Merging Neutron Stars: What We Know So Far

[mfb]

It seems strange that this observation that light and gravity are behaving so similarly (forming waves that propagate at the same speed) can't somehow be used to show that gravitons belong in the standard model like photons. There has got to be a way.

The SM, by definition, doesn't include gravity. And just because both photons and hypothetical gravitons are massless doesn't mean the phenomena would be similar in any way. We know they are not, otherwise we would have a consistent theory of quantum gravity by now.

@ohwilleke: You underestimate the number of events we will get. Let's say conservatively 1 NS event in 1 year of running, ignoring that the sensitivity improved over time. In a year, with twice the sensitivity, we would expect 8 events per year. In ~3 years, with the full sensitivity, we might get something like 20-30 per year. Add KAGRA and the number will get even larger. Add INDIGO and we might get more than 100 events per year. And then we can build the Einstein telescope which will see these events routinely.
Of course we might have been extremely lucky with this NS event, but that is unlikely. For binary black hole mergers the situation is even better, with 4 events observed already we can be quite sure they are not extremely uncommon.

The matter of the observed collision was noted, followed by the comments (unsourced) that this resulted in the production of a large quantity of gold and platinum, and that consequently we now have a clear idea where heavy metals come from.

That was said by the scientists during the press conference.

 

[phyzguy]

So apparently the GW signal was too weak to get any information on the ring-down phase. At least I haven't seen it mentioned anywhere. Did I just miss it? Has anyone seen any information on a ring-down phase?

 

[Sophrosyne]

Gravitational waves in GR propagate at the speed of light and so do EM waves in Maxwell's classical EM equations. The speed of propagation is something that the classical theories and quantum ones have in common, not a distinction.

Sure. But currently we have a quantum theory of light, but not gravitation. Isn't there anything in these findings of observed gravitational waves that we can use to try to confirm or rule out some of the current proposals for a quantum theory of gravity, whether strings or quantum loop or whatever?

 

[Vanadium 50]

Isn't there anything in these findings of observed gravitational waves that we can use to try to confirm or rule out some of the current proposals for a quantum theory of gravity, whether strings or quantum loop or whatever?

No.

 

[websterling]

So apparently the GW signal was too weak to get any information on the ring-down phase. At least I haven't seen it mentioned anywhere. Did I just miss it? Has anyone seen any information on a ring-down phase?

As the inspiral progressed, the frequency of the gravitational waves increases to higher than the frequency range the detectors are sensitive to. I think the merger and ringdown were missed because of this.

 

[ohwilleke]

You underestimate the number of events we will get. Let's say conservatively 1 NS event in 1 year of running, ignoring that the sensitivity improved over time. In a year, with twice the sensitivity, we would expect 8 events per year. In ~3 years, with the full sensitivity, we might get something like 20-30 per year. Add KAGRA and the number will get even larger. Add INDIGO and we might get more than 100 events per year. And then we can build the Einstein telescope which will see these events routinely.
Of course we might have been extremely lucky with this NS event, but that is unlikely. For binary black hole mergers the situation is even better, with 4 events observed already we can be quite sure they are not extremely uncommon.

This is very encouraging. I knew there would be some improvement in sensitivity and the number of detectors over time, but I had no idea that it would be so dramatic.

 

[snorkack]

How long after merger did visual spectrum start to show lines (whether emission or absorption)?

 

[OmCheeto]

How long after merger did visual spectrum start to show lines (whether emission or absorption)?

I believe physzguy kind of answered that in Borek's "splinter" thread:

It took about 11 hours for the optical telescopes to start looking, mostly because they were waiting for it to get dark. As soon as they started looking, the optical radiation was there, but it shifted to longer wavelengths as the ejecta cooled off over time.

ps. I've been thinking of starting a "splinter" thread myself, after a bunch of people on the internet started whining that; "Those aren't real pictures!" to NASA, et. al.
My very bad maths told me yesterday that it would take a telescope that is 3 light years in diameter to get a clear visual image of the event.
I googled feverishly to get confirmation that my maths was wrong, and found a web site* that said an optical telescope of that size would be so massive, it would collapse into a black hole.

[edit] *found it in my browsing history: http://quarksandcoffee.com/index.php/2015/07/08/aliens-and-dinosaurs/

 

[phyzguy]

The optical pictures you see are not resolved images of the merger. They just capture the total light emitted. Optical photos of distant stars are never resolved. The stars are just points of light.

 

[OmCheeto]

The optical pictures you see are not resolved images of the merger. They just capture the total light emitted. Optical photos of distant stars are never resolved. The stars are just points of light.

I know that. But how do you explain to someone who might find out that all of the data comes from a pin point, how the [FAKE!] pictures are created?

ps. I really wish that Goddard had put in time stamps for their little movie.

​

I like watching it at 1/4 speed, at max resolution.

 

[phyzguy]

I know that. But how do you explain to someone who might find out that all of the data comes from a pin point, how the [FAKE!] pictures are created?

I see your problem. Some of the movies are artists' conceptions and some are computer simulations. It's hard to know which are which without digging into it.

 

[phyzguy]

As the inspiral progressed, the frequency of the gravitational waves increases to higher than the frequency range the detectors are sensitive to. I think the merger and ringdown were missed because of this.

Thanks, that makes sense. That explains why the frequency/time graph disappears off the top. I wonder if the upgrades will help with this.

 

[stefan r]

I believe physzguy kind of answered that in Borek's "splinter" thread:
ps. I've been thinking of starting a "splinter" thread myself, after a bunch of people on the internet started whining that; "Those aren't real pictures!" to NASA, et. al.
My very bad maths told me yesterday that it would take a telescope that is 3 light years in diameter to get a clear visual image of the event.
I googled feverishly to get confirmation that my maths was wrong, and found a web site* that said an optical telescope of that size would be so massive, it would collapse into a black hole.

[edit] *found it in my browsing history: http://quarksandcoffee.com/index.php/2015/07/08/aliens-and-dinosaurs/

That is only if you wanted to look at a T-rex. A Texas sized pixel would make a very nice picture/video. A 10au diameter solid glass lens could collapse into a black hole. A mirror only needs to be a few atoms thick. There are a lot of photons coming out of 10⁴⁴J events. An array orbiting the sun could get the resolution.

 

[Urs Schreiber]

Isn't there anything in these findings of observed gravitational waves that we can use to try to confirm or rule out some of the current proposals for a quantum theory of gravity

At least all theories involving violation of Lorentz invariance (e.g. doubly-special-relativity, Horava-gravity, many non-commutative-spacetime models) now face stronger constraints by up to ten orders of magnitude.

 

[mfb]

The 11 hours are the time until an optical telescope found the source. That doesn't mean there was no optical light emission before, we just don't have data about this time. Telescopes that can find both dim sources and cover a large area in the sky are rare, and the localization based on the gamma ray burst and the gravitational wave still left a large area to search in the sky (relative to the field of view of telescopes).

Optical photos of distant stars are never resolved.

Be careful with overly general statements, they might be wrong. Okay, you could argue Antares is not that distant...

In addition, the event produced jets much larger than stellar objects, it might be possible to see them in the future. ELT with its 5 mas resolution could see structures as small as 3 light years across at this distance. I don't know if the jets are bright enough for that, however.

 

[ohwilleke]

Matt Strassler has another more lengthy and exceptionally accessible explanation for those who are less familiar with the relevant topics: https://profmattstrassler.com/2017/...rdays-gravitational-wave-announcement-an-faq/

 

[Urs Schreiber]

At least all theories involving violation of Lorentz invariance (e.g. doubly-special-relativity, Horava-gravity, many non-commutative-spacetime models) now face stronger constraints by up to ten orders of magnitude.

Similarly, MOND-like modifications of the laws of gravity are further constrained by the new data, see here:

• Jose María Ezquiaga, Miguel Zumalacárregui, "Dark Energy after GW170817" (arXiv:1710.05901)
• Jeremy Sakstein, Bhuvnesh Jain, "Implications of the Neutron Star Merger GW170817 for Cosmological Scalar-Tensor Theories" (arXiv:1710.05893)
• Sibel Boran, Shantanu Desai, Emre Kahya, Richard Woodard, "GW170817 Falsifies Dark Matter Emulators" (arXiv:1710.06168)

Of course MOND faces bigger problems already,

• Scott Dodelson, "The Real Problem with MOND", Int. J. Mod. Phys. D, 20, 2749 (2011). (arXiv:1112.1320)

but still.

 

[ohwilleke]

Similarly, MOND-like modifications of the laws of gravity are further constrained by the new data, see here:

• Jose María Ezquiaga, Miguel Zumalacárregui, "Dark Energy after GW170817" (arXiv:1710.05901)
• Jeremy Sakstein, Bhuvnesh Jain, "Implications of the Neutron Star Merger GW170817 for Cosmological Scalar-Tensor Theories" (arXiv:1710.05893)
• Sibel Boran, Shantanu Desai, Emre Kahya, Richard Woodard, "GW170817 Falsifies Dark Matter Emulators" (arXiv:1710.06168)

Of course MOND faces bigger problems already,

• Scott Dodelson, "The Real Problem with MOND", Int. J. Mod. Phys. D, 20, 2749 (2011). (arXiv:1112.1320)

but still.

Dodelson (2011) is really just attacking a straw man. Everyone has known since the beginning that MOND-proper is a toy model and needs to be generalized to be relativistic and doesn't capture cluster phenomena. And, it was generalized with TeVeS and a similar approach was made with Moffat's MOG theory (that work for clusters and cosmology). Apparently Boran (2017) is a blow to those relativistic approaches that are more general and not pure toy models.

Ezquiaga (2017) and Sakstein (2017) are not primarily going after MOND-like modifications. They are instead addressing a different group of gravity modifications usually pushed by GR theorists (e.g. some f(R) theories of gravity) designed only to deal with dark energy and not with dark matter - almost the opposite of what MOND-like gravity modification theories do, MOND-like gravity modification theories often don't address dark energy phenomena at all. Ezquiaga argues that gravity doesn't propagate at the speed of light in TeVeS, but I'm skeptical of that claim (he relies on another paper for this throw away statement in his conclusion) and it is certainly a theory specific argument and not a generalized modified gravity argument. Ezquiaga (2017) also makes clear that some modified gravity theories do make the cut:

Motivated by these results, we identify the theories that avoid this constraint and thus can still be used to explain DE (see a summary in Fig. 3). Within Horndeski’s theory, the simplest models such as quintessence/kessence, Kinetic Gravity Braiding or Brans-Dicke/f(R) are the ones surviving. Beyond Horndeski theory, viable gravities can be obtained in two ways. One can apply a derivative-dependent conformal transformation to those Horndeski models with cg = 1, since it does not affect their causal structure. An example of this is the derivative conformal transformation of GR. Alternatively, one can implement a disformal transformation, which does alter the GWs light-cone, designed to precisely compensate the original anomalous speed of the theory. Specific combinations of Horndeski and GLPV Lagrangians are representatives of this class.

In more general grounds, the bounds on cg severely restrict the kinetic term of gravity to be canonical (of the Einstein-Hilbert form), up to field redefinitions that preserve the causal structure. This requirement provides a strong selection criteria for viable modified theories of gravity, applicable also to theories other than scalartensor gravity. Massive gravity [21], bigravity [55] and multi-gravity [56] all fall in the safe category as long as matter couples minimally to one of the metrics. Note that in the cases with more than one dynamical metric(s), tensor perturbations of the auxiliary, uncoupled metrics will in general travel at a different speed. In minimally coupled scenarios this effect is only detectable by graviton oscillations with the physical metric [57].

Boran (2017) does place significant limits on the parameter space of MOND-like theories that use gravity modification to explain phenomena attributed to dark matter. But I'll defer further commenting on that paper as I haven't had a chance to really dig into it yet. FWIW, at first glance it looks to me like the case of Boran (2017) is probably overstated, but I'm willing to keep an open mind for now.

Certainly, nothing in Boran (2017) in any way impairs the approach taken in the following series of papers that involve a massless boson as a force carrier:

* A. Deur, "A possible explanation for dark matter and dark energy consistent with the Standard Model of particle physics and General Relativity" (2017).
* A. Deur, "Self-interacting scalar fields in their strong regime" (November 17, 2016).
* Alexandre Deur, "A correlation between the amount of dark matter in elliptical galaxies and their shape" (July 28, 2014).

Incidentally, I don't agree that Deur's approach is actually consistent with classical GR as currently formulated, although the tweak that he makes in coming up with his own regime that handles rotation curves, cluster data, elliptical galaxies and cosmology tests, at least at a back of napkin level of precision, are very subtle and very principled. In both results and theoretical motivation it is probably the best of the current gravitational explanations of dark matter phenomena, although it has been ill developed as the author has had to devote most of his work to his day job in QCD and doesn't have the funding, support or following necessary to really kick the tires of this approach.

The other point to recognize is that dark matter particle theories are in very deep trouble in ways which this data point doesn't address. Truly collisionless dark matter is all but ruled out, and the parameter space of self-interacting dark matter theories is also highly constrained. See, e.g., Lin Wang, Da-Ming Chen, Ran Li "The total density profile of DM halos fitted from strong lensing" (July 31, 2017); Paolo Salucci and Nicola Turini, "Evidences for Collisional Dark Matter In Galaxies?" (July 4, 2017).

 

[jerromyjon]

Sounds to me like we need another "Einstein" to figure out the cosmological situation. That was his biggest "blunder", as he put it.

 

[Urs Schreiber]

Boran (2017) does place significant limits on the parameter space of MOND-like theories that use gravity modification to explain phenomena attributed to dark matter. But I'll defer further commenting on that paper as I haven't had a chance to really dig into it yet.

Thanks for the detailed comments. I'll be interested in your take on Boran17.

The other point to recognize is that dark matter particle theories are in very deep trouble in ways which this data point doesn't address. Truly collisionless dark matter is all but ruled out, and the parameter space of self-interacting dark matter theories is also highly constrained. See, e.g., Lin Wang, Da-Ming Chen, Ran Li "The total density profile of DM halos fitted from strong lensing" (July 31, 2017); Paolo Salucci and Nicola Turini, "Evidences for Collisional Dark Matter In Galaxies?" (July 4, 2017).

Sure. It seems you are now passing from the question "How does the coincident GW+EM radiation from GW170817 constrain modifications of basic physical laws?" to a general discussion of the problem of DM+DE. How about ultralight axion models? They seem to be in decent shape.

 

[Buzz Bloom]

Paolo Salucci and Nicola Turini, "Evidences for Collisional Dark Matter In Galaxies?" (July 4, 2017).

I have been unsuccessfully trying to find a definition of the term "collosional dark matter" or the term "collosionless dark matter". I assume these terms are opposite in meaning so that that finding one definition would therefore be sufficient. I searched the Internet for these terms and found quite a few articles, but in none of the abstracts was a definition given.

Merriam Webster gives the definition of "collisionless" as

of, relating to, or being a plasma in which particles interact through charge rather than collision
https://www.merriam-webster.com/dictionary/collisionless​

The Free Dictionary gives the definition of "collisional" as

A brief dynamic event consisting of the close approach of two or more particles, such as atoms, resulting in an abrupt change of momentum or exchange of energy.
https://www.thefreedictionary.com/collisional​

If these are correct definitions in the context of "dark matter", then presumably "collosionless dark matter" means dark matter that interacts via EM, but everything I have read so far about dark matter says this is inconsistent with the lack of any observational evidence for such interaction.

Here is a quote from the P Salucci and N. Turini paper:

Moreover, the analysis of the CMB fluctuations spectrum
and a number of cosmological measurements unavoidably point to a scenario in which a
Dark Massive Particle is the responsible for the mass discrepancy phenomenon in Galaxies
and Clusters of Galaxies ( Planck Collaboration (2016)).​

 

[SciencewithDrJ]

What's next in GW research, I wonder.

 

[phyzguy]

I have been unsuccessfully trying to find a definition of the term "collosional dark matter" or the term "collosionless dark matter". I assume these terms are opposite in meaning so that that finding one definition would therefore be sufficient. I Internet searched the Internet for these terms and found quite a few articles, but in none of the abstracts was a definition given.

I think the term "collisionless dark matter" is a shorthand for "dark matter that only interacts with ordinary matter through gravity, and has no other interactions", while collisional dark matter is dark matter that has some other interaction, not necessarily electromagnetic.

 

[phyzguy]

What's next in GW research, I wonder.

My understanding is that LIGO has gone down for about one year to increase the sensitivity. This will allow it to look deeper into space and see events at a higher rate. KAGRA is planned to come on line in Japan next year, and Indigo in India after that. This Wikipedia entry shows some of the plans.

 

[SciencewithDrJ]

How do GW detection findings help further refine the calculation of the age of the Universe?

 

[phyzguy]

How do GW detection findings help further refine the calculation of the age of the Universe?

We already know this value to about 1% or better. How accurate do you need it to be? Did you notice a change when Planck refined this value from 13.7 billion years to 13.82 billion years?

 

[SciencewithDrJ]

We already know this value to about 1% or better. How accurate do you need it to be? Did you notice a change when Planck refined this value from 13.7 billion years to 13.82 billion years?

Thanks for the speedy response. What I meant is (not being a physicist) what is it in the findings from GW detection that enables us to get a more accurate calculation. What's the connection?

 

[Jonathan Scott]

What's next in GW research, I wonder.

I'm hoping for mergers with masses heavier than the recent neutron stars to help establish the physics for the threshold for collapse into a black hole. Or, even more exciting, more mergers of objects which ought to be black holes but which are accompanied by electromagnetic radiation (as initially appeared to be the case for the first detection), suggesting the need for new theory!

 

[PAllen]

I recently attended a talk by Edo Berger on this, and, in person, he expressed strong confidence that the remnant became a BH very quickly. The data and argument are based on the following paper. However the paper’s conclusion on the final state is stated more weakly than the in person presentation.

https://arxiv.org/abs/1710.11576

The discussion of the nature of the remnant begins at the end of p.9.

A comment in the talk was that the remnant would either be the heaviest known neutron star or the lightest known BH, so a first either way. The suggestion is that a 2.7 solar mass BH is now known.

 

[phyzguy]

I recently attended a talk by Edo Berger on this, and, in person, he expressed strong confidence that the remnant became a BH very quickly. The data and argument are based on the following paper. However the paper’s conclusion on the final state is stated more weakly than the in person presentation.

From my reading and discussions, it seems that the preferred model is that the SGRB, which occurred approximately 1.7 seconds after the time of merger inferred from the GW signal, is when the BH formed. Was this discussed in the verbal presentation? When you say "very quickly", do you mean on the time scale of seconds?