly disappear without converting it to another form of energy.

Suppose some process could instantly convert the Sun's mass into a uniform spherical shell of electromagnetic energy, expanding at the speed of light. At any point outside that expanding shell, the gravity of the shell's energy would be exactly equal to what had been the gravity of the Sun's mass. At any point inside the shell, the net gravity due to the shell's energy would be zero (as explained by Newton's shell theorem). So the planets would continue to orbit as before until the instant the shell of light passed them. Eight minutes after the sun's mass suddenly became light, Earth would be hit by a terrible flash of light; then the sky would go completely dark (except for other stars), and the sun's gravity would disappear at that instant.

If anyone on the night side of Earth survived the cataclysm, they would see the outer planets still in their original orbits until, one by one, they are illuminated by the shell of light. As the shell of light passes each planet, it shines blindingly for a few seconds before going dark. The planet then stops accelerating toward where the Sun had been and goes off on a tangent.

So there would be an eight-minute delay before Earth would stop feeling the Sun's gravity, but that is not evidence of gravity propagating at the speed of light. If gravity did propagate at the speed of light, the Sun's gravity would pull us in the direction where we see the Sun, not the direction where the Sun is. Therefor, we would be pulled forward into a higher and higher orbit and eventually ejected from the solar system. The very existence of solar systems is proof that gravity propagates many times faster than light. Tom Van Flandern calculated that gravity has to be at least twenty billion times faster than light to explain the lack of forward acceleration of planets. While I poo poo some of his wild theories, I believe he was right about the speed of gravity.

If instead of a uniform spherical shell of light, the Sun were converted to a pair of equal and opposite photon torpedoes, the shell theorem would not be applicable. At distances many times the light travel distance since the event, the gravity would be equal to that of the Sun's mass at the center between the two photon torpedoes. But as the angular size of the separation between the photon torpedoes, becomes significant, the gravity would have to be calculated by some formula which I don't know. I leave that to the mathematicians. I suspect an observer would be strongly attracted to a photon torpedo passing near him, while the other one would have negligible effect. The gravity would gradually increase until the photon torpedo passes the observer, and then it would gradually diminish. The observer would not see the photon torpedo unless and until it hits him; light is not an emitter of light, but I believe it does have a gravitational field.

Note: Newton's shell theorem tacitly assumes that gravity propagates at infinite speed. Otherwise, there would have to be a light-speed delay factor in the proof of the theorem. The absence of such a factor has no effect as long as the shells are static. If the shells expand or contract, a speed of gravity factor would be needed to prove the validity of the theorem. Phractality, Wed, 7th Aug 2013

>No one has yet detected those gravitational waves. I guess this is now obsolete denshade, Sat, 13th Feb 2016

Sorry, but relying on the 'neutron star' explanation for pulsars is a flawed theoretical concept. There is a far better theory for pulsars that does not rely on massive objects rotating at crazy 43,000rpm speeds. Let's have another proof for gravity speed please. Cigarshaped, Wed, 6th Jul 2016

The results of the gravitational wave detection announced in 2016 showed a delay of about 6ms between the two detectors (about 3000 miles apart).
This is consistent with gravitational waves traveling at about the speed of light.

More accurate confirmation will take a few more active detectors. evan_au, Thu, 7th Jul 2016

See the whole discussion | Make a comment

Not working please enable javascript
Powered by UKfast
Genetics Society