Dark matter particle search at CERN vs cosmology simulations

[ohannuks]

Dark matter particle candidates are being searched at CERN and the various dark matter models are being probed by cosmological simulations. The usual way to probe models via cosmology is to plug a candidate into a cosmological simulation and then compare the results with observations. Yet another way is to calculate cross-sections of potential candidates and estimate their effects near compact objects ( e.g. http://arxiv.org/abs/1506.04143 )

There is also research in modelling various effects of different models.

How much do you believe the various probes at cosmological models will help in searching for the dark matter particle? Do you believe building various computational models for different DM candidates is premature at this point?

 

[Drakkith]

Do you believe building various computational models for different DM candidates is premature at this point?

Not at all. Building and testing models is exactly how science gets done.

 

[ohannuks]

Not at all. Building and testing models is exactly how science gets done.

Thank you for your answer Drakkith. I was referring to the process of building further while testing. One can evaluate the pay-off of building and testing a model. This pay-off can be huge, minimal, or somewhere in-between.

Perhaps I can elaborate; I would like to know if putting much effort into building computational (cosmology) models for investigating dark matter candidates at this point will bear fruit, taking into account the fact that particle physics experiments could soon tell us much about the viable candidates.

E: Further elaboration: As an example, I would like to know people's opinion on investigating e.g. effects of one specific candidate on rotational evolution of neutron stars or similar phenomena. Should the models be built so far for every candidate in the zoo of dark matter candidates at this point?

 

[Chronos]

It requires great ingenuity to devise a working experiment to test any theoretical model, but, it does happen. These instances are typically hailed as scientific breakthroughs.

 

[Finny]

I would like to know if putting much effort into building computational (cosmology) models for investigating dark matter candidates at this point will bear fruit, taking into account the fact that particle physics experiments could soon tell us much about the viable candidates.

Uh, 'particle physics experiments' is exactly what CERN does. Their detectors as focused on pinpoint particle detections and associated telltale signs...tracks, scatterings, particle productions and so forth.
I've not heard of them doing any macroscopic neutron star sized experiments.

I would like to know people's opinion on investigating e.g. effects of one specific candidate on rotational evolution of neutron stars or similar phenomena. Should the models be built so far for every candidate in the zoo of dark matter candidates at this point?

I did not know the rotational evolution of a neutron star has much to do with dark matter. Other types of exotic matter...Neutron, quark, strange, degenerate matter...yes.

Regarding model building and experimental detection in general: Science has a rather incredible history of mathematical predictions sometimes leading to discoveries. If you don't know what you are looking for, what kinds of detections to expect, it's a 'hit or miss' proposition.

For example, people had been looking at galaxies for a long time before Vera Rubin came along and decided to look at rotational speeds of galaxies...She had different ideas, so of course conventional establishment types didn't pay much attention. She found anomalies that all before her had missed. Some decades later her discovery led to the 'discovery' of dark matter.

But experimentalists do lead the way in other cases. I think dark matter itself is an example. It was predicted way before discovery in 1933 by that crazy theorist Fritz Zwicky, but he was dismissive of his peers and they of him. I don't think Zwicky had a 'model', just some rather incredible intuition.

Probably best to model and or seek what your intuition tells you.

Of course most models do not always pan out. Even Einstein had to try over and over to get the stress energy tensor mathematics for GR in a format that would lead to the predictions he foresaw. He tried a bunch and I think it was his math professor, Herman Minkowsky, who found the math Einstein needed.

Here are some major experiments at CERN...seem to be particle sized efforts.

http://home.web.cern.ch/about/experiments

 

[ohannuks]

Thanks Chronos.

Uh, 'particle physics experiments' is exactly what CERN does. Their detectors as focused on pinpoint particle detections and associated telltale signs...tracks, scatterings, particle productions and so forth.
I've not heard of them doing any macroscopic neutron star sized experiments.

Yes, I mean CERN. What do you mean by macroscopic neutron star sized experiments?

I did not know the rotational evolution of a neutron star has much to do with dark matter. Other types of exotic matter...Neutron, quark, strange, degenerate matter...yes. [\QUOTE]

I'm sure you can dig something up.

Regarding model building and experimental detection in general: Science has a rather incredible history of mathematical predictions sometimes leading to discoveries. If you don't know what you are looking for, what kinds of detections to expect, it's a 'hit or miss' proposition.[\QUOTE]
Many of the dark matter candidates already have fairly narrowed properties, and any viable candidate should be derived from an existing theory.

For example, people had been looking at galaxies for a long time before Vera Rubin came along and decided to look at rotational speeds of galaxies...She had different ideas, so of course conventional establishment types didn't pay much attention. She found anomalies that all before her had missed. Some decades later her discovery led to the 'discovery' of dark matter.

But experimentalists do lead the way in other cases. I think dark matter itself is an example. It was predicted way before discovery in 1933 by that crazy theorist Fritz Zwicky, but he was dismissive of his peers and they of him. I don't think Zwicky had a 'model', just some rather incredible intuition.

Probably best to model and or seek what your intuition tells you.

Of course most models do not always pan out. Even Einstein had to try over and over to get the stress energy tensor mathematics for GR in a format that would lead to the predictions he foresaw. He tried a bunch and I think it was his math professor, Herman Minkowsky, who found the math Einstein needed.

Here are some major experiments at CERN...seem to be particle sized efforts.

http://home.web.cern.ch/about/experiments

[/QUOTE]
Cheers. The truth is that there are also many theorists out there making very strange theories without any predictions or backup from observations. Some even work on theories which are doomed to fail or they're investigating things in an inefficient way.

Perhaps I did not articulate the topic clearly.
This topic is about discussing whether currently investigation into various dark matter models via computational simulations of cosmology is currently fruitful. As an example, it seems some people are starting to believe heavy WIMPs are not that great candidates: http://www.nature.com/news/crunch-time-for-pet-theory-on-dark-matter-1.16757 .

 

[Finny]

What do you mean by macroscopic neutron star sized experiments?

https://en.wikipedia.org/wiki/Neutron_star#Rotation

"A neutron star is a type of compact star that can result from the gravitational collapse of a massive star after a supernova. Neutron stars are the densest and smallest stars known to exist in the universe; with a radius of only about 12–13 km (7 mi), they can have a mass of about two times that of the Sun."

In other words, point particle detections are microscopic, neutron stars are macroscopic...big.

 

[Chronos]

The case for WIMPS has seen little progress over the past decade. Despite a number of detection efforts, no definitive results have been achieved to date. While other models like sterile neutrinos and axions have gathered interest, they suffer this same deficit. There remains hope the current LHC run may produce something noteworthy, but, it looks like a long shot

 

[mfb]

Perhaps I can elaborate; I would like to know if putting much effort into building computational (cosmology) models for investigating dark matter candidates at this point will bear fruit, taking into account the fact that particle physics experiments could soon tell us much about the viable candidates.

Everything that is not excluded yet can be interesting, and the models the LHC experiments and others can exclude or find are the most interesting ones - the experimentalists need input where to search for them and how to exclude or find them.
Models that predict new particles at 10¹⁵ GeV and nothing else are quite pointless as there is no way to test them in the foreseeable future.

 

[ohwilleke]

Dark matter particle candidates are being searched at CERN and the various dark matter models are being probed by cosmological simulations. The usual way to probe models via cosmology is to plug a candidate into a cosmological simulation and then compare the results with observations. Yet another way is to calculate cross-sections of potential candidates and estimate their effects near compact objects ( e.g. http://arxiv.org/abs/1506.04143 )

There is also research in modelling various effects of different models.

How much do you believe the various probes at cosmological models will help in searching for the dark matter particle? Do you believe building various computational models for different DM candidates is premature at this point?

No. The cosmological models and the astronomy observations are by far the more promising avenues at this point.

CERN and previous high energy physics experiments have all but ruled out any dark matter candidates that has electromagnetic, weak or strong force interactions up to several hundred GeV of mass, and direct dark matter detection experiments have ruled out particles that interact at least a strongly as a neutrino down to about 1 GeV.

The CERN PR campaign to sell it as a means of detecting dark matter, to be honest, verges on scientific dishonesty. There is virtually no realistic prospect that a dark matter candidate will be detected there. If it could have been, direct dark matter detection experiments would have already found it, and it would really be a poor fit to astronomy data constraining the DM parameter space.

Whole universe cosmology observations tell you how much dark matter should be in the universe as a whole, but are not very sensitive to the mass of the dark matter candidate - anything from 1 keV to many TeV could fit the lambdaCDM model.

The best prospect for identifying DM candidates at this point is to use astronomy data in the range from dwarf galaxies to satellite galaxies of spiral and elliptical galaxies to galactic clusters to pin down the properties of DM based upon what can be inferred from the dynamics of luminous objects, from gravitational lensing, and from the overall structure of the universe at that scale which is observed at various red shifts of age relative to what would be predicted in simulations of universes with various kinds of DM in it at those ages. Those dynamics and the behavior in simulations can allow astronomers to infer distributions of DM in the universe relative to luminous objects, and use that to reverse engineer the properties DM would have to have to distribute itself like that. Colliding galactic clusters that allow for differentiation of luminous and DM halo shapes and inferred interaction cross-sections can be particularly helpful. And, the more we understand precisely what the make up of ordinary matter (especially gas and dust) in a galaxy or galactic cluster is, the more accurately we can pinpoint the inferred behavior of DM.

This work has been ongoing since the 1980s, and improved astronomy observations from satellite based telescopes along with improved computational power from better and cheaper computers has dramatically improved the search.

The good new is that this endeavor has been quite fruitful. The bad news is that some subtle details of those simulations, some of which have been adopted for sake of computational ease, turn to matter a great deal in the results, and to some extent, the DM candidate searches are overconstrained.

In the 1980s there was a great deal of hope that weakly interacting massive particles predicted to exist by supersymmetry models at the tens to hundreds of GeV mass scale would produce dark matter with the right properties to match observation, at the right relic density, with the right distribution. Scientists calculated from first principles what kind of distribution the simplest kind of dark matter would naturally clump into which is called the NFW distribution (after the initials of the people who discovered it). The generic dark matter description used for the lamdaCDM model worked well to reproduce the observed cosmology. Potential baryonic dark matter (i.e. dark matter made of ordinary matter like gas giants and hydrogen and helium gas) and potential neutrino dark matter was ruled out. Very crude simulations crudely reproduced what we see in the universe. A number of cosmic ray lines from deep space have been proposed as dark matter annihilation signals.

The reality a generation later is more problematic. Almost all of the cosmic ray signals have been ruled out, one by one, in favor of more mundane sources, although a couple are still being actively investigated. The parameter space of supersymmetry models has been pushed ever higher (with the lightest superpartners in at least the hundreds of GeV masses and most superpartners in the TeV energy scale) while astronomy observations are finding that thermal relic particles that are the best fit to the data would have to have masses on the order of keV and no weak force, strong force or electromagnetic interactions (although lighter or heavier non-thermal relic particles with the same mean velocity would also work) which are not predicted to exist by supersymmetry and have not been seen in experiments (because this would be almost impossible to detect). Direct dark matter searches have come up empty. Dynamically inferred dark matter halos turn out not to look much like the NFW distribution (an issue called the core-cusp problem) and instead look more like galactic scale rugby balls with an "isothermal" distribution. Regularities in galactic dynamics like the Tully-Fischer relation between galaxy size and rotation curve behavior are tighter fits than random accumulations of dark matter in halos for which more scatter would be expected would imply. There are too many bulgeless spiral galaxies for the hypothetical galaxy formation process in CDM theories, and there are too few satellite galaxies for thermal relic DM candidates in the 1-500+ GeV mass range. DM theories also have a poor track record of accurately predicting new phenomena in advance and many have too many parameters. Models with more than one DM fermion and one DM self-interaction boson almost always underperform simpler models and there is suggestive evidence from the small number of parameters needed by modified gravity theories to match the data over broad ranges, that the right DM solution should not require much complexity in terms of a complex dark sector to get it right.

There have been a few responses to these criticisms. First, if DM is really and truly sterile it may just be effectively impossible to directly detect it or to produce it at CERN, and every indications so far is that this is the case. All evidence points to a DM candidate that does not interact with ordinary matter in any appreciable way. Second, some of the failure of DM candidates to perform in simulations may be due to a failure to accurately model the interactions of ordinary matter with dark matter via gravity properly, rather than the wrong DM properties put into the simulations. Third, there are still perhaps five plausible DM candidates left standing, roughly in order of viability: (1) singlet keV mass sterile neutrino like DM, (2) GeV or TeV mass scale DM that allows DM to interact with other DM via MeV mass scale bosons (sometimes called dark photons) with a coupling constant strength comparable to that of the electromagnetic force, (3) axions to which some of the usual DM rules don't apply because these very light particles are not thermal relic, (4) some sort of simple bosonic DM probably in the form of a Bose-Einstein condensate of some form, and (5) some sort of Chayplin gas (I'm sure I'm spelling that wrong) which addresses both DM and dark energy in the same substance. Fourth, there have been a few reasonably good modified gravity alternatives to DM do a decent job matching the astronomy data, although others have crashed and burned when confronted with data. We are pretty much down to testing less than a dozen remaining plausible DM and modified gravity models (with some subtle tweaks to each of the possibilities for fine tuning purposes) against the data after having ruled out everything else.

Right now, the most promising course is to make more astronomy observations that can discriminate between models, and more set up better simulations, in the honest hope that the data don't leave us overconstrained to the point where all possible DM candidates and all conceived modified gravity theories are ruled out.

In truth, CERN is exceedingly unlikely to discover a dark matter candidate, because the DM candidate parameter space that CERN hasn't explored but could explore in the near future, is already strongly disfavored by direct dark matter detection experiments and inferred dark matter behavior. TeV mass and near TeV mass particles that have interactions that would allow them to be produced or detected at the LHC are almost surely not good DM candidates, and even less so if they are not stable, yet that is what the LHC will find, if it finds any new particles in the next few years.

Likewise, direct dark matter detection experiments are nearing the end of their rope. The only area direct dark matter detection experiment can probe, but haven't is sub-GeV mass dark matter that interacts in some way with ordinary matter. But, any such fundamental or composite DM candidate that can be produced by a W or Z boson decay, or by photons or gluons, has already been ruled out long ago in HEP experiments. If light DM doesn't show up in experiments of those kinds, it is unlikely to show up in direct detection experiments either which rely on the same fundamental forces to produce "hits".