Simulating the Local Universe
Dr. Klaus Dolag
Local Project ID:
HPC Platform used:
SuperMUC-NG at LRZ
Our Cosmic Home, which is the local volume of the Universe centered on us, contains very prominently visible structures, extending over almost one billion light-years. Such structures, ranging from the Local Group over the Local Void and the most prominent galaxy clusters like Virgo, Perseus, Coma and many more, represent a formidable site where extremely detailed observations exist. Therefore, cosmological simulations of the formation of galaxies and galaxy clusters within the Local Universe, rather than any other, randomly selected part of the cosmic web, are perfect tools to test our formation and evolution theories of galaxies and galaxy clusters down to the details. However, at these detailed levels, such simulations are facing various challenges, from the computational point as well as from the treatment of various physical processes needed to properly capture the evolution of galaxies and galaxy clusters. Doing this within constrained simulations on the one hand allows a deeper understanding of the evolution of the non-thermal components and thereby improves our understanding of cosmic ray physics and on the other hand sheds light on galaxy formation processes. Including magnetic fields allows us to additionally address questions about the propagation of ultra high energy cosmic rays (UHECRs) and other messenger particles. Figure 1 shows a full sky map of the thermal Sunyaev Zeldovich effect, e.g. the shadow of the prominent local structures caused onto the cosmic microwave background, as predicted by the first, hydro-dynamical simulations performed within this project.
Validating the reproduction of local clusters
As the observations put constraints in different quality across the simulation volume, it is very important to understand in detail what are the corresponding structures within the simulations and the observations. Especially cross checking the results against independent observations delivers quite a good understanding of the level of matching between simulated and observed structures. We started to do a detailed comparison based on our large set of dark matter simulations, as well as on our first hydro-dynamical simulations.
Figure 2a shows a comparison of the mass of clusters, inferred from different observations, namely dynamical mass (from galaxy position and velocities), Sunyaev Zeldovich effect (the shadow of the ICM on the CMB) and X-ray Temperature measurements. Besides the large uncertainties in the observations reflected in the different masses obtained from the different (and independent) observations, a very clear trend is visible that the constrained simulations re-produce the cluster masses on individual bases for many of the systems quite well. The next steps will be to use directly the observational properties (like SZ and X-Ray signal) from the first hydro-dynamical counterparts.
The additional advantage of the simulations is that this now also allows to investigate the evolution of these objects within their (constrained) environment, as for example discussed for the Virgo cluster (see Sorce+ 2021). Here we can now extend this to many galaxy clusters in the local universe, as shown in figure 2b, which displays the evolution of the mass of some of the simulated, local clusters (Hernandez+ 2022, in preparation). Strikingly, one can immediately see that Coma has a recent major merger (as expected from observations), while Perseus has a very quiet recent accretion history, as expected by being observed as a cool-core cluster. Next step will be to more quantitatively derive the dynamical state of the simulated clusters directly from observables (mainly SZ and X-Ray) and directly compare them to the ones derived from the real clusters. We will also investigate, how well merger driven, non-thermal emissions like the radio halo of the Coma cluster can be reproduced within the MHD/CR versions of the simulations.
Galaxy cluster dynamics from velocity waves
Galaxy clusters are excellent cosmological probes provided their mass estimates are accurately determined. Fueled with large imaging surveys, stacked weak lensing is the most promising method though it provides mass estimates within relatively small radii. Given the large amount of accompanying redshift and spectroscopic data overlapping the imaging surveys, we must take the opportunity to calibrate also with a reasonable accuracy a method based on galaxy dynamics. Two independent measures hold indeed better constraints on the cosmological model. Infall zones of galaxy clusters are probably the less sensitive to baryonic physics thus systematics and probe large radii. These manifestations of a tug of war between gravity and dark energy provide a unique avenue to test modified gravity theories when comparing resulting mass estimates to those from stacked weak lensing measurements. Combined with stacked weak lensing results, they might even yield evidence that departure from General Relativity on cosmological scales is responsible for the expansion acceleration.
The accurate calibration of the relation between infall zones properties and cluster masses starts with careful comparisons between cosmological simulations and observations. Our cosmological simulations of the local Large Scale Structure have sufficient resolution to study the effect of the gravitational potential of massive local halos onto the velocity of (sub-)halos and compare with that of their observational cluster counterparts, as shown in figure 3. Clear to see that velocity waves stand out in radial peculiar velocity – distance to a box-centered synthetic observer diagram. The agreement between velocity waves, caused by the most massive halos of the simulations and those born from their observational local cluster counterparts, is simply remarkable especially for the clusters the closest to us that are the best constrained (e.g. Virgo, Centaurus). Secondary waves due to smaller groups in (quasi) the same line-of-sight as the most massive clusters stand out equally.
Currently, several simulations are running, among them a simulation including galaxy formation physics like star-formation and AGN feedback, full MHD simulations including in addition the treatment of cosmic rays.
The result of these simulations will be in addition extensively used within LOCALIZATION , a newly formed, collaborative project that is jointly funded by ANR/DFG and started in December 2021. Here especially these and future constrained simulations will be used to investigate the so-called σ8 tension (e.g. the difference between the linearly evolved amplitude of the matter power spectrum measured with the CMB and galaxy clusters), as well as CMB large-scale anomalous features.
 Sorce, J. Mohayaee, R. , Aghanim, N., Dolag, K. & Malavasi, N., 2022, Submitted to A&A