The $\Lambda\rm CDM$ (cold dark matter with cosmological constant) is the current standard model of cosmology because the model comes with a long list of phenomena successfully explained by it. However, there are a remaining handful of problems which are not yet resolved in the context of $\Lambda\rm CDM$, for instance (with links to my picks for reasonably accessible and/or up to date technical papers):

These (perceived? it is as yet unclear whether these issues can be resolved within the $\Lambda\rm CDM$ framework) problems have prompted many research groups to look at alternate theories. Some examine modifications or alternatives to general relativity - since much of the evidence for the existence of dark matter assumes GR, an alternate theory of gravity might make DM obsolete. For the purposes of this question, I want to retain the assumption that GR is correct. The other approach is to question the cold part of CDM. There is evidence ruling out hot (i.e. relativistic) dark matter, but so called "warm" dark matter (WDM) is an area of active research. There's also been some buzz about self-interacting dark matter (SIDM, i.e. interaction couplings within the dark sector). There are a number of papers claiming solutions to the $\Lambda\rm CDM$ problems in the framework of WDM or SIDM, or rather more conservatively the magnitude of the problems can be at least alleviated with alternate dark matter models.

However, I assume that the scientific community hasn't fully embraced WDM because it has trouble in other areas where CDM is just fine. What is/are the observations that WDM/SIDM/other-alternate-DM have trouble explaining that prevent them from replacing CDM as the standard model of cosmology? Or is one of these alternate models now competitive with CDM and we just need compelling evidence that WDM solves the remaining problems in CDM before re-writing the textbooks, so to speak?

  • $\begingroup$ Hi Kyle, I subbed the Nature link to the (now available) arXiv version; I imagine this is in the spirit of the post. Good question! $\endgroup$ Mar 28, 2016 at 19:22
  • $\begingroup$ @EmilioPisanty thanks, yes of course it is. Still hoping for a decent answer to this, though it's admittedly a bit of an unsolved problem. $\endgroup$
    – Kyle Oman
    Mar 29, 2016 at 21:12
  • $\begingroup$ Yeah, well, that's sort of what happens when you ask big, broad, ambitious questions like this one ;-). $\endgroup$ Mar 29, 2016 at 21:26
  • $\begingroup$ Extraordinary claims require extraordinary evidence. CDM is simpler than WDM and all the problems you mention may have simpler explanations based on the effects of baryonic physics. $\endgroup$
    – Virgo
    Dec 31, 2016 at 1:39
  • $\begingroup$ @Virgo You are incorrect. WDM is not a more complex theory or an extraordinary claim relative to CDM. The only difference is in the mass of the hypothetical dark matter particle. They are essentially identical theories except for the value of a single parameter. $\endgroup$
    – ohwilleke
    Jan 8, 2017 at 22:40

3 Answers 3


I stumbled on this question and thought I will give it a try, hopefully it is still relevant to you. The thing with all the issues you cite is that they have a solution within $\Lambda$CDM

  1. Too Big to Fail problem I think this problem is overrated. Within a couple of weeks of its publication people already found solutions, even with more colorful names: Vera-Ciro et al. (2012), Wang et al. (2012), Brooks et al. (2012) ... and the list continues. Point is that it is not an unsolvable problem within the list of things we know. For example, the first two references argue that the number of massive satellites strongly depends on the mass of the MW, which is actually a pretty tough number of measure, so moving the virial mass of the MW to the lower end of the spectrum is perfectly valid.

    Baryonic solutions to the problem are also available, based mostly on feedback. But here is the deal: if you throw away CDM in favor of WDM, these two things (MW mass and feedback) are still going to be there, the result is that you're going to overshoot the problem in the other direction. Now you will end up with the opposite problem. I will call this argument (A1).

  2. Cusp-Core problem Again, we know solutions for this problem within $\Lambda$CDM, a couple of SN can cause a lot of damage. Even with a modest stellar formation history in a gas-poor dwarf you can expect to have a modification in the inner slope of the host dark halo, e.g. Breddels et al. 2015. I've never tested this myself, but I have the feeling that if you run a simulation with enough resolution to resolve small structures in WDM and add feedback in a self consistent way you will end-up with no cusps at all. Again (A1) applies.

    Unfortunately, I don't think WHM is popular enough to try this type of simulations with the level of details of say Illustris

  3. Missing satellite problem I don't think this is a problem at all. It just reflects the free market nature of physics: big things end up bigger, small things ... well. You can come up with a convoluted story to about the way dark matter could affect this problem, but at the end a simpler solution will work better: cooling in small halos is insanely inefficient.


I will stop here, but I think you can see the pattern. Cooling is a real issue, baryonic feedback is real, mass uncertainties are real. If you solve all the issues you post with replacing CDM with WDM all these processes will still be there, affecting how galaxies form, and the solution that WDM offers will not longer be valid.

  • 1
    $\begingroup$ There are simulations with WDM on par with projects like Illustris in terms of detail, looking at these problems. The one I know most about is a variant on the APOSTLE simulations (themselves a high-resolution variant of the EAGLE simulations, which are similar in many respects to Illustris) with sterile neutrino dark matter. See e.g. this and this. One of the main results is that many of the same solutions (or small variations) work for the same problems. $\endgroup$
    – Kyle Oman
    Dec 17, 2018 at 12:49
  • $\begingroup$ ...but anyway, nice answer and +1. $\endgroup$
    – Kyle Oman
    Dec 17, 2018 at 12:50
  • $\begingroup$ @KyleOman I see, unfortunately I lost track of these simulations when I left the field. Happy holidays :) $\endgroup$
    – caverac
    Dec 17, 2018 at 13:05
  • $\begingroup$ You don't explain what MW and SN are. $\endgroup$ Mar 5, 2019 at 14:46
  • $\begingroup$ @HarryWilson Sorry for the lack of detail: MW = Milky Way, SN = Supernova. Hopefully it makes sense now $\endgroup$
    – caverac
    Mar 5, 2019 at 16:43

The "C" for cold in lambda CDM is a false friend of the "C" for cold in CDM as contrasted with WDM (with "W" for warm).

In the lambda CDM model, cold dark matter is defined more broadly to include any dark matter that is not "hot" (for example, thermal relic dark matter that has particle mass >> 1 eV), because for cosmology purposes one doesn't need a very specific definition of cold dark matter. The lambda CDM definition includes both WDM and CDM.

In contrast, when you are distinguishing between WDM models and CDM models, WDM refers to dark matter particles on the order of keV masses, while CDM refers to particles on the order of 1 GeV+ masses. The lambda CDM model is agnostic as between WDM and CDM models.

A SIDM model would complicate the math of the lambdaCDM model without improving its fit to the parameters that this model describes, so it is rejected out of parsimony until such time as WDM and CDM are rejected empirically, if ever.

  • $\begingroup$ While this may be true, it doesn't really answer the actual question, which is why are CDM models (as in ~1GeV WIMPs) the de facto standard assumption rather than WDM (e.g. sterile neutrinos). Neither is ruled out, and there seems to be perhaps a bit more weak circumstantial evidence in favour of WDM than in favour of CDM. Perhaps it's just that CDM has been well-entrenched for a long time? $\endgroup$
    – Kyle Oman
    Oct 13, 2016 at 18:54
  • 1
    $\begingroup$ I honestly don't think that there is much overlap between the people doing the lamdaCDM cosmology work with data from WMAP and Planck and BAO, who honestly don't care very much, and the people doing the mostly galaxy scale WDM and CDM and GAMA (GAlaxy Mass Assembly) work who aren't very concerned about cosmology and in that community there are also lots of non-astronomers pushing links to speculative particle physics ideas who simply refuse to look at evidence in other DM subfields that disfavor their latest and greatest idea. Inertia is the main factor. Also CDM is more SUSY friendly. $\endgroup$
    – ohwilleke
    Oct 13, 2016 at 22:30

I think that both CDM and WDM have first to conform to the constraint imposed by the Cosmological Principle. If I remember correctly, Einstein stated this principle which may be found in the book, "The Principle of Relativity", Dover Publications, English translation published in 1952. I paraphrase his principle as...Observables and objects in the Universe may be treated as being uniformly distributed...we live in a solar system, within a galaxy, which is nothing special. This is simply an expanded version of the useful Copernicus viewpoint.

Up to now CDM has always been located "out there" and "over there". Never in our solar system. If one allows for the Cosmological Principle then the center of our earth should be dense with CDM. I have not read of such claim by geologists. I have read of such claim of CDM residing in our sun, in an arXiv article a few years ago. So should one abandon the Cosmological Principle or the idea of WDM and CDM? To embrace both concepts does not seem consistent with the data.

  • 1
    $\begingroup$ The cosmological principle applies on large scales, not small ones where there are extremely obvious and very well-known variations in very many properties: the existence of planets, for instance, does not invalidate the cosmological principle. $\endgroup$
    – user107153
    Nov 21, 2016 at 0:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.