Shining (the wrong?) light on Self-Interacting Dark Matter

Title: Brightest Cluster Galaxy Offsets in Cold Dark Matter

Authors: Cian Roche, Michael McDonald, Josh Borrow, Mark Vogelsberger, Xuejian Shen, Volker Springel, Lars Hernquist, Ruediger Pakmor, Sownak Bose, Rahul Kannan

First Author’s Institution: Department of Physics and MIT Kavli Institute for Astrophysics and Space Research, Massachusetts Institute of Technology, USA

Status: Published in The Open Journal of Astrophysics [open access]

One of the biggest and most fascinating questions remaining in astrophysics is the question of dark matter (DM) – the mysterious substance that makes up 85% of the matter in the universe but doesn’t seem to interact at all with light. From the moment it was first hypothesized astronomers have been trying to explain what DM is. Observations have ruled out many models of how DM could behave, but many more models still remain possible.

The model that’s currently primarily used is known as Λ Cold Dark Matter (ΛCDM) – DM that can only interact with itself and the rest of the universe via gravity. In today’s paper, the primary alternative to this model that’s being discussed is Self-Interacting Dark Matter (SIDM) – a model where, as the name suggests, DM particles can interact with each other on a given spatial scale. This model shows up as a flattening of the dark matter distribution in the cores of DM halos. Because, under this model, DM particles scatter against (bounce off) each other, they can’t pack as close together as regular ΛCDM can, and so the centers of DM halos are prevented from getting as dense.

Like any idea in science, DM models need to be tested with observations. The issue with dark matter is, well, it’s dark! We can’t measure the position of dark matter directly, so we need to develop indirect probes of dark matter distributions. Dwarf galaxies are particularly useful for testing this, since they tend to have much more DM than normal matter, so the things in the galaxy that emit light (stars, gas) follow the DM much more reliably than in other galaxies. On the other end of the spectrum, massive clusters of galaxies are also a good probe of DM distributions. These are the most massive objects in the universe, so they’re surrounded by the biggest dark matter halos and have the densest dark matter at their cores. Today’s authors focus on one method in particular to measure DM distributions with galaxy clusters – the movement of the Brightest Cluster Galaxy, or BCG.

BCGs are extremely interesting objects on their own (they’re the uniquely bright, massive galaxy right at the center of a galaxy cluster – see some astrobites on the subject), but this strategy ignores all of that. Instead, the BCG is treated as a single solid object, floating around in the gravitational potential set up by the galaxy cluster dark matter halo – a ‘test particle‘. With regular ΛCDM, the BCG should wobble a bit around the center of the gravitational potential of the cluster DM halo, but mostly stay quite close. In a SIDM model, however, the cluster potential will be much flatter, so there won’t be as much of a distinct center and the BCG will wobble much more. This wobble happens on a timescale much too long for us to see in a single object, but if we look at many BCGs, the distribution of their position with respect to the cluster center should be representative of how a typical single BCG moves. This has been shown to be a plausible probe of dark matter behavior in simulations that assume SIDM, and some observations of galaxy clusters have shown BCG offsets significant enough to maybe point towards a SIDM model.

In today’s paper, the authors use several different popular hydrodynamical simulations to test BCG ‘wobble’ as a dark matter probe from the other end – they take simulations that assume a ΛCDM model, and see what BCG-cluster offsets they produce. The simulations they use include IllustrisTNG, MilleniumTNG, and BAHAMAS. For each code, they generate many different simulated clusters, and measure the distance between the center of the cluster DM distribution and the BCG.

The first conclusion of the authors’ simulation work is that simulations using ΛCDM do indeed predict very small offsets between the BCG and the center of the cluster gravitational potential. For each of the simulation codes they ran, the authors found that the median offset between the BCG and the DM center was below the ‘gravitational softening length’ (an effective gravitational resolution) of the simulation – see Figure 1.

Figure 1: The median offsets between the BCG and the dark matter center of galaxy clusters generated with different simulation methods. The error bars show the 68% interval bounds. These are plotted as a function of the ‘softening length’ of the simulation – the length scale below which gravitational interactions are artificially decreased, functionally a gravitational resolution. In each case, the offset is below the softening length and thus is not significant. Adapted from Fig. 2 of Roche et al. 2024.

However, the authors also found that the measured offset distribution depended a lot on how you define the center of the galaxy cluster. The center of the BCG is pretty easy to find, whether in a simulation or an observation, because the BCG is small and bright. The center of the galaxy cluster, however, is more difficult. With simulations, you know exactly where all the matter (dark or otherwise) is in your simulation, so you can find the exact center of the gravitational potential. This is the method the authors used for Figure 1. You can also find the center of the projected DM mass distribution. Again, though, dark matter is dark! When not working with simulations we need to rely on other tracers of the dark matter distribution, which may not trace the real distribution perfectly.

Figure 2: The distribution of BCG-cluster center offsets measured from the simulations using four different ways of measuring the center of the galaxy cluster. The actual dark-matter center is the blue line, and is what was used to create Figure 1. The ‘Lensing’ and ‘Gas’ distributions (described below) are the most like methods used by observers, and show distributions that are much more significant than the actual offset distribution. Fig. 3 of Roche et al. 2024.

The authors test two other methods of finding the dark matter center, both of which more realistically simulate actual observations. These include reconstructing the mass distribution from gravitational lensing of sources near the center of the cluster, and measuring the centroid of the hot intracluster medium gas in the cluster, taking that as a proxy for the DM centroid. Both of these methods overpredict the offsets between the BCGs and the galaxy cluster centroids considerably – the lensing approach by a factor of 10, and the gas approach by a factor of 30. The distributions of their offsets are shown in Figure 2. This is concerning, because these are the methods most often used in observational measurements of the BCG-cluster offset, and this overprediction looks a lot like the SIDM signature. It might be that the tentative evidence we’ve been seeing for an exotic DM model is just light acting up!

Ultimately, the conclusion that the authors come to is that dark matter is a very hard thing to measure. If we’re going to constrain its nature, we need to be absolutely confident that the available observations, the techniques used to analyze them, and the simulations used to test findings are all absolutely trustworthy, and all consistent with each other. In this specific case, and for this specific tracer, we’re probably not there yet. Still, that just means we have a very exciting mystery to solve!

Astrobite edited by Skylar Grayson

Featured image credit: The IllustrisTNG Collaboration (TNG50)

This bite was edited on Sept. 27th, 2024 to more accurately describe the cluster center determination, and to fix the measured offset overpredictions from lensing and gas measurements, which had previously been switched.

About Delaney Dunne

I'm a PhD Candidate at Caltech, where I study how galaxies form and evolve by mapping their molecular gas! I do this using COMAP, a radio-frequency Line Intensity Mapping experiment based in California's Owens Valley.

Discover more from astrobites

Subscribe to get the latest posts to your email.

Leave a Reply

Discover more from astrobites

Subscribe now to keep reading and get access to the full archive.

Continue reading