Gravitational fields affecting each other?

[Buckethead]

TL;DR Summary

Does an external gravitational field affect the dynamics of a self gravitating system?

Regarding this paper:

https://iopscience.iop.org/article/10.3847/1538-4357/abbb96

In the opening sentence of the Abstract the following is stated:

"The strong equivalence principle (SEP) distinguishes general relativity (GR) from other viable theories of gravity. The SEP demands that the internal dynamics of a self-gravitating system under freefall in an external gravitational field should not depend on the external field strength. "

I"m confused by this as I would think we are just talking addition of vectors here where the flatness of spacetime around a star for example would be affected by an opposing gravitational field of another nearby star. For example, if you have a rock situated exactly between 2 stars of the same mass, wouldn't that rock remain stationarily situated? Doesn't this mean the gravitational fields are affecting each other?

I'm thinking this is not what the authors are talking about because of this statement:

"Tidal effects from neighboring galaxies in the Λ cold dark matter (CDM) context are not strong enough to explain these phenomena."

So I'm thinking that they are not talking about just the simple summing of vectors as I refer to above, but instead are referring to some deeper influence of a gravitational field from neighboring galaxies affecting the kinematics of the galaxy in question when it should not.

What is it about a distant gravitational field (other than tidal forces) that would affect the kinematics of a galaxy?

Thanks.

 

[kimbyd]

Internal dynamics is the critical phrase here. It's more or less a statement that if you're in a closed box in free fall (so you can't detect light, particles, etc. from outside), you can't do any experiment that will tell you about your motion relative to other objects in the universe.

This isn't completely true, because tidal forces will still be present for a finite-sized box. But it is true for an infinitesimal one. Which is why they mention tidal forces.

 

[PeterDonis]

I"m confused by this as I would think we are just talking addition of vectors here

No, for two reasons: first, the "gravitational field" in GR cannot be described by a vector, it requires tensors (the metric tensor, the Riemann tensor, etc.); and second, GR is nonlinear, so you can't just add the fields for two individual bodies to get a resultant field when both are present.

However, you are making a more fundamental error here as well: you are confusing the motion of an object in the global spacetime geometry, with the object's internal structure. The latter is what the SEP is talking about. For example, the Earth is in free fall motion about the Sun: that means the Sun's "gravitational field" (as opposed to tidal effects) does not affect the Earth's internal structure, which is determined by the Earth's gravity. But of course the Sun's "gravitational field" affects the Earth's motion about the Sun; the SEP certainly doesn't rule that out.

 

[Buckethead]

So in this paper, are they referring to the internal structure of the galaxy as a whole and saying the rotational curve of the galaxy due to its internal structure should not be affected by the external gravitational fields of neighboring galaxies but appears to be. Is this correct?

 

[Ibix]

Is this correct?

In GR, yes. In MOND, no. I believe the paper says that their results are consistent with MOND.

 

[PeterDonis]

I believe the paper says that their results are consistent with MOND.

Yes, the paper is basically arguing that standard GR cannot explain galaxy rotation curves because standard GR does not allow for any "external field effect", while MOND does.

 

[Buckethead]

OK, got it. I'm still trying to figure out what an external field my look like. Is there such a thing as a "strong" flat spacetime for example due to an even distribution of zillions of neighboring galaxies and a "weak" flat spacetime due to a sparse neighborhood? I'm saying flat spacetime because I'm assuming an external curved spacetime is the same as a tidal force which is not what this paper is referring to and would influence a galaxy.

If this is the case that there is a strong and weak flat spacetime, then in GR this makes no difference, but in MOND, this difference has an effect on the dynamics of the galaxy?

 

[PeterDonis]

I'm still trying to figure out what an external field my look like.

What the paper means by "external field" is something like the Newtonian concept of "gravitational potential"--roughly speaking, a lot of external masses means a large gravitational potential, but not many external masses means a small gravitational potential.

If this is the case that there is a strong and weak flat spacetime

It is possible in standard GR for there to be regions of flat spacetime with different gravitational potential (for example, the interior of a hollow sphere as compared with asymptotically flat spacetime at infinity), but the "external field effect" described in the paper, as far as I can tell, does not require the region of spacetime in which it occurs to be flat; the effect is simply separate from the usual tidal effects of spacetime curvature (and of course there is no such "external field effect" in standard GR). And of course any such difference in gravitational potential requires there to be spacetime curvature somewhere in the spacetime.

 

[Ibix]

roughly speaking, a lot of external masses means a large gravitational potential, but not many external masses means a small gravitational potential.

...the absolute value of which is indetectable in Newtonian gravity (only differences in potential can have physical effects), but is detectable in MOND, as I understand.

 

[kimbyd]

So in this paper, are they referring to the internal structure of the galaxy as a whole and saying the rotational curve of the galaxy due to its internal structure should not be affected by the external gravitational fields of neighboring galaxies but appears to be. Is this correct?

That's what they're trying to say, yes.

I may be biased here, but to my mind this isn't really telling us much. MOND simply doesn't fit other observations to a pretty spectacular degree. And observations like this particular one are sensitive to system dynamics that are incredibly hard to model accurately.

The crux of the matter here is that the smaller the system is, the more the behavior of normal matter plays a role. And normal matter is incredibly hard to model accurately. The entire dynamical behavior of individual galaxies is impacted by things like supernova explosions and the behavior of their active galactic nuclei, for example. And the impact of these factors just can't be accurately simulated with current technology, period.

But MOND only really seems to work at these smaller cosmic scales. Such as this study which is looking at dynamics internal to various galaxies. The moment you start looking at systems larger than galaxies, MOND starts to break down in pretty dramatic ways. To my knowledge, MOND has always struggled to explain the behavior of galaxy clusters. And now that we have a number of examples of collisions of galaxy clusters which separate normal matter from dark matter to a dramatic degree (such as the Bullet Cluster), MOND has an even harder time explaining things. I know that some years ago, the only modified gravity model to fit the Bullet Cluster could only do so if they added heavy neutrinos to the mix, so the model still had dark matter, just less of it.

And MOND completely fails to explain the CMB data.

So when I see MOND proponents doing yet more studies of internal galactic dynamics, it's just incredibly disappointing. I honestly feel it's scientifically irresponsible to focus only on the areas where their model does well while ignoring the glaring discrepancies between their model and observations.

If you look at the other side of the coin, for instance, physicists who are working on dark matter models tend to pretty actively engage in the areas where the model fits things poorly. Such as galactic centers.

 

[PeterDonis]

...the absolute value of which is indetectable in Newtonian gravity (only differences in potential can have physical effects), but is detectable in MOND

In at least some versions of MOND, yes. I don't think the "external field effect" referred to in the paper is present in all versions of MOND.

 

[Vanadium 50]

But MOND only really seems to work at these smaller cosmic scales

I would say instead that it works on galactic scales and nowhere else. As such, it might not be telling us anything abut gravity, but might be telling us something about galaxies and their formation. And certainly MOND does better with the baryonic Tully-Fisher relationship than ΛCDM.

And MOND completely fails to explain the CMB data.

I think that's a bit unfair. MOND predicted the first to second peak ratio before ΛCDM did. It's true that it doesn't get the third peak. Also, the CMB makes a BBN prediction for ⁷Li production that misses the observation by a factor of ~4. The reaction of the community seems to be "pay no attention to the lithium beyond the curtain". Maybe there's some astrophysics that makes the lithium go away. Maybe there isn't.

 

[Buckethead]

I'm not a fan of either MOND or of CDM. My interest in this paper is the relationship between the rotational curve of galaxies and the density of their neighborhoods which I find fascinating. It's another interesting twist to the DM saga.

 

[kimbyd]

I think that's a bit unfair. MOND predicted the first to second peak ratio before ΛCDM did. It's true that it doesn't get the third peak. Also, the CMB makes a BBN prediction for ⁷Li production that misses the observation by a factor of ~4. The reaction of the community seems to be "pay no attention to the lithium beyond the curtain". Maybe there's some astrophysics that makes the lithium go away. Maybe there isn't.

That's a far, far greater problem than you are implying here. That third peak is the entire point.

With MOND, you have a sharp drop-off between the first and second peaks because the image of the CMB is blurry: the plasma in the early universe doesn't all condense into a gas at the same time, so the surface we see is cloudy. This effect is amplified in MOND because it's a purely normal matter effect.

With ΛCDM, there isn't as much normal matter (proportionately), which narrows the surface of last scattering, making it less blurry. The second peak (in fact, every even-numbered peak) is instead suppressed because dark matter doesn't experience pressure like normal matter does: it doesn't bounce after falling into a gravitational well.

These are two wildly different pictures of what is going on with the CMB data, and now that we've measured much more than just the first two peaks, it is abundantly clear which model fits the data better. Who came first on those first two peaks is simply irrelevant.

 

[Buckethead]

To my knowledge, MOND has always struggled to explain the behavior of galaxy clusters. And now that we have a number of examples of collisions of galaxy clusters which separate normal matter from dark matter to a dramatic degree (such as the Bullet Cluster)

The Bullet Cluster became an instant celebrity because of it's support for DM and the difficulty it caused MOND but then along comes Abell 520 (Train Wreck Cluster) a few years later which shows DM following the gas in the center and not the galaxies that have passed through each other and left everyone scratching their heads. The thing is no-one talks about Abell 520 and it's an embarrassment to the CDM hypothesis. Yes there have been refutes to the result, but it is clearly an opposing discovery to the Bullet Cluster.

The fact is, both the CDM theory and MOND have serious problems. There are some alternate hypotheses, such as Erik P. Verlinde's idea:

https://arxiv.org/abs/1611.02269

which I do not have an opinion about, but the thing is, there are different ideas that need shaking down. I think focusing too much on CDM and MOND at this point is counter productive. New ideas need to be generated.

I would think it would be benificial to focus on the odd evidence that has been showing up to try and come up with an alternative hypothesis. For example:

1. The influence of neighboring galaxies which is the subject of this post
2. Eliptical galaxies show less DM than Spirals:
https://astrobites.org/2013/04/04/do-elliptical-galaxies-have-dark-matter-halos/
3. Milky Way galaxy shows little evidence for DM:
https://arxiv.org/abs/1411.2625

For some odd reason, it seems to be the consensus that the only alternative to CDM is MOND. That its got to be one or the other. Why the limited thinking? I can think of one idea off hand that predicts #1, 2, and 3 above although I can't say what it is as its not in a published paper. All I'm saying is there are alternatives.

 

[kimbyd]

You're making the false assumption that an alternative hypothesis needs to be made in the first place.

As far as Abell 520, stuff is complicated sometimes. That one is at the crossing of three filaments, and is therefore not a simple two-cluster collision like the bullet cluster. Here we again have an important point: when observations are at their simplest, dark matter is the clear winner.

 

[PeterDonis]

New ideas need to be generated.

They are. You yourself noted Verlinde's paper. He's not the only one generating alternative hypotheses.

The issue is not so much generating new ideas as testing them. New observations are always coming in, but they aren't being made in order to try to differentiate between any particular hypotheses; they're just being made by default, because we have telescopes and things that are always looking at new parts of the sky. We don't really have the capability to do much more, which means that it will take a long time to make progress on differentiating between hypotheses.

I can think of one idea off hand that predicts #1, 2, and 3 above although I can't say what it is as its not in a published paper.

Then get it published. Either you will, in which case it will get considered along with all the other ideas that have been published, or you won't. But until it is published, it's off limits for discussion here, as you know quite well, and that includes offhand remarks like this one.

 

[Buckethead]

You're making the false assumption that an alternative hypothesis needs to be made in the first place.

Respectfully I believe alternative hypothesis always need to be considered in absolutely every aspect of science, otherwise you end up with dogma and I think the idea of CDM is getting dangerously close to dogma. Consider this: The idea of DM is the very simplest and easiest hypothesis that could possibly have been created to explain rotational curves of spiral galaxies and galactic clusters. A six year old could have thought of it. But of course that is the very reason it needed to be thoroughly investigated and I fully support all the money and research that has gone into it. But now it's time to reflect. Not stop, just reflect. DM has reached a point where we are running out of detection options and not for lack of trying. Maybe DM simply does not exist. MOND also has issues, as you are aware. So yes, alternative hypothesis do need to continue to be created up until the point where DM is beyond a shadow of a doubt detected.

They are. You yourself noted Verlinde's paper. He's not the only one generating alternative hypotheses.

The issue is not so much generating new ideas as testing them. New observations are always coming in, but they aren't being made in order to try to differentiate between any particular hypotheses; they're just being made by default, because we have telescopes and things that are always looking at new parts of the sky. We don't really have the capability to do much more, which means that it will take a long time to make progress on differentiating between hypotheses.

Agreed, but I think my point was more taking the awesome amount of data supporting new anomolies that has been coming in and trying to tie them together. What seems to be happening is new observations are getting pushed aside if they are difficult to explain using DM, and supporting observations are getting exhaulted. The Bullet Cluster/Train Wreck is an example of this. How much of the Bullet Cluster have we seen even in popular books on DM and You Tube videos on DM and on and on and how we have seen virtually nothing on the Train Wreck. There seems to be this underlying current to first try and explain something within the DM paradigm, and if it can't, just keep trying! It's the inertia behind the DM hypothesis that scares me.

Then get it published. Either you will, in which case it will get considered along with all the other ideas that have been published, or you won't. But until it is published, it's off limits for discussion here, as you know quite well, and that includes offhand remarks like this one.

I apologize for this. I did feel a bit of self indulgent sarcasm when I posted it.

 

[kimbyd]

One thing I'd like to add is that there is certainly a lot of work to be done to understand dark matter. From my current understanding, I generally feel that:
1) Dark matter's existence is close to certain. Dark matter's relative strength in predicting the simplest systems in the universe is a testament to this.
2) The precise nature of dark matter is very much unknown.
3) The current differences between observations and various dark matter models are likely accounted for by some combination of complexities we haven't yet been able to simulate and the precise behavior of dark matter (e.g., its annihilation rate and temperature).

That said, I thought I'd go on a little rant about why I still think the CMB is by far the strongest evidence in favor of dark matter. I won't go much into the details of how dark matter physics works. I mentioned it a bit above, but I thought I'd highlight some other facts.

1) The gravitational behavior that is relevant to CMB observations is exceedingly simple. Specifically, it can be accurately modeled using linearized gravity. The linear approximation works because of two features that are relevant: the the differences in density by the time the CMB was emitted were small, and the relevant distances being examined are large. The fact that linear equations are accurate in this situation makes the calculations extremely easy. The predicted statistics of the hot and cold spots on the CMB can be computed in a fraction of a second on a modern computer, for instance.
2) The CMB data gives an incredibly accurate measure of the overall amount of normal matter in the universe. This is pretty easy to understand: because all of the normal matter in the early universe was illuminated (because it was a plasma), we actually see it. This has direct and clearly-visible effects on the CMB data. As of the latest Planck data, the amount of normal matter in the observable universe is known to roughly 0.7% accuracy.
3) CMB data gives an extremely accurate estimate of the ratio of normal matter to dark matter. I described a bit why before, but the crux of the matter is just that dark matter and normal matter behave differently, and these differences can be calculated very precisely because of the linear behavior of the system. Because of this, the latest Planck release, by measuring this ratio, measures the amount of dark matter in our observable universe to within about 1%.

It's hard to translate these numbers into how confident we can be that dark matter exists at all. But it gives a hint of how confident we can be. The fundamental issue is if you want to try to replicate this effect, you have to recreate a signal that has a number of interesting features very precisely using something completely different. And that's just really really hard to do.

By way of comparison, we can look at the error bars typically quoted for measures of galaxies. Consider, for example, Andromeda. This is the closest galaxy to us outside of our own, and you would therefore think that its mass would be the easiest for us to measure. As of this paper, we currently know the mass of the Andromeda galaxy to only about 30% accuracy (see section 6.4, page 16).

This isn't the whole picture, not by a long shot. But it is reflective of how much more precisely we can measure some aspects of our observable universe using the CMB than using galaxies. And the existence and amount of dark matter is one of those aspects. Other things can be best-measured by looking at the nearby universe, of course (the CMB alone isn't very sensitive to dark energy, the Hubble expansion rate, or spatial curvature, for instance).

 

[PeterDonis]

alternative hypothesis always need to be considered in absolutely every aspect of science

Not every aspect, since in some areas of science we already have confirmation of our best current theories to extremely high accuracy in particular domains. The reason it is reasonable to consider alternative hypotheses in the area of science under discussion in this thread is that it is not within one of those particular domains.

 

[Buckethead]

That said, I thought I'd go on a little rant about why I still think the CMB is by far the strongest evidence in favor of dark matter. I won't go much into the details of how dark matter physics works. I mentioned it a bit above, but I thought I'd highlight some other facts.

Thank you for the detailed response. I must admit I am quite un-versed as to the importance of the CMB in the DM paradigm. I have focused exclusively on rotation curves of spirals, elipticals, and clusters and of course the all important lensing aspect. With the limited research I've done with regard to CMB, I still can't grasp how the 3 peaks in the density graphs translate into dark matter/normal matter/dark energy ratios. But apparently from what I've read, according to some, that pesky third peak puts a nail in the coffin. I won't agree till I learn more, but I didn't realize the degree of importance this held.

That said. Here is my list of why I think CDM is in trouble:
1. What started out as a occum's razor explanation for rotational curves has grown into a hypothesis of complexity rivaling alternative theories of GR if for no other reason than the CDM particle is not in the standard model. That is HUGE! We are no longer talking about a dust particle here, we are talking about a particle that challenges everything we know about particle physics.
2. CDM has had much more time as a theory to evolve and therefore has many more proponents. For every scientist that opposes CDM as a hypothesis there are perhaps 100 that support it and that inertia makes the playing field lopsided and difficult for dissenters. I expect this will shake out eventually, but for now, it's lopsided.
3. More to the point, there have been a great number of experiments done to detect DM particles within very narrow energy levels and with high levels of confidence that have all failed. Very expensive, very precise and very good experiments.
4. The search for DM and for CDM in particular has gone on for quite a long time now and still no cigar.
5. More importantly, as more interesting data comes in, we are discovering that there is a wide fluctuation in just how much DM surrounds various galaxies. For example, Eliptical galaxies show less DM than spirals (although there are exceptions) and some dwarf galaxies have large amounts of DM and others have none. Then there is the anomaly pointed out in my original post relating neighborhood densities with rotational curves. This can't be addressed in a CDM paradigm. I'm sure there are other interesting phenomenon that I am unaware of that will also add to the confusion.
6. With regard to the CMB I might say perhaps there is an assumption there that is throwing things off. For example the rate of inflation or how long it lasted? Or, if before 380K years after the big bang, if there was 5X as much CDM in the soup and it doesn't interact with light, why did light have such a hard time getting out before then? I'm drawing at straws with my examples, but my point is, the dynamics of the big bang surely must be extremely mysterious in those early years.

That's all I can think of right now but I think that's a pretty good list.

 

[PeterDonis]

I have focused exclusively on rotation curves of spirals, elipticals, and clusters and of course the all important lensing aspect.

Yes, which means all of your judgments about various theoretical hypotheses are skewed, because you are only familiar with the evidence in one particular regime. But any valid scientific model of our universe has to account for all of the evidence, in all regimes. So you can't properly evaluate any models or any hypotheses without considering the evidence in all regimes.

I'm drawing at straws with my examples

Yes, you certainly are. Please bear in mind the PF rules on personal speculation.

my point is, the dynamics of the big bang surely must be extremely mysterious in those early years

I think you need to spend a lot more time learning about how much we actually know about the dynamics of the Big Bang in those early years.