Signatures of dynamical activity in the hot gas profiles of groups and clusters in the FLAMINGO simulations

TL;DR

Using the FLAMINGO suite, the paper investigates dynamical activity signatures in the hot gas of groups and clusters up to $z=1$ by comparing a theoretical halo mass accretion rate $Γ$ with observational proxies $ΔM^*_{14}$, $c_x$, and $⟨w⟩$. It finds $⟨w⟩$ to be the most robust dynamical-state tracer across mass/redshift, while $c_x$ and $ΔM^*_{14}$ are more sensitive to resolution and feedback; the dark-matter density–$Γ$ relation shows a consistent radial sign change, mirrored by the hot gas when using $⟨w⟩$. Gas-density proxies correlate with $Γ$ in a radially dependent manner, strongest in clusters and at low redshift, with correlations weakening in groups and under stronger feedback. The intrinsic scatter of gas profiles declines with redshift, and the radius of minimum scatter increases with feedback strength, suggesting a potential diagnostic for feedback in upcoming X-ray/SZ surveys. These results inform how to model dynamical-state selection and interpret high-redshift, lower-mass systems in future cosmological investigations.

Abstract

In anticipation of upcoming cosmological surveys, we use the large volume Flamingo hydrodynamical simulations to look for signatures of dynamical activity, focusing on the hot gas profiles of groups and clusters out to redshift $z=1$. To determine the dynamical state of each object, we consider the halo mass accretion rate, $Γ$, as well as three observational proxies: stellar mass gap, $\Mstar$; X-ray concentration, $c_\mathrm{x}$, and X-ray centroid shift, $\left<w\right>$. In general, the median values of these indicators vary in accordance with an increase in dynamical activity with both mass and redshift. We find $\left<w\right>$ to be the most reliable proxy, while $c_\mathrm{x}$ and $\Mstar$ are more sensitive to resolution and feedback model details. Looking at the profiles, the correlation between dark matter density and $Γ$ has a characteristic radial dependence, being negatively (positively) correlated at small (large) radii. This trend is insensitive to both halo mass and redshift. Similar behaviour is also seen for the hot gas densities in low redshift clusters, particularly when using $\left<w\right>$, but the correlations become weaker in groups, at higher redshift and when stronger feedback is employed. We also find the intrinsic scatter in the gas density profiles to decrease with redshift, particularly in groups, contrary to what is seen for the dark matter. Interestingly, the radius of minimum gas density scatter increases with feedback strength, suggesting that this property could be a useful feedback diagnostic in future observational studies.

Figures (12)

• Figure 1: Corner plots for the L1_m8 groups: $\log_{10}(\mathrm{M}_\mathrm{500c}/\mathrm{M_{\odot}}) =13.5 - 14$ at $z=0$ (top) and L2p8_m9 clusters: $\log_{10}(\mathrm{M}_\mathrm{500c}/\mathrm{M_{\odot}}) =14.5 - 15$ at $z=0$ (bottom), comparing the accretion rate, stellar mass gap, X-ray concentration and X-ray centroid shift. In the upper right corners, we present a correlation matrix with the Pearson coefficients for each indicator pair. The dashed vertical and horizontal lines represent the relaxed and disturbed limits for each indicator. Note that the $\Gamma$ limits are calculated for each mass and redshift subset.
• Figure 2: Star maps for a relaxed (top) and disturbed (bottom) cluster of similar mass, ($M_\mathrm{500c} \sim 5 \times 10^{14} \mathrm{M}_{\odot}$) at $z=0$ in the Fiducial L1_m9 run. Highlighted in dashed white lines are the soft band ($0.5 - 2$ keV) X-ray luminosity contours for $\log_{10}(L_\mathrm{x}/L_\mathrm{x,max}) = [-2.5, -1.5, -1.0]$. Both $R_\mathrm{500c}$ and $0.15R_\mathrm{500c}$ apertures are denoted as yellow circles with centroids for the 8 apertures between these radii being marked with cyan crosses. The BCG and 4th brightest object are shown as blue and white 50kpc apertures. The following values are dynamical indicator measurements for each cluster: Merging ($\Gamma = 4.0$, $\mathrm{\Delta M^*_{14}} = 0.09$, $c_\mathrm{x}=0.10$, $\log_{10}\left<w\right> = -1.4$) and Relaxed ($\Gamma = 1.5$, $\mathrm{\Delta M^*_{14}} = 3.2$, $c_\mathrm{x}=0.29$, $\log_{10}\left<w\right>= -2.4$).
• Figure 3: Top: X-ray centroid shift vs X-ray concentration contour plots for clusters at $z=0$ with $2.5\times10^{14} < \mathrm{M}_\mathrm{500c}/\mathrm{M_{\odot}} < 10^{15}$ in the L1_m8 and L2p8_m9 simulations compared to the sample of 61 clusters studied in chex_2022 in this same mass regime with $z<0.2$. Bottom: Same as top but for the Jet, $\mathrm{fgas}-8\sigma$ and $\mathrm{fgas}+2\sigma$ runs. The dashed lines denote the m8 (orange) and m9 (blue) relaxed and disturbed dynamical limits for the centroid shift and concentration.
• Figure 4: Median dynamical state indicator vs redshift for the different mass bins and resolutions for the Fiducial model as listed in Table \ref{['tab:sim_params']}. The error bars included are the bootstrap standard error of the median.
• Figure 5: Medians of each dynamical state indicator for the $\log_{10}(M_\mathrm{500c}/\mathrm{M_\odot})=14-14.5$ mass bin for varying feedback runs alongside the Fiducial L1_m9 and L1_m8 cases.