PDA

View Full Version : MOND observed between binary stars



borman
2011-Jul-08, 10:50 PM
MOND observed between binary stars

The new update provides an additional study with better signal to noise ratio that confirms the observations of the first version of the paper.

In short, when the distance between two stars that are gravitationally bound grows to where the acceleration threshold between them falls below a critical value, a_0, the stronger Dark Matter Effect, or perhaps modified gravity, or something else we donít yet understand, takes ascendance over the typical inverse squared law of Newton and their dynamics are governed by the additional force.

Wide binaries as a critical test of Classical Gravity

http://arxiv.org/abs/1105.1873

Tensor
2011-Jul-09, 04:11 AM
MOND observed between binary stars

The new update provides an additional study with better signal to noise ratio that confirms the observations of the first version of the paper.

In short, when the distance between two stars that are gravitationally bound grows to where the acceleration threshold between them falls below a critical value, a_0, the stronger Dark Matter Effect, or perhaps modified gravity, or something else we donít yet understand, takes ascendance over the typical inverse squared law of Newton and their dynamics are governed by the additional force.

Wide binaries as a critical test of Classical Gravity

http://arxiv.org/abs/1105.1873

It's nice, but there are two problems (well there are more, but I'll stick with two for now). One, as several commenters at Cosmic Variance point out, is that both studies are based on the proper motions of the binaries. Usual errors of this type can be 10 to 20 times the value of the measurement.

The other problem is that they claim there is no experimental evidence for dark matter. However, this paper (http://arxiv.org/pdf/1106.0650v2) announces a annual signal consistent with WIMPs. However, the signal is right at threshold. That's not encouraging. I'm not blaming Hernandez et al, as this paper was released after their paper.

Now, I will point out that neither of the papers have yet been peer reviewed, as of yet. There are good probabilities that both papers may end up going away after peer review and after further studies.

There is good info at Cosmic Variance (http://blogs.discovermagazine.com/cosmicvariance/2011/05/22/arxiv-find-breakdown-of-classical-gravity/) and The e-Astronomer (http://andyxl.wordpress.com/2011/06/13/dark-matter-heats-up/). In both the blogs and the comment sections.

borman
2011-Jul-09, 08:16 PM
Tensor,
Thank you for your comments, but you might want to look at the 2nd version of the paper. It has a new title, is nearly twice as long, and has 10 charts as opposed to the original 4. There is also the inclusion of a study on the Hipparcos data, using only 280 samples as opposed to the over 1200 from the SDSS data, but with improved S/N ratio. Many of the questions regarding your first point are addressed in the extended paper.

The tenor of their statement that no physical evidence of Dark Matter Particles still holds. We don't have a "beaker" full of DM particles yet. The recent CoGeNT, CDMS II, DAMA/LIBRA, and EDELWEISS results seem to indicate an annual signal of some sort. With DAMA/LIBRA the signal has high sigma, over 8. However, DM is not proven; rather it is being suggested as a possible explanation. There have been many recent papers dealing with compatibility between the various signals and whether they can be reconciled with each other considering whether the interaction is elastic or not and the possible mass range of the putative WIMP. There are still the negative results to be explained as well fom other detectors. A possibility has been a lighter WIMP in the upper 20's of GeV. However, this is clearly in the reach of the LHC and there have been no reports of any created masses in this range whether mostly hidden (WIMPs) or completely hidden (NIMPs) that are needed to conserve momentum.

Considering that the phnomenon that requires either some modified gravity theory or Dark Matter Particles to preserve Newton-Kepler-Einstein is apparent at extra-galactic, galactic especially including dwarf and LSB, and the globular cluster level, it is not surprising that it is found at the bound two star level as well. It is just that the LCDM paradigm, successful at the very large scale was already challenged at the galactic level, severely challenged at the globular cluster level, and pretty much broken at the binary level.

All the same, I, at least personally, am not confident that MOND or similar modified gravity theories are yet the right answer. Rather, they are a better first step in a possible direction. As it is they are being "protected" by a 7000 AU buffer zone and are of no help to addressing anomalies already observed deep within our solar system. For a while Milgrom and others attempted to go after some of these anomalies by abandoning Modified Gravity in favor of Modified Inertia. But, at least to me, they keep making the same (erronous) step that the LCDM makes: They all make the same arbitrary, a priori, fundamental assumption that there is only one co-efficient of inertia. Slightly further in a promissing direction is HOND which retrieves the observed inverse linear for the Dark Matter force instead of invese squared of Newton.

We can still keep the good Einstein theory so long as the used co-efficient of inertia is 3+1, the HOND-like co-efficinet is 2+1, Dark Energy is 1+1, and the "elephant in the Dark Matter livingroom" that dissipated back around z=2, cosmological dark matter is 0+1.

Cougar
2011-Jul-11, 12:33 PM
Alternatively, one can propose a contrasting scenario where gravity exhibits a change of regime at acceleration scales a<a0, and obtain just as good a fit to observations across astrophysical scales.

From what I've heard of MOND, the above assertion is false. When a0 is set to fit galactic dynamics, then it does not fit cluster dynamics. Therefore, such a scenario is not "just as good a fit to observations across astrophysical scales."

Tensor
2011-Jul-12, 04:33 AM
Tensor,
Thank you for your comments, but you might want to look at the 2nd version of the paper. It has a new title, is nearly twice as long, and has 10 charts as opposed to the original 4. There is also the inclusion of a study on the Hipparcos data, using only 280 samples as opposed to the over 1200 from the SDSS data, but with improved S/N ratio. Many of the questions regarding your first point are addressed in the extended paper.

Actually, I missed the first version and did read the second version (your link goes right to it). While they've done a rather thorough job of trying to eliminate errors, there are still several that they may not recognize (I say this as I've read the paper on Hipparcos (http://www.astro.umd.edu/~eshaya/binary/#A0.T1) data they used).

One of the problems is that the selection process, for some of the binaries, according to the Hipparcos paper, includes injecting a Keplerian orbit into the inputs (it's one of several different inputs) and then checking to see if any observed pairs match that simulation. If it does, it's included into the definition of a wide binary. If MOND works, as Hernandez et al claim, then an injected Keplerian orbit would not produce a match for observation, except by chance. Which bring up this part of the paper:

The paper states:
The only parameter of the catalogue to which the results of figure (1) are sensitive, is the threshold in the probability above chance alignment which one uses to define the final sample.

If Keplerian parameters don't work, as is claimed by the paper, then almost all of the alignments are chance.

The paper states:
The upper limit clearly shows the flat trend expected from modified gravity theories, clearly inconsistent with Keplerís third law, shown by the s−1/2 solid line.

And yet, one of the data used to verify stars are binaries in the Hipparcos catalogue, uses Keplers third law.

Another problem(well, for me) is the lack of knowing exactly what binaries they used from Hipparcos. Some of them were directly measured. If Hernandez et al used those, there is a lot less to be concerned about.



The tenor of their statement that no physical evidence of Dark Matter Particles still holds. We don't have a "beaker" full of DM particles yet.

Snip...

been no reports of any created masses in this range whether mostly hidden (WIMPs) or completely hidden (NIMPs) that are needed to conserve momentum.

I think you misunderstood me. I was just pointing out that both papers, each providing support to opposing sides, have not yet passed peer review. Both may not see publishing beyond Arxiv. We'll just have to wait and see.


Considering that the phnomenon that requires either some modified gravity theory or Dark Matter Particles to preserve Newton-Kepler-Einstein is apparent at extra-galactic, galactic especially including dwarf and LSB, and the globular cluster level, it is not surprising that it is found at the bound two star level as well.

Well, another problem is whether the stars in the catalogue are bound or escaping binaries. Henandez et at are a bit cavalier for my mind in dismissing Jiang & Tremaine definitions (http://arxiv.org/pdf/0907.2952v2) of escaping or dissolving systems. Hernandez's comments that
"...the total absence of any tell tale feature on crossing the 1.7pc tidal limit expected under Newtonian gravity argues in the same direction."

Indicates they either didn't read the paper or didn't understand the argument. As there isn't a sharp cutoff at 1.7pc.


It is just that the LCDM paradigm, successful at the very large scale was already challenged at the galactic level, severely challenged at the globular cluster level, and pretty much broken at the binary level.

I'll agree with you at the galactic and globular cluster level, not so much at the binary level.

borman
2011-Jul-24, 12:57 PM
Thanks for the Hipparcos paper.

Both the new paper and the Hipparcos paper have overlooked the impact upon wide binaries crossing the cusp divide. Depending on which paradigm one prefers, the LCDM says there is a cuspy hole in the dark matter halo closer to the galactic bulge where the rotation rates are more Newtonian and the faster rotation of wide binaries would be prohibited. Within the MOND paradigm, there is a similar External Field Effect, or EFE, where the ambient acceleration due to a larger external field rises above a_0 preventing a MOND effect from occuring. The Hipparcos paper only concerns itself with inverse squared law apllied to tidal mechanisms and only treating Dark Matter as sub-clumps that can be modelled as localised masses; but not considered is wide binaries within a larger halo nor the transition into the cusp region as regards the impact upon dissasociation. I.e., they are already rotating too fast about each other when the MOND or LCDM halo gravitational "rug is pulled out from underneath them" when they wander into either the cusp region or an EFE region. The effects upon wide binaries can be used to probe how dynamics are altered to see how or even if these regions behave or exist. The details of this research are completely buried in the data of both papers; one would have to sort out disassociated wide binaries near these cusp zones to see if their trajectories differ greatly from tidally disrupted wide binaries. The signature should be quite different for removal inverse linear effect as opposed to inverse squared tidal effects. Wandering into either a Dark Matter Particle Free Zone or a high ambient acceleration (>a_0) zone (MOND) would effectively "cut" the dark matter effect out and the resulting excess angular momentum between the wide binaries would cause them to fly apart from each other in a way that could not be explained by tidal disassociation... their redshifts would appear anomalous compared to tidal disruption of the binary.

borman
2012-Feb-13, 04:36 AM
Now, I will point out that neither of the papers have yet been peer reviewed, as of yet. There are good probabilities that both papers may end up going away after peer review and after further studies.


Version 4 has been reviewed and is being published. The publlished version still refers to both Hipparcos and SDSS. Their conclusions remain unchanged. While not supporting any particular theory be it MOND or F(r), curves tend to flatten where a<a_0 between wide binaries contrary to the expectations of purely Newtonian or Keplerian expectations. From a Dark Matter Particle perspective, each binary needs to be encased within their own personal bubble of dark matter when the distance between them exceeds 7000 AU in order to remain consistent with Newton.

Tensor
2012-Feb-13, 06:09 AM
Version 4 has been reviewed and is being published. The publlished version still refers to both Hipparcos and SDSS. Their conclusions remain unchanged. While not supporting any particular theory be it MOND or F(r), curves tend to flatten where a<a_0 between wide binaries contrary to the expectations of purely Newtonian or Keplerian expectations. From a Dark Matter Particle perspective, each binary needs to be encased within their own personal bubble of dark matter when the distance between them exceeds 7000 AU in order to remain consistent with Newton.

I still have a problem with it. As I pointed out before, the wide binaries in the Hipparcos catalog (they were binned in 0-25, 25-50, and 50-100, but the smallest was 11.3 parsecs) were determined to be wide binaries using Keplerian orbits. Yet, version 4 still says that their data excludes Keplerian orbits. How can they claim the orbits in their data exclude Keplerian data, when the orbits in their data were selected using Keplerian criteria?

Jerry
2012-Feb-17, 06:41 AM
I don't see a problem with sprinkling in a little dark stuff here; as it is currently in vogue to use it where ever observations are at odds with predictions.

On the other hand, what a good idea; to look for non-keplarian results at on the smallest reasonable scale. The authors may be in a gray zone; but hopefully this will attract some additional attention...and funding. Good Gaia project?

Meanwhile: http://arxiv.org/abs/1202.3629 MOND as the weak-field limit of an extended metric theory of gravity

I didn't include this in this list of fun papers, but I probably should have.

borman
2012-Feb-19, 09:52 PM
I don't see a problem with sprinkling in a little dark stuff here; as it is currently in vogue to use it where ever observations are at odds with predictions.

On the other hand, what a good idea; to look for non-keplarian results at on the smallest reasonable scale. The authors may be in a gray zone; but hopefully this will attract some additional attention...and funding. Good Gaia project?

Meanwhile: http://arxiv.org/abs/1202.3629 MOND as the weak-field limit of an extended metric theory of gravity

I didn't include this in this list of fun papers, but I probably should have.

The GAIA mission should deliver a third independent set of data to confirm Hernandez et al obsersavtions. This time velocity data may be read directly rather than being calculated as projected Delta V data. This was brought up in reference 16 of the paper, also refered to by Tensor as the Jiang and Tremaine paper.

Two possible observations regarding the paper: In paragraph 3, the authors state that no Dark Matter Effect was ever indicated in a zone of high surface gravity>>a_0. However, it is not yet ruled out that some of the anomalies deep within our solar system well are not associated with the effect. These include the AU, Lunar eccentricity secular increase, flyby anomalies, and possibly a portion of the Pioneer anomaly. Dark Matter is ruled out as its gravity would affect everything equally gravitationally, not just spacecraft. The conclusion that this represents a failure of Classical gravity may be a bit too strong. It may show the limitations of Classical gravity and that it is not the complete answer rather than being wrong. It could still be right but there is something *else* also going on that we do not quite understand yet .

The paper you cite could be added along with the others that purport either a modified or extended gravity theory, assuming it has anything to do with gravity in the classical sense, that the paper mentioned. The main goal of the paper was not to support any particular modified gravity model, but to test whether classical gravity gives the complete description of the expected velocities between wide binaries especially where the distance between them is so great that a<<a_0. The study supports that the usual Newtonian description works, including evolution and dissolution into galactic tides, where a>>a_0, but supports generic modified gravity theories where the Newtonian gravity is too weak having fallen below a_0 where the termianl velocity is a fourth power function of the combined mass, eq 2 of the paper,

borman
2012-Feb-19, 10:33 PM
I still have a problem with it. As I pointed out before, the wide binaries in the Hipparcos catalog (they were binned in 0-25, 25-50, and 50-100, but the smallest was 11.3 parsecs) were determined to be wide binaries using Keplerian orbits. Yet, version 4 still says that their data excludes Keplerian orbits. How can they claim the orbits in their data exclude Keplerian data, when the orbits in their data were selected using Keplerian criteria?

The point of the paper was to inquire if indeed all the binaries of the Shaya and Ollings paper who claim fewer than 10% false positives are really all Keplerian. To this end they took the data from their paper using delta nu and and delta theta columns data to caluculate the projected delta V and projected s data to do thier study. They made further exclusions of three star systems, rejecting binaries whose S/N fell below 0.3. For clarity reagrding possible 4th power mass functions intoducing fog into the picture, they also used binaries whose both masses were one solar mass so that the second equation did not produce an overly large range of values and so that the resulting Delta V would be in a similar range for most values. I.e. try to make Delta V constant for the chosen samples since (GMa_0)^1/4 would then be constant. In the appendix they did a consistency check by rejecting data, and so used a smaller sample, below a S/N of 2.2 . Their reuslts remained unchanged.

The issue with Jiang and Tremaine is not because they find fault with their study or else they would not cite it (reference 16). This study dealt only with Newtonian gravity and evolution and dissolution of binaries within Newtonian constraints without any regard of the possible impact of modified gravity theories where the ambient acceleration drops below a_0. Hernandez et al do take a look at the diferences in tidal disruption occuring in the two different regimes of Newton, a>>a_0 and Modified gravity, a<<a_0, and find that the weak regime is more robust against dissolution than the Newtonian regime. The result is that the situation is more complicated since dissolution is going on under not one or the other, but both regimes where the criteria of whether a_0 is being exceeeded or not. We have a mixture of both and they are not color coded to help separate one from the other.

Tensor
2012-Feb-21, 05:54 AM
The point of the paper was to inquire if indeed all the binaries of the Shaya and Ollings paper who claim fewer than 10% false positives are really all Keplerian. To this end they took the data from their paper using delta nu and and delta theta columns data to caluculate the projected delta V and projected s data to do thier study. They made further exclusions of three star systems, rejecting binaries whose S/N fell below 0.3. For clarity reagrding possible 4th power mass functions intoducing fog into the picture, they also used binaries whose both masses were one solar mass so that the second equation did not produce an overly large range of values and so that the resulting Delta V would be in a similar range for most values. I.e. try to make Delta V constant for the chosen samples since (GMa_0)^1/4 would then be constant. In the appendix they did a consistency check by rejecting data, and so used a smaller sample, below a S/N of 2.2 . Their reuslts remained unchanged.

You (and they) still didn't answer my question. Taken from the paper: "We have taken this catalogue and kept only binaries with a probability of non-chance alignment greater than 0.9." So, how did the Hipparcos catalog determine the probability on whether the binaries were in a chance alignment or actually in orbit around each other? By injecting Keplerian parameters into orbital simulations, and comparing what the simulations showed against what was observed. So, if the actual binaries matched something from a simulation, they were classified as probable binaries. To quote from the paper "...The upper limit shows the flat trend expected from modified gravity theories, at odds with Keplerís third law..." But, if they are using Keplerian parameters, how do you determine the orbits are not Keplerian, when you used Kepelerian data to determine the orbit?


The issue with Jiang and Tremaine is not because they find fault with their study or else they would not cite it (reference 16). This study dealt only with Newtonian gravity and evolution and dissolution of binaries within Newtonian constraints without any regard of the possible impact of modified gravity theories where the ambient acceleration drops below a_0. Hernandez et al do take a look at the diferences in tidal disruption occuring in the two different regimes of Newton, a>>a_0 and Modified gravity, a<<a_0, and find that the weak regime is more robust against dissolution than the Newtonian regime. The result is that the situation is more complicated since dissolution is going on under not one or the other, but both regimes where the criteria of whether a_0 is being exceeeded or not. We have a mixture of both and they are not color coded to help separate one from the other.


No, it's not just dissolution of the binaries. It also dealt with the what the difference in the change in relative velocity, dependent on the separation, and how that compared with the different theory predictions. I see where the the breakup of the binary system at > 7000 pc needs a mechanism in both systems. I can't and won't comment on the other catalogs that were used, as I don't (at this time) know how the other catalogs determined whether binaries were or were not actually orbiting each other. But, that claim of finding non-Keplerian parameters when Keplerian parameters were used bothers me.