Thursday, February 27, 2020

Galaxy Formation Simulations Disfavor Warm Dark Matter

A new paper compares simulated galaxy formation in cold dark matter (CDM), self-interacting dark matter (SIDM), and warm dark matter (WDM) models. In the simulation, SIDM models produce galaxies at about the same time as CDM models, while WDM models produce galaxies much later.

Empirically, galaxies are observed to form significantly earlier than predicted in CDM models. This is called the "impossible early galaxies" problem. But, according to this stimulation, self-interacting dark matter models do not to solve the slow galaxy formation problem found in CDM models, and warm dark matter models have the same late galaxy formation problem as CDM models, but one of that is much worse in magnitude, because galaxies form more slowly in WDM models than in CDM or SIDM models. 

This is also a big problem for the dark matter paradigm generally, because the CDM model does not produce the galaxy scale inferred dark matter distributions that are observed. This is the main problem that the SIDM and WDM models were devised to solve and is why these two dark matter theories are the most promising dark matter particle theories currently being considered. 

Of the two, WDM was more attractive than SIDM in many respects, because WDM requires just one kind of dark matter particle, while SIDM requires both a new fundamental dark matter particle and a new dark matter interaction force carrying particle and more free parameters. Also, WDM might have been possible to tie into neutrino physics and the hints of a possible sterile neutrino in measurements of neutrinos oscillations at nuclear reactors.

But, each of these solutions to the galaxy scale problems of CDM (and neither SIDM nor WDM really do a great job of solving that) fails to address the impossible early galaxies problem of CDM. SIDM provides no improvement on this front, and WDM makes the impossible early galaxies problem of CDM significantly worse. So, this paper is a meaningful blow to WDM theories and provides no signs of encouragement for SIDM theories, which an optimist who had not seen the simulation data might have wishfully hoped could have solved the impossible early galaxies problem in addition to improving the galaxy scale behavior of CDM theories.

The paper and its abstract are as follows:

Local Group star formation in warm and self-interacting dark matter cosmologies

Mark R. Lovell (1,2), Wojciech Hellwing (3), Aaron Ludlow (4), Jesús Zavala (1), Andrew Robertson (2), Azadeh Fattahi (2), Carlos S. Frenk (2), Jennifer Hardwick (4) ((1) University of Iceland, (2) ICC Durham, (3) Warsaw, (4) ICRAR/UWA)
The nature of the dark matter can affect the collapse time of dark matter haloes, and can therefore be imprinted in observables such as the stellar population ages and star formation histories of dwarf galaxies. In this paper we use high resolution hydrodynamical simulations of Local Group-analogue (LG) volumes in cold dark matter (CDM), sterile neutrino warm dark matter (WDM) and self-interacting dark matter (SIDM) models with the EAGLE galaxy formation code to study how galaxy formation times change with dark matter model. We are able to identify the same haloes in different simulations, since they share the same initial density field phases. We show that the stellar mass varies systematically with resolution by over a factor of two, in a manner that depends on the final stellar mass. The evolution of the stellar populations in SIDM is largely identical to that of CDM, but in WDM early star formation is instead suppressed. The time at which LG haloes can begin to form stars through atomic cooling is delayed by 200~Myr in WDM models compared to CDM. 70~per~cent of WDM haloes of mass >108M collapse early enough to form stars before z=6, compared to 90~per~cent of CDM and SIDM galaxies. It will be necessary to measure stellar ages of old populations to a precision of better than 100~Myr, and to address degeneracies with the redshift of reionization, in order to use these observables to distinguish between dark matter models.
Comments:17 pages, 13 figures. To be submitted to MNRAS. Contact: lovell@hi.is
Subjects:Astrophysics of Galaxies (astro-ph.GA); Cosmology and Nongalactic Astrophysics (astro-ph.CO); High Energy Physics - Phenomenology (hep-ph)
Cite as:arXiv:2002.11129 [astro-ph.GA]
 (or arXiv:2002.11129v1 [astro-ph.GA] for this version)

8 comments:

neo said...

maybe the estimates of age of galaxies is off

andrew said...

Way down the list of plausible explanations.

neo said...

what conclusion would you draw if a paper gets posted that shows Galaxy Formation Simulations Disfavor MOND and any modification of gravity like Deur's?

andrew said...

"what conclusion would you draw if a paper gets posted that shows Galaxy Formation Simulations Disfavor MOND and any modification of gravity like Deur's?"

I'd kick the tires carefully to see if it is flawed.

If it favored another view, I'd take that seriously and modify my views as I have more than once in the past. I don't think that a simulation can ever been truly definitive in a matter so complex. There are so many ways to screw it up and even the best simulations are much more crude than most people realize, to a significant extent, due to poor resolution.

Science is a puzzle with lots of pieces. Sometimes, when you part way though you have an insight about what the whole picture looks like. Often you're reasonable close, but sometimes, you are wrong. But, it is also true that as you get more pieces of the puzzle, a major paradigm change is increasingly unlikely. The evidence favoring some manner of gravity modification theory over dark matter particles is pretty strong at this point.

neo said...

i'd like to see simulation using both MOND and dark matter

andrew said...

"i'd like to see simulation using both MOND and dark matter"

To get something that worked, you' have to devise some kind of dark matter that confines dark matter mostly to galactic clusters and it isn't obvious what particle properties could cause that to happen.

neo said...

what about MOND + neutrino hot dark matter?

while neutrinos appear to have masses less than 1 ev, what if there are more of them than predicted by CMB?

andrew said...

So far, there are no neutrino discrepancies from lambdaCDM.

Neutrino hot dark matter cannot be the source of any dark matter or dark energy phenomena. The dark matter phenomena we see, generically, gives rise to more structure than expected with Newtonian gravity and baryonic matter, which is what we see. Hot dark matter would give rise to less structure than expected, which is not what we see.