This is a review of the recent paper, to be published in Nature, by Andrew Potzen and Fabio Governato Cold dark matter heats up. The paper is itself a review, containing information published in previous studies. I gave a presentation about it at our weekly Astrophysics journal club. The figure in the header is M82, and we can see recombining hydrogen in outflowing gas in purple. Recent observations of exactly such outflowing gas will become important in the exposition of the paper.
Modern cosmology holds what is know as concordance model to be true-one incorporating both a cosmological constant, known as lambda Λ and cold dark matter or CDM. Together these components comprise over 95% of the mass-energy budget of the cosmos. The model thus is deemed ΛCDM and comprises our best theoretical candidate to fit observations.
The good news it that as CDM interacts only gravitationally so we can model it in a more straightforward manner in large scale simulations than more complicated substances. When these simulations are performed they are on the whole a robust match with observations. We wouldn’t expect them to be a perfect match, however, because such simulations ignore many physical processes pertaining to baryonic matter-the stuff that you and I are made of, among other reasons, resolution and a phenomenon known as cosmic variance among them. One such place this mismatch is apparent is if you measure the density as a function of radius or the density profile of the object from its center to its outskirts, there is a discrepancy in the center of dark matter only simulations when compared to cosmological systems-from faint dwarf galaxies to massive clusters. Namely the density at the center of these real world objects is smaller, or “cored”-that is its density curve vs. radius is much flatter in a dark matter only simulation than compared with observations.
The authors of this paper hypothesize this mismatch can be accounted for by taking into account the in more difficult to simulate baryonic matter because of its additional collisional and electromagnetic nature, and with relatively weak effects on the cosmological scale for the effort put in. This might not be so surprising to the reader due to my foreshadowing above, but the observational and theoretic community’s opinion as to just how much baryonic matter can influence the behavior of dark matter has been modified in recent years due to better simulation resolution and better observational data. It is now thought to be a much more effective influence and thus a former thorn seems to be resolved. These improvements in simulation and observation are outlined in the paper and summarized here.
Recent spectroscopic observations reveal that there are massive galaxy outflows carrying significant gas mass away from star forming galaxies-in contrast to previous thoughts that star formation impacted passively the DM population. As we see in Box 1 these outflows create non adiabatic corrections. This box illustrates how dark matter and baryons interact through adiabatic contraction-and adiabatic modeling can fail due to outflows. If gas is removed on a short timescale, the dark matter flies outwards due to reduce centripetal force. That means 
Figure 1 shows the observational evidence for these outflows. M82 appears on the left and the right plot shows a compilation of many galaxies, their outflow rate vs. their star formation rate. The takeaway is that even galaxies with low star formation rate can support winds above 100km/s and that these two are correlated. The greater the star formation rate, the faster the wind.
Figure 2 shows the observational discrepancy with theory and Section 3 of the paper outlines this, along with the difficulty of obtaining the observations. The black curve is the simulated result. The plot is the slope of the power law vs. radius, that means that lower alpha is steeper. Thus we can see the observations support shallower than dark matter only simulation results. The right panel shows the probability distribution function of beta (negative alpha), that is the slope of the power law. Here as beta decreases alpha increases, so it is getting shallower. Likewise as beta increases alpha decreases, so the profile gets steeper. Simulation results correspond to 1, so the combined constraints show shallower to dark matter only simulations.
Figure 3 shows the scaling with the process with systems of different mass – there’s a transition between core creation and persistent cusps below a certain mass. Cores would be small for systems below $latex 10^6 M_\odot$. This figure provides the primary support for the author’s thesis, that hydrodynamical simulations are important to create the necessary cores observed. Specifically if there is more than a certain amount of gas in the simulation, the crosses, or results of the hydrodynamical simulations, are in general cored (that is alpha is less than -1).
Finally the authors sketch alternatives to CDM and their hypothesis which can include some modified form of dark matter or “non-minimal” dark matter model. Warm dark matter or self interacting dark matter for example. These models in general could also prevent the central high density cusp. However, neither of these provide a complete alleviation, and no DM model can alone alleviate removing low angular momentum baryons from the center of galaxies. The authors do add a note of caution both models incorporating baryonic physics and alternative models such as WDM/SIDM etc. may have competing, canceling or otherwise conflated effects which complicates these, so including baryonic physics fully will involve reconsidering and re-constraining non-minimal dark matter models as well.