Stellar Populations in Satellite Galaxies in Close Pairs

TL;DR

This study tests whether satellite galaxies in close pairs exhibit conformity in stellar population properties by reanalyzing a high-S/N SDSS sample with new semi-empirical SSP templates that span a wide range of [$\alpha$/Fe]. By fitting 15 optical Lick indices to stacks of satellites at fixed velocity dispersion $\sigma$ and comparing two extreme mass ratios $\mu_1$ and $\mu_3$, the authors find robust age conformity: satellites around more massive primaries tend to be older by up to ~2 Gyr at the same $\sigma$, while metallicity and [$\alpha$/Fe] show no clear environmental offsets. This age conformity persists even when allowing [$\alpha$/Fe] to vary, and there is a suggested upturn in $[$\alpha$/Fe] at the lowest $\sigma$. A turnover in the age trends at the highest $\sigma$ implies environment-related processes in the most massive groups, offering a sensitive constraint for subgrid physics in simulations and for assembly-bias scenarios in galaxy evolution.

Abstract

Satellite galaxies that are near to massive primary galaxies in close pairs can have stellar population ages that are more similar to their primaries than expected. This is one way in which close pairs of galaxies show galactic conformity, which is thought to be driven by assembly bias. Such conformity is seen in ages, morphologies and star formation rates in different samples. This paper revisits a high signal-to-noise SDSS spectroscopic sample, by spectral fitting of new stellar population models, to investigate satellite galaxy properties of age, metallicity and alpha-element abundance. We find the clear signature of age conformity, as previously seen, but no clear evidence for conformity in metallicity or abundance ratios. The offsets showing age conformity are not caused by age-metallicity degeneracies. There is a suggestion in these data that lower velocity dispersion satellites have increased [alpha/Fe] compared to a control sample of passive galaxies, however this needs further observations to be verified. Our results also suggest an intriguing turnover in the age trends of the satellites at the highest velocity dispersion, perhaps reflecting the onset of environment-related processes in the most massive groups.

Figures (6)

• Figure 1: We show some details of the stacks corresponding to the second velocity dispersion bin ($\sigma_2$). The top panel shows the ratio between the close pairs stack ($\mu_1$ in orange and $\mu_3$ in purple) and the corresponding control stack. The middle panel shows the S/N as a function of wavelength, and the bottom panel corresponds to the effective spectral resolution. The characteristic transition in resolution from the blue to the red arm of the SDSS spectrograph is evident Smee:13.
• Figure 2: Equivalent widths of the stacks shown as a function of velocity dispersion. From top to bottom we show the 4000Å break strength, D$_n$(4000), Balmer absorption (H$\delta_A$), and the [MgFe]$^\prime$ index. Note the turnover of the age sensitive indices at higher velocity dispersion for the $\mu_3$ stacks. For a consistent comparison, the spectra have been corrected for emission lines and smoothed to a common velocity dispersion of 250 km/s (see text for details). Error bars are propagated from the error in the mean of the stacks, shown at the 1 $\sigma$ level.
• Figure 3: SSP fits to SDSS stacked satellite galaxy data (from Ferreras19a). Light-weighted ages (top panel), metallicities Z=[M/H]$_{SSP}$ (middle panel) and [$\alpha$/Fe] ratios (lower panel) are given, based on fitting 15 Lick line indices, plotted against stellar velocity dispersion in the middle of each bin ($\sigma_0$). Orange points are for satellites with masses just below their nearby primary galaxies ($\mu_1$); purple points are for satellites with much lower masses than their primaries ($\mu_3$); black points are for the control sample without environment constraints; small grey points are from the sample of passive, early-type galaxies (ETG) described in Section \ref{['sec:Full']}. Uncertainty ranges are obtained from our Monte-Carlo perturbations of the spectra being fitted (see Section \ref{['sec:SSPfitting']}).
• Figure 4: This figure shows normalised $\chi^2_{\nu}$ contours for SSP fits to the satellite galaxy stacks at different stellar velocity dispersions, indicated by the five different colours. Contours are plotted at 1-sigma confidence level for three fitted parameters, fitting 15 Lick indices. Here $\sigma_1$ to $\sigma_5$ are the lowest to highest velocity dispersion bins as listed in Table \ref{['SSPfits_Table']}. At each $\sigma$, satellite stacks with low $M_{SAT}/M_{PRI}$ ($\mu_3$) are shown with a darker contour line and satellite stacks with higher $M_{SAT}/M_{PRI}$ ($\mu_1$) are shown with a lighter contour line. On the left are plotted contours in Age and metallicity ($=Z=$[M/H]$_{SSP}$) space, at the best fitting [$\alpha$/Fe]. On the right are plotted contours in Age and [$\alpha$/Fe], at the best fitting metallicity.
• Figure 5: Same as in Figure \ref{['SSP_Contours']} but for SSP fits marginalised over the third parameter not plotted in each case.