Speed of Gravity

[Sommers]

I recently read a story on the web stating that the speed of gravity was recently calculated to be 6X10E+18 m/s. Thats amazingly fast (.006 s to the Alpha Centauri system and 158 s accross the Milky Way). Is this story **?

[publius]

Yes, it's pretty much **. While, to be correct, the speed of gravitational propagation has never been yet directly measured experimentally (there's a long story there), according to General Relativity, our best theory of gravity, the speed of gravitational propagation is c, the speed of light.

The reason for the stuff you read about gravity being much faster than light is due to some fundamental misunderstandings. Gravity has some little tricks up its sleeve that compensate for the effects of propagation delay. It's not perfect, but it is much better than what happens if you naively add a propagation delay to Newtonian gravity.

This latter business is where the claims about gravity being orders of magnitude faster than light come from. It is simply wrong. :)

-Richard

[Sommers]

I really wanted to believe this one.

I guess I can scratch off my follow up questions.

[Jens]

Just as an aside, I don't think there's anything to be bummed out about. If you want to understand how gravity works, it's an excellent place to start. You heard something and heard it's wrong. Now you might be interested in exploring how planetary orbits are stable even though gravity is not transmitted instantaneously (they would not be in Newtonian gravity with a delay). It's actually a very interesting issue.

[Sommers]

I'm trying to grasp how they would'nt be in a Newtonian delay. Couldn't the planets just constantly be encountering the gravitational field that originated a moment prior?

.........In trying to understand this better, I am envisioning a ball on a string being swing in orbit by a person. The string (tension force) holding the ball in orbit doesn't curve but remains taught and straight. Is this somehow the same concept that would explain the gravitation force pointing to the source's true real-time location? Or is their a minute lateral curvature in that string (and the associated path of the tension force)?

[cosmocrazy]

Don't be to dis-heartened, as Publius pointed out, based on GR and our understanding of the implications regarding FTL then yes gravity propagation is believed to travel at C. But, this does not mean the case is closed. The propagation has not been directly measured so FTL can not be totally ruled out. However, GR is the best mainstream model we have that fits the observational evidence we have so it appears that gravity must follow the laws of relativity also.

[swampyankee]

According to at least one report (here (http://physicsworld.com/cws/article/news/16644)), the speed of gravity has been measured, although I believe that is has not been confirmed.

[cjameshuff]

I'm trying to grasp how they would'nt be in a Newtonian delay. Couldn't the planets just constantly be encountering the gravitational field that originated a moment prior?

With Newtonian gravity, planets would be accelerating toward the point where the sun was a short time before, and the sun accelerates toward the points where they used to be. This leads to orbits spiraling in or out. GR gravity has velocity dependent characteristics that compensate for this.



The propagation has not been directly measured so FTL can not be totally ruled out.

It has been measured. And while it hasn't been measured with great precision and certainty, other measurements related to relativity make it pretty certain that FTL propagation of gravity would allow information to be sent back in time, allowing for causality violation. This is considered very unlikely.

[publius]

According to at least one report (here (http://physicsworld.com/cws/article/news/16644)), the speed of gravity has been measured, although I believe that is has not been confirmed.


That was part of the "long story" I mentioned above. It's long, and I'll try to make it short. :) Kopeikin and Fomalont claimed to have measured the "speed of gravity" via measurement of the "bending" (Shapiro delay more properly) of light from a quasar around Jupiter.

Now, Clifford Will and Steve Carlip, two "high priests" of GR as I like to call them, and two guys I trust on things well beyond my own mathematical abilities, have pointed on the flaws in Kopeikin's work.

It's very subtle and complicated. Basically, in the rather complex equations of GR, the parameter 'c' is in there all over the place. Now, how do you separate the one that accounts for the speed of gravity, vs the one that is the regular speed of light (present in space to time conversion and energy to mass relations)?

That is not obvious. :) As Will showed in a paper, Kopeikin simply made a mistake and varied 'c' in the wrong place.

Kopeikin's result, IIRC, has a speed of gravity term proportional to first order in (v/c). That should be a clue right there, as no relativistic (deviations from classical behavior) effects occur at anything less the second order there.

Second, gravity "compensates" for the propagation delay to high order, and one would intuitively expect that if does this, then it must do the same thing with any effect due to moving sources. It has to be to be consistent. And we know propagation delay effects are related to gravitational radiation, which comes in at order (v/c)^5. So, we'd expect that any "speed of gravity" effects will only be apparent when (v/c)^5 is significant. It ain't for Jupiter by a long shot, well below the threshold of measurement.

-Richard

[publius]

I didn't want to ramble on so much, but for completeness, more can be said about Kopeikin's purported speed of gravity experiment.

From SR, we know 'c' is special. That's the speed that is the same for all frames. Speeds less than 'c' in one frame will be different in other frames. Thus, if the speed of gravity, call it c_g were less than c, it would appear to propagate at different speeds in different frames.

If c_g > c, we'd have the same thing, but with causality problems to boot. As we've discussed in long threads about causality, relativity, and FTL, the only way information move faster than light and still have the Lorentz transform hold (in a mechanical way) is to have a absolute frame.

It turns out that any difference between c_g and c, greater or less, translates in what the "high priests" call "preferred frame effects" of gravity -- IOW, gravity would not be invariant in the way it must be to compatible with the base principle of relativity. In the PPN framework (weak field post Newtonian framework for experimental gravitational work), this translates into one of the parameters that is zero for GR, but non-zero for any gravity with preferred frame characteristics.

This parameter has been constrained by precise solar system observations over the years to some very small value, actually. Anyway, when you look at that, it's obvious that any such effects would well below the "signal to noise" of Kopeikin's measurements. Well below.

Kopeikin simply made a mistake (nothing to be ashamed of there -- this ain't romper room 'rithmetic), but he refuses to admit his mistake and stubbornly insists he did measure the speed of gravity.

And for the record, this does not mean than anybody thinks the speed of gravity is different from c. This is just about getting the math and experiments right.

-Richard

[WayneFrancis]

Thanks for being complete publius, very good info.

[sirius0]

Am I correct in thinking to myself that if we detect gravity waves then we would be able to say something quite firm about the speed of gravity? I assume the several detectors we have set up at the moment have been configured on the assumption of gravity being at c. So if it wasn't c, then if we had only one detector, we may not know that we were measuring an alias of the true wave but because we have a few aliasing would give different results for each detector. But if it all proves consistent then the gravity wave would demonstrate the speed of gravity, would it?


Sorry I think my grammar switch might be off!

[WayneFrancis]

The gravity wave detectors as I understand them are actually multiple detectors linked at a distance, they intrinsically will measure the speed of gravity if they detect the waves.

[publius]

Yes, if (I hope to say when) gravitational waves are directly observed, that will pin down the speed, in the same way electromagnetic radiation (light) pins down the "speed of electricity". :)


-Richard

[sirius0]

Yes, if (I hope to say when) gravitational waves are directly observed, that will pin down the speed, in the same way electromagnetic radiation (light) pins down the "speed of electricity". :)


-Richard
Yes I agree, when we detect the waves.

Not that I am one qualified to assess as yet, but I would like to add Kip Thorne to those giants mentioned earlier.

What I meant about aliasing earlier is that if gravity has a different speed then calculations made for gravity wavelengths from the same event will be different for different detectors and the speed would have to be corrected in order to get the differing results to match. I don't think this will be the case at all. But the fact that this could be done if the detectors did disagree in itself lends weight to a consistent result for g @ c when the waves are detected.

[astromark]

Sorry my sensible switch is broken... Please do not take me Sirius0 ly...:)
If gravity waves... wave back.

This conversation and question has been dealt with not to long ago... and a reasoned argument was conducted with a result not agreed apon...:boohoo:

Light at c. Gravity most probably at c. It seems like and, most likely is... No argument for change has been found.:neutral:

[trinitree88]

Am I correct in thinking to myself that if we detect gravity waves then we would be able to say something quite firm about the speed of gravity? I assume the several detectors we have set up at the moment have been configured on the assumption of gravity being at c. So if it wasn't c, then if we had only one detector, we may not know that we were measuring an alias of the true wave but because we have a few aliasing would give different results for each detector. But if it all proves consistent then the gravity wave would demonstrate the speed of gravity, would it?


Sorry I think my grammar switch might be off!

sirius0. The coincidences seen during SN1987a between the neutrino detectors at the IMB, Mont Blanc, Kamiokande, and Baksan neutrino detectors, and the Rome and Maryland bar gravitational wave detectors peer reviewed and published ~ 25 times indicate that the speed of the gravitational waves matched that of the neutrinos...(also c ) to less than two seconds difference over a light path of ~ 170,000 light years (the distance to the Large Magellanic Cloud). The timing uncertainty is likely due to the fact that some of the neutrino detectors were not at the time set to universal time, leaving the distinct possibility that they are in fact identical.
Regardless, off by two parts in.approximately 5,364,645,120,000 seconds......................................... is pretty small. At least one theory predicted this effect.

[jwestern]

I appreciate this discussion on the "Speed of Gravity". There was reference to research, but no specific materials were identified. Can any of you suggest a few good reference material documents on the topic. There seem to be a general agreement the speed of gravity is the speed of light, but maybe less. In our local universe, I assume the gravitons of all mass are in touch with one another. If one mass moves all of the other masses know. As we move farther and farther away from our local universe, at what point (distance or velocity) does graviton communication between masses stop. As the universe is expanding faster than speed of light, it seems to me there will be no graviton communication with masses outside of our local universe (distant galaxies). Personally, I like the thought of gravitons of all mass in the entire universe being in constant communication with one another. If one mass moves anywhere in the universe, all of the other mass gravitons know its movement. Thank you for recommending any thoughtful reference materials on the topic.

Jeff

[sirius0]

sirius0. The coincidences seen during SN1987a between the neutrino detectors at the IMB, Mont Blanc, Kamiokande, and Baksan neutrino detectors, and the Rome and Maryland bar gravitational wave detectors peer reviewed and published ~ 25 times indicate that the speed of the gravitational waves matched that of the neutrinos...(also c ) to less than two seconds difference over a light path of ~ 170,000 light years (the distance to the Large Magellanic Cloud). The timing uncertainty is likely due to the fact that some of the neutrino detectors were not at the time set to universal time, leaving the distinct possibility that they are in fact identical.
Regardless, off by two parts in.approximately 5,364,645,120,000 seconds......................................... is pretty small. At least one theory predicted this effect.
Hi trinitree88,
Could you perhaps link to a paper about these gravitational waves detected. I was of the impression that they had not been detected as yet. I am looking forward to this first detection and would hate to think I had missed it! I assume you involve neutrinos because they were emitted by the same event? If so this would be very strong evidence for gravity at c.

[Cougar]

Hi trinitree88,
Could you perhaps link to a paper about these gravitational waves detected.

I would also be interested in that. Here (http://www.springerlink.com/content/13k308x245152rj3/) is the abstract of one paper, but you can't get to the actual paper without paying or subscribing, and it doesn't come up anywhere else. If this 'correlation' was a bit stronger, I imagine it would have been splashed all over the news. Still, intriguing....

[undidly]

I appreciate this discussion on the "Speed of Gravity". There was reference to research, but no specific materials were identified. Can any of you suggest a few good reference material documents on the topic. There seem to be a general agreement the speed of gravity is the speed of light, but maybe less. In our local universe, I assume the gravitons of all mass are in touch with one another. If one mass moves all of the other masses know. As we move farther and farther away from our local universe, at what point (distance or velocity) does graviton communication between masses stop. As the universe is expanding faster than speed of light, it seems to me there will be no graviton communication with masses outside of our local universe (distant galaxies). Personally, I like the thought of gravitons of all mass in the entire universe being in constant communication with one another. If one mass moves anywhere in the universe, all of the other mass gravitons know its movement. Thank you for recommending any thoughtful reference materials on the topic.

Jeff

""Personally, I like the thought of gravitons of all mass in the entire universe being in constant communication with one another. If one mass moves anywhere in the universe, all of the other mass gravitons know its movement.""

I like it too but wonder if the gravitons communicate instantly or at C.
So far no one has detected a graviton or a gravitational wave.

[sirjon]

I believe it is more important to 'know' first how gravity 'works' and from it everyone would agree the mesurement of its speed, if gravity is really some sort of a 'force' like the other three fundamental forces of nature. I reserve my belief that it isn't a force.
(Care for personal opinion? I'll appreciate if you do).

[undidly]

I believe it is more important to 'know' first how gravity 'works' and from it everyone would agree the mesurement of its speed, if gravity is really some sort of a 'force' like the other three fundamental forces of nature. I reserve my belief that it isn't a force.
(Care for personal opinion? I'll appreciate if you do).

""I reserve my belief that it isn't a force.""

Me too.

It is not a force just like centrifugal (centripetal) force is not a force.

[sirius0]

Thanks for the clue Cougar!
Below is a paper that re-visits the event.

arXiv:0810.3759v1 [gr-qc] 21 Oct 2008 (http://www.google.com.au/url?sa=t&source=web&ct=res&cd=5&ved=0CB4QFjAE&url=http%3A%2F%2Farxiv.org%2Fpdf%2F0810.3759&rct=j&q=SN+1987+A%3A+Correlations+between+the+Maryland+a nd+Rome+gravitational+wave+detector+data+and+the+M ont+Blanc+and+Kamiokande+neutrino+detector+data&ei=Cj8aS7uNK46gkQXd0tDNAw&usg=AFQjCNFKR7-gJ38qYu117DsckP5Nrf-gWA)



Just going to have to waddle through this now...

[sirius0]

This looks interesting too.

www.aei.mpg.de/~schutz/papers/schutz_60270.pdf (http://www.google.com.au/url?sa=t&source=web&ct=res&cd=10&ved=0CDAQFjAJ&url=http%3A%2F%2Fwww.aei.mpg.de%2F%7Eschutz%2Fpape rs%2Fschutz_60270.pdf&rct=j&q=SN+1987+A%3A+Correlations+between+the+Maryland+a nd+Rome+gravitational+wave+detector+data+and+the+M ont+Blanc+and+Kamiokande+neutrino+detector+data&ei=Cj8aS7uNK46gkQXd0tDNAw&usg=AFQjCNHi4x_gZcXC2tVYmTGZGH9N264cOg)

[JCoyote]

If gravity waves... wave back.

Wait a second, is it my hand that's waving back, or the space my hand is in that's waving back?

Sorry, had to say it. :lol:

[Hlafordlaes]

From SR, we know 'c' is special. That's the speed that is the same for all frames. Speeds less than 'c' in one frame will be different in other frames. Thus, if the speed of gravity, call it c_g were less than c, it would appear to propagate at different speeds in different frames.

If c_g > c, we'd have the same thing, but with causality problems to boot. As we've discussed in long threads about causality, relativity, and FTL, the only way information move faster than light and still have the Lorentz transform hold (in a mechanical way) is to have a absolute frame.

It turns out that any difference between c_g and c, greater or less, translates in what the "high priests" call "preferred frame effects" of gravity -- IOW, gravity would not be invariant in the way it must be to compatible with the base principle of relativity. In the PPN framework (weak field post Newtonian framework for experimental gravitational work), this translates into one of the parameters that is zero for GR, but non-zero for any gravity with preferred frame characteristics....

Thanks for that, I think I (almost) understood it. Surprised me I was able to. Have you authored any physics books/articles for laymen (read:dopes)?

[Hornblower]

""I reserve my belief that it isn't a force.""

Me too.

It is not a force just like centrifugal (centripetal) force is not a force.

Centrifugal and centripetal forces are two different things. Which one do you mean?

[trinitree88]

Hi trinitree88,
Could you perhaps link to a paper about these gravitational waves detected. I was of the impression that they had not been detected as yet. I am looking forward to this first detection and would hate to think I had missed it! I assume you involve neutrinos because they were emitted by the same event? If so this would be very strong evidence for gravity at c.

sirius0. Sorry for the late reply...try post #31 in this thread:http://www.bautforum.com/space-astronomy-questions-answers/35564-mass-speed-light-2.html

there's about 13 of ~ 20 articles listed. Pete

[Hornblower]

I believe it is more important to 'know' first how gravity 'works' and from it everyone would agree the mesurement of its speed, if gravity is really some sort of a 'force' like the other three fundamental forces of nature. I reserve my belief that it isn't a force.
(Care for personal opinion? I'll appreciate if you do).

In my admittedly non-expert opinion, your reasoning is backward. If we succeed in measuring the speed of propagation of some aspect of gravity, it might help in firming up our "knowledge" of how gravity "works".

An analogy: Our ancestors long ago found ways to measure the speed of light, long before anyone knew about its electromagnetic properties, not to mention its quantum-mechanical properties.

[astromark]

As is often the case... where is the argument here ?
Has a question been asked ?
I can not justify an argument for gravity working at a quicker rate than instant. Instant in this universe would seem to mean at c. We have no example of any other fact. Is there one ?
How does it work ? Hmmm... Ummm... I do not know.
What is it ? Ummm... Errr... and other than repeating what we all know...' That matter seems to attract matter.' and that. 'Mater has mass proportional to its mass' Just in case I am being misunderstood... A Black hole has a very strong gravitational signature because it has a great deal of mass.
Then we come to... It just is.

[cosmocrazy]

I'd have to agree with you on this Astro, If Relativity is correct, which currently it seems so, then instant in this universe means at C. For me based on this, the concept of faster than C is like talking about beyond Infinity or below zero, in the physical sense of things that is. They just don't have any real definition within this universe.