Thread by @ReadDark

We have found observational evidence that dark matter is “heated up” at the centres of dwarf galaxies:

arxiv.org/abs/1808.06634…

This solves the long-standing “cusp-core” problem in LCDM, and suggests that dark matter is a cold, collisionless, fluid.

[A thread]

I gave a ~20 minutes talk at KITP about this paper which you can watch here:

online.kitp.ucsb.edu/online/cdm-c18…

For a Twitter summary of the results, please read on ... [1/N]

On large scales, the standard cosmological model (LCDM) gives an excellent description of the cosmic microwave background radiation and the growth of structure. However, on small scales, in our cosmic backyard known as the Local Group, there have been long-standing puzzles. [2/N]

The oldest of these is the “cusp-core” problem. Gas rich dwarf galaxies have much less dark matter (DM) in their inner regions than early numerical models predicted. This could point to more exotic DM models, or to a failure of the cosmological model. [3/N]

However, there is a simpler solution. The early numerical models, above, considered a universe that contains only DM. Recent models that include gas cooling, star formation, and gas blow-out due to exploding stars find that star formation (SF) in dwarfs is bursty. [4/N]

This bursty SF occurs due to repeated cycles of gas inflow and outflow. These cause the inner gravitational potential of the dwarf galaxy to fluctuate, kinematically “heating up” the DM and lowering its inner density [review -> arxiv.org/abs/1402.1764]. [5/N]

A key prediction of such models is that, at a fixed DM halo mass, more SF leads to more DM “heating” and, therefore, a lower inner DM density, at least in dwarf galaxies. This is the prediction we set out to test in this paper. [6/N]

We developed new tools to measure the inner DM density of 16 dwarf galaxies with excellent quality data and a wide range of SF histories. These tools were extensively tested on mock data for which we knew the correct answer. [7/N]

Some of our sample of dwarfs stopped forming stars just ~2 billion years after the big bang; others are still forming stars today. If “DM heating” is a real phenomenon, the former should be substantially denser than the latter. [8/N]

These two plots compare the DM density profile of the Draco dwarf that stopped forming stars ~2 billion years after the big bang (left) with the Fornax dwarf that stopped forming stars just ~1 billion years ago (right). Notice how much denser Draco is than Fornax. [9/N]

We then compared the inner DM densities of our full sample of dwarfs against their stellar masses. We found that galaxies that stopped forming stars long ago (black) are systematically denser than those that formed stars until recently (blue). [10/N]

This scatter may owe, however, to these galaxies inhabiting DM halos with different pre-infall masses. For this reason, we estimated pre-infall halo masses for all of our dwarfs using abundance matching (c.f. this paper: arxiv.org/abs/1807.07093). [11/N]

With this, we can plot the inner DM density versus halo mass. Notice that galaxies with no recent SF (black) lie along the grey track, consistent with no DM heating. Those with recent SF (blue) lie along the blue track, consistent with maximal DM heating. [12/N]

We can see this another way by plotting the inner DM density versus the ratio of stellar mass to halo mass. Now we see a clear anti-correlation: galaxies with more SF (i.e. higher M*/M200) have lower inner DM densities. This is exactly as predicted by DM heating models. [13/N]

Finally, we can ask: are these results are consistent with other DM models? Most models designed to solve the cusp-core problem predict no correlation between the inner DM density and SF. This is not what we find here. “Alternative gravity” models also run into trouble. [14/N]

We find a pair of galaxies - Carina and Draco - with very similar stellar mass profile and orbit around the Milky Way, but very different gravitational fields. In the standard cosmological model, this is explained by their different pre-infall halo masses. [15/N]

In MOND, Carina and/or Draco must be out of equilibrium. This is hard to understand given their benign orbits. Assuming equilibrium, the MOND predictions give a poor match to the data (compare the solid and dashed lines with the contours in this plot). [16/N]

The above results suggest that, to leading order, DM is a cold, collisionless, fluid that can be kinematically “heated up” and moved around. This is consistent with DM being a new weakly interacting particle. But until we find the particle, we must remain open-minded. [End]