X-ray shocks in the cool cores of galaxy clusters: insights from TNG-Cluster

TL;DR

This work presents the first systematic study of AGN-driven shocks in a large sample of galaxy clusters from the TNG-Cluster cosmological zoom-in simulations. By generating 600 ks mock Chandra observations for 100 mass-matched halos and applying standard observational shock-detection techniques, the authors identify 50 shocks in 30 clusters and quantify their Mach numbers, energetics, and spatial relation to X-ray cavities. They find that shocks are common in cool-core systems, have weak Mach numbers around $M \,\sim\,1.1$, and, together with cavities, contribute energetically to offset cooling within the cooling radius, with shocks typically at larger radii than cavities. When compared to observed clusters, the simulated shocks reproduce the key demographics and energetics, though extreme high-power shocks are more prevalent in real data, likely due to selection biases; overall, the results support a multi-channel AGN feedback scenario in which shocks and cavities heat the ICM across radii.

Abstract

Shock fronts driven by active galactic nuclei in galaxy cluster cores represent a promising mechanism to heat the intracluster gas by converting kinetic energy into thermal energy through gas compression, thereby offsetting radiative cooling. Despite their potential importance, such shocks are challenging to detect, requiring deep X-ray exposures, and have only been identified in ten clusters. We present the first systematic detection and characterization of AGN-driven shocks in simulated clusters from the TNG-Cluster magnetohydrodynamic cosmological zoom-in simulations of galaxies. TNG-Cluster exhibits a rich variety of X-ray structures, including realistic populations of X-ray cavities, as well as shocks, produced by its AGN feedback model, without collimated, relativistic jets, nor cosmic rays. We produce mock Chandra observations with deep exposure times, for a sample of 100 clusters, mass-matched (M$_{500c}=1.2$ - $8.5 \times 10^{14}$ M$_\odot$) to the ten observed clusters with shocks. Using observational techniques, we identify shocks through surface brightness edges fitted with broken power laws and associated density and temperature jumps. We detect 50 shocks in 30 of the 100 clusters, with ~35% hosting multiple shocks. These shocks lie within a hundred kiloparsec of the central SMBH, are weak (Mach number < 2, median ~ 1.1), and are associated with cavities in about half of the cases. Both in observations and in TNG-Cluster, shocks tend to be located at larger radii than cavities, with median offsets of 46 and 27 kpc, respectively. The observationally inferred shock powers are comparable to the cluster cooling luminosities (10$^{44-46}$ erg s$^{-1}$), suggesting that shocks in the simulation are crucial heating mechanisms. Our results indicate that shocks play a role as important as cavities in balancing cooling in cluster cores, acting isotropically and up to larger distances.

Figures (11)

• Figure 1: TNG-Cluster mass-selected sample of 100 simulated clusters from TNG-Cluster (blue circles) randomly chosen to broadly match the M$_\text{500c}$ of the 10 known observed clusters hosting shocks (Table \ref{['fig:sample_selection_panel']}, rectangle markers). Each observational point is color-coded by its total Chandra exposure time (left) and redshift (right). Although the TNG-Cluster subsample was selected based solely on mass, it exhibits soft X-ray luminosities broadly consistent with observations -- except for a small subset with slightly higher L$_\text{x500c}$ -- and is representative of the full TNG-Cluster population within this mass range.
• Figure 2: Gallery of simulated galaxy clusters from TNG-Cluster exhibiting X-ray shocks. Mock Chandra surface brightness images of the central $200\times200$ kpc region of six selected clusters in TNG-Cluster, alongside images processed with unsharp masking or Gaussian gradient magnitude filters to highlight the shock fronts. Images are centered on the AGN position. White dashed arrows mark identified shock fronts, while the green arrow points the specific shock for which we show, in Fig. \ref{['fig:panel_shocks_fit_temp_jump']}, the surface brightness profile and associated temperature jump. The gallery illustrates the diversity of AGN-driven shocks in TNG-Cluster cores, including features located at cavity edges or at larger radii, and exhibiting a range of morphologies, from arc-like fronts with varying opening angles to complete ellipses.
• Figure 3: Profiles across selected shock fronts from the TNG-Cluster simulations (blue arrows in the figure above). Left: Surface brightness profiles at the shock locations, with the quoted Mach numbers inferred from the best-fit broken power law. Right: Temperature profiles across the same shock fronts, showing the pre-shock (colder) and post-shock (hotter) temperature jumps.
• Figure 4: Demographics of clusters with and without identified shocks and X-ray cavities in the TNG-Cluster subsample of 100 clusters (at snapshot 99). Left: Distribution of the number of shocks (mint green) or X-ray cavities (gray) per cluster, showing the percentage of clusters with none, one, two, three, or four detected features. Right: Among the 45 clusters with at least one identified shock or cavity, the fraction hosting only shocks, only cavities, or both. Shocks are detected in 30 out of the 100 clusters in our subsample, indicating that they are reasonably common features in simulated clusters. While shocks frequently appear in clusters that also host X-ray cavities, about 18% of the detected shocks are found in clusters where no cavities were previously identified in Prunier2025a.
• Figure 5: Shock properties in the TNG-Cluster subsample. Top Left: Distribution of Mach numbers for the 50 shocks detected in the TNG-Cluster subsample (filled histogram) and in the observational sample (line histogram, values compiled from the literature, see Table \ref{['tab:shocks']}). The simulated and observed distributions align, with comparable medians. In both cases, shocks are weak (Mach $< 2$), with 16th-84th percentile Mach 1.05–1.18, and three exceeding Mach 1.4. Top Right: Mach number as a function of distance from the central SMBH. To emphasize the (weak) trend of decreasing Mach number with increasing distance, only the bulk of the distribution is shown (higher $>$ 1.25 Mach shocks are represented with clipped triangle markers). Bottom panels, left to right: Distributions of shock (thermo)dynamical properties: radial distance of the shock from the SMBH, estimated shock age, pressure, and temperature jump across the shock front (the latter two measured from deprojected spectral fitting).