# Thread: speed of gravitational waves

1. Established Member
Join Date
Jan 2010
Location
Wisconsin USA
Posts
2,533

## speed of gravitational waves

When will LIGO confirm the speed of gravity relative to light or what will it take?

2. Order of Kilopi
Join Date
Oct 2005
Posts
25,837
I presume it will take enough stations to over-constrain the direction of the source, and then they should be able to use correlations in the stations to determine the speed of the wave between them. That would seem to require 4 stations, because you need 3 to get the direction.

3. Established Member
Join Date
Jan 2010
Location
Wisconsin USA
Posts
2,533
Originally Posted by Ken G
I presume it will take enough stations to over-constrain the direction of the source, and then they should be able to use correlations in the stations to determine the speed of the wave between them. That would seem to require 4 stations, because you need 3 to get the direction.
Could this then be the big news?

4. Originally Posted by Copernicus
Could this then be the big news?
If I am not mistaken, the theory predicts that these waves propagate at the speed of light. Thus I would expect it to be big news if we found some other velocity, which would force us back to the theoretical drawing board.

5. Like, nothing odd so far, Bounding the speed of gravity with gravitational wave observations

Using a Bayesian approach that combines the first three gravitational wave detections reported by the LIGO collaboration we constrain the gravitational waves propagation speed c_gw to the 90% credible interval 0.55 c < c_gw < 1.42 c, where c is the speed of light in vacuum.
[...]
Of order twenty detections by the two LIGO detectors will constrain the speed of gravity to within 20% of the speed of light, while just five detections by the LIGO-Virgo-Kagra network will constrain the speed of gravity to within 1% of the speed of light.
Last edited by slang; 2017-Oct-10 at 05:32 AM. Reason: fixed URL

6. Originally Posted by Copernicus
Could this then be the big news?
No, because the last "big news" was that there are now three detectors online.

7. Order of Kilopi
Join Date
Mar 2010
Location
United Kingdom
Posts
6,672
Originally Posted by Hornblower
If I am not mistaken, the theory predicts that these waves propagate at the speed of light. Thus I would expect it to be big news if we found some other velocity, which would force us back to the theoretical drawing board.
Not just theory - indirect measurements also support this, so it would be interesting if direct measurements then don't.

8. Established Member
Join Date
Jan 2010
Location
Wisconsin USA
Posts
2,533
Originally Posted by Shaula
Not just theory - indirect measurements also support this, so it would be interesting if direct measurements then don't.
Iwould expect a direct measurement that it is the speed of light would be big news.

9. Originally Posted by Copernicus
Iwould expect a direct measurement that it is the speed of light would be big news.
I disagree. That would merely be one more item in good agreement with GR, which has been upheld repeatedly for a century. I would expect something not in agreement, like those reputed superluminal neutrinos, to be a bigger news item.

10. Established Member
Join Date
Jun 2009
Posts
2,131
Originally Posted by Ken G
I presume it will take enough stations to over-constrain the direction of the source, and then they should be able to use correlations in the stations to determine the speed of the wave between them. That would seem to require 4 stations, because you need 3 to get the direction.
I'm not quite sure how that would work. The triangulation from the first 3 stations assumes the speed is c (which I'm willing to bet it is). How do we use a 4th station to measure the speed without assuming a speed for the first 3?

Seems to me that if we could triangulate events and then match them up to some visible phenomena(GRB, etc...), then that would confirm that the assumed speed is correct.

11. The "big news" to be announced on Oct 16 is very, very, very likely to describe the results of the gravitational wave detection of a merging pair of neutron stars, _and_ the followup observations in optical, radio, X-ray and IR. I would guess that an electromagnetic counterpart was detected in at least one case, as that would make the event much more exciting.

You can read a bit about what we know of the followup observations at

http://spiff.rit.edu/classes/ast613/...tro/intro.html

12. Order of Kilopi
Join Date
Oct 2005
Posts
25,837
Originally Posted by ShinAce
I'm not quite sure how that would work. The triangulation from the first 3 stations assumes the speed is c (which I'm willing to bet it is). How do we use a 4th station to measure the speed without assuming a speed for the first 3?
I presume by checking consistency. That's often how one escapes the need to assume something-- if you make the assumption for 3, and the 4th doesn't agree, you can correct your assumption.
Seems to me that if we could triangulate events and then match them up to some visible phenomena(GRB, etc...), then that would confirm that the assumed speed is correct.
Sure, but there's no guarantee there will be a visible phenomenon, though there is a rumor that neutron star mergers might come with counterparts. But even then, you don't know exactly when the light is emitted, for example with supernovae it is delayed from the neutrino burst. With 4 detectors, I would think you can test the speed in a single event with no visible counterpart, though the precision of the test might not be as good as you can do with a distant source that does have a visible counterpart. In any event, since there are only 3 now, the 4-site approach might be superceded if the rumors are true.
Last edited by Ken G; 2017-Oct-08 at 01:39 AM.

13. Established Member
Join Date
Jan 2010
Location
Wisconsin USA
Posts
2,533
Originally Posted by StupendousMan
The "big news" to be announced on Oct 16 is very, very, very likely to describe the results of the gravitational wave detection of a merging pair of neutron stars, _and_ the followup observations in optical, radio, X-ray and IR. I would guess that an electromagnetic counterpart was detected in at least one case, as that would make the event much more exciting.

You can read a bit about what we know of the followup observations at

http://spiff.rit.edu/classes/ast613/...tro/intro.html
That would be awesome!

14. Established Member
Join Date
Jun 2009
Posts
2,131
Originally Posted by Copernicus
That would be awesome!
That would be awesome, but I wouldn't count on it. The detections thus far have all been black hole binaries where the BH's had around 25 solar masses. The second observation is a smaller pair, but also a much noisier signal. If it is a neutron star merger, it needs to be much closer because LIGO is 'tuned' to frequencies matching larger objects.

Reminds me of the initial extrasolar planet discoveries. A bunch of hot jupiters.

15. NGC 4993 _is_ much closer than the previous black-hole binary mergers. It's only about 40 Mpc away.
Last edited by StupendousMan; 2017-Oct-08 at 10:13 PM.

16. Contrast:

GW150914 440 (+160/-180) Mpc
GW151226 440 (+180/-190) Mpc
GW170104 880 (+450/-390) Mpc
GW170814 540 (+130/-210) Mpc
NGC 4993 40 Mpc

Edit to add: On tuning, LSC News, 2017 July 7

The average reach of the LIGO network for binary merger events has been around 70 Mpc for 1.4+1.4 Msun, 300 Mpc for 10+10 Msun and 700 Mpc for 30+30 Msun mergers, with relative variations in time of the order of 10%.
Last edited by 01101001; 2017-Oct-08 at 11:50 PM. Reason: Add size/distance reach

17. Order of Kilopi
Join Date
Oct 2005
Posts
25,837
So if those statistics were to hold up in the long run, it means that last one is of a type that is some 250 times more common but has a gravitational signal some 10 times weaker. Apparently that would be how neutron-star mergers compare to black-hole mergers.

What I don't understand is for the black hole mergers themselves, which is, where are the 5 solar-mass black holes? I mean, if we assume the energy released in gravitational waves is proportional to the initial mass of the star, then the fact that the distance from which the source can be seen scales as the square root of the energy means that the volume over which it can be seen scales as energy to the 3/2 power. Meanwhile, the Salpeter initial mass function says that the number of stars with initial mass above M scales as M-1.35. Combining these means we expect to see a mass distribution that scales like the product of these, or M0.15, meaning very little M dependence. So we should be seeing a mass distribution that is fairly flat, not 4 biggies and 1 smallie, if those assumptions are correct.

Of course the data is still sparse, but there might be something starting to flesh out here that suggests the assumptions I gave are not correct. It might be that the more massive systems are the ones that are more likely to merge, since they might interact more with each other, or it might be that the Salpeter IMF didn't apply at the early ages of the universe that we can see out to for the high-mass guys. Either of those might not be so surprising, but even so, astronomers are not very used to talking about stars that give rise to 50 solar mass black holes-- they're oddballs among what we are used to seeing in our neck of the woods.
Last edited by Ken G; 2017-Oct-09 at 12:05 AM.

18. Order of Kilopi
Join Date
Aug 2008
Location
Wellington, New Zealand
Posts
3,668
Originally Posted by Copernicus
When will LIGO confirm the speed of gravity relative to light or what will it take?
That was done with the first detection announced in 2016: Speed of gravity - Direct measurements of gravitational waves
In 2016: On constraining the speed of gravitational waves following GW150914
Earlier this year: Bounding the speed of gravity with gravitational wave observations

Note that the papers give a wide range of values, e.g. 0.55 c to 1.42 c. Better bounds will come with more detections and more detectors.

19. Established Member
Join Date
Jun 2009
Posts
2,131
Ken: We're detecting larger(~50 solar masses) mergers because LIGO is most sensitive to those corresponding frequencies. Again, we're detecting 'hot Jupiters' because that's what we're looking for.

20. Established Member
Join Date
Jan 2010
Location
Wisconsin USA
Posts
2,533
Originally Posted by Reality Check
That was done with the first detection announced in 2016: Speed of gravity - Direct measurements of gravitational waves
In 2016: On constraining the speed of gravitational waves following GW150914
Earlier this year: Bounding the speed of gravity with gravitational wave observations

Note that the papers give a wide range of values, e.g. 0.55 c to 1.42 c. Better bounds will come with more detections and more detectors.
I think that is a great find by you Reality Check. Don't know how to interpret this for sure, but it seems it must be around 1.7 sigma range.

21. Originally Posted by ShinAce
Ken: We're detecting larger(~50 solar masses) mergers because LIGO is most sensitive to those corresponding frequencies. Again, we're detecting 'hot Jupiters' because that's what we're looking for.
Are you sure that's true for Advanced LIGO? The informal chart at Caltech: Gravitational Wave Spectrum seems to indicate that neutron-star binaries were on the edge of detectability with original LIGO, but that they are now part of the expected events for Advanced LIGO. Now (or before too long), it is neutron-star asymmetries (not pairs) that are at the edge of detection.
Last edited by 01101001; 2017-Oct-09 at 02:00 PM. Reason: Covered for not knowing exact Advanced LIGO improvement progress

22. Order of Kilopi
Join Date
Oct 2005
Posts
25,837
Originally Posted by ShinAce
Ken: We're detecting larger(~50 solar masses) mergers because LIGO is most sensitive to those corresponding frequencies. Again, we're detecting 'hot Jupiters' because that's what we're looking for.
Ah, there's not just a wave amplitude issue, but also a frequency issue. So that's something I left out. But it seems strange the frequency sensitivity would work out as you say. The mirrors are 4 km apart, so that means they are most sensitive to detecting gravitational wave sources that are of similar scale, since we expect the merger to give maximum signal when moving close to the speed of light, and in any case cannot move faster than that. But the Schwarzschild radius is about 3M km, where M is in solar masses, so I would have thought 4 km would be well suited to finding neutron-star mergers, and quite a bit small for finding 50 solar mass black hole mergers. So that looks to make the situation even worse, in terms of why we are finding such huge and rare objects, unless the earlier-universe initial mass function was significantly weighted toward higher mass stars, or the higher-mass binaries are the ones that tend to coalesce, both of which could certainly be true. But you are saying the frequency is actually weighted toward 50 solar-mass black holes, but if the mirrors really are 4 km apart, that would seem to me to be too high frequency to be skewed that way.
Last edited by Ken G; 2017-Oct-09 at 09:27 PM.

23. Established Member
Join Date
Jun 2009
Posts
2,131
Originally Posted by Ken G
Ah, there's not just a wave amplitude issue, but also a frequency issue. So that's something I left out. But it seems strange the frequency sensitivity would work out as you say. The mirrors are 4 km apart, so that means they are most sensitive to detecting gravitational wave sources that are of similar scale, since we expect the merger to give maximum signal when moving close to the speed of light, and in any case cannot move faster than that. But the Schwarzschild radius is about 3M km, where M is in solar masses, so I would have thought 4 km would be well suited to finding neutron-star mergers, and quite a bit small for finding 50 solar mass black hole mergers. So that looks to make the situation even worse, in terms of why we are finding such huge and rare objects, unless the earlier-universe initial mass function was significantly weighted toward higher mass stars, or the higher-mass binaries are the ones that tend to coalesce, both of which could certainly be true. But you are saying the frequency is actually weighted toward 50 solar-mass black holes, but if the mirrors really are 4 km apart, that would seem to me to be too high frequency to be skewed that way.
The underlying issue is the sources of noise. While advanced LIGO does improve the bandwidth, LIGO is still most sensitive to frequencies around 200 Hz. The edge of detectability is one thing, but we shouldn't be surprised to find more events matching the instruments frequency profile.

see:
https://inspirehep.net/record/963331/plots

24. Order of Kilopi
Join Date
Oct 2005
Posts
25,837
OK, so it's a bit more complicated than a characteristic frequency of c/R, where R is the Schwarzschild radius, since then the frequencies would be much larger than 200 Hz even for very massive black holes.

25. Established Member
Join Date
Jun 2009
Posts
2,131
Originally Posted by 01101001
Are you sure that's true for Advanced LIGO? The informal chart at Caltech: Gravitational Wave Spectrum seems to indicate that neutron-star binaries were on the edge of detectability with original LIGO, but that they are now part of the expected events for Advanced LIGO. Now (or before too long), it is neutron-star asymmetries (not pairs) that are at the edge of detection.
I haven't done a complete analysis. However, we can see that the important factors are not independent.

1) The strength of the signal. The bigger the black hole, the bigger the signal.
2) The rarity of the signal. While a merger of black holes with a million suns worth of mass might be powerful, they are expected to be rare. Galaxies only collide so often.
3) The frequency of the signal. We, so far, have only detected signals lasting less than a second. We catch the final moment of the merger, and hence the frequency of that final 'blip' will be the related to the mass of the objects.

It's no surprise that we are detecting events with frequencies on the order of 100 Hz. These are strong signals with the right frequency. Sure, they seem to be more common than anyone expected, but then again, hot jupiters seemed more common than anyone expected when we first started looking for planets with a method particularly well suited to find hot jupiters.

This is an exciting time in astronomy, but I think it's important to just be patient and wait. We will find more events, and gleam more information. However, there is much work to be done. What is the correlation between GRBs and gravitational mergers? Are there galaxies that are very active with mergers, indicating a galactic merging event, which seem pristine optically?

Heck, I'm still waiting on the B-mode polarization of the CMB. I'm also waiting on absolute neutrino masses, and even whether they are fermionic or majorana.
Last edited by ShinAce; 2017-Oct-10 at 12:16 AM.

26. Established Member
Join Date
Jan 2010
Location
Wisconsin USA
Posts
2,533
Thanks all for your input. Pretty exciting times with gravitational waves. Would like to acknowledge UW-Milwaukee Physics department for their work in this area. I graduated from UW-Milwaukee with bachelors in nursing.

27. The August 17, 2017, neutron-star merger observed by LIGO and 70 observatories around the world, had the Fermi spacecraft seeing the gamma-ray burst within 2 seconds of the gravitational-wave climax of the merger.

That leaves very little wiggle room for grav waves traveling the speed of light.

No speed details, but background: LIGO: GW20170817 Press Release

28. Established Member
Join Date
Jan 2010
Location
Wisconsin USA
Posts
2,533
Originally Posted by 01101001
The August 17, 2017, neutron-star merger observed by LIGO and 70 observatories around the world, had the Fermi spacecraft seeing the gamma-ray burst within 2 seconds of the gravitational-wave climax of the merger.

That leaves very little wiggle room for grav waves traveling the speed of light.

No speed details, but background: LIGO: GW20170817 Press Release
Would the result that the gravitational waves arrived earlier than light waves indicate a granular spacetime?

29. Originally Posted by Copernicus
Would the result that the gravitational waves arrived earlier than light waves indicate a granular spacetime?
No, because no one assumes that the gravitational-wave maximum happened at the same instant the gamma rays were released. As I heard it explained at the presser, there are some (fairly swift) mechanical processes involved to get the gamma rays going.

As an extreme example, the radio-wave maximum is now, and expected to increase; they will watch that for months or years. That really takes time to happen.

30. Established Member
Join Date
Jan 2010
Location
Wisconsin USA
Posts
2,533
Originally Posted by 01101001
No, because no one assumes that the gravitational-wave maximum happened at the same instant the gamma rays were released. As I heard it explained at the presser, there are some (fairly swift) mechanical processes involved to get the gamma rays going.

As an extreme example, the radio-wave maximum is now, and expected to increase; they will watch that for months or years. That really takes time to happen.
Fair enough. Lets say next time this happens the distance is 10 times farther and the lag is 10 times more, would that indicate granular spacetime?

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•