Spatially Resolved Stellar Kinematics of Field Early-Type Galaxies at z=1: Evolution of the Rotation Rate

TL;DR

This study investigates the stellar kinematics of early-type galaxies at z=1, revealing that the fraction of rotating galaxies remains constant over time and examining the evolution of their mass-to-light ratios.

Contribution

It provides the first detailed analysis of galaxy rotation at z=1 using spatially resolved spectroscopy and models the kinematics with surface brightness data.

Findings

No significant evolution in the fraction of rotating early-type galaxies since z=1.

S0 and Sa galaxies are generally rotationally supported.

M/L ratios from Jeans models and virial estimates agree for non-rotating galaxies, but differ for rotating ones.

Abstract

We use the spatial information of our previously published VLT/FORS2 absorption line spectroscopy to measure mean stellar velocity and velocity dispersion profiles of 25 field early-type galaxies at a median redshift z=0.97 (full range 0.6<z<1.2). This provides the first detailed study of early-type galaxy rotation at these redshifts. From surface brightness profiles from HST imaging we calculate two-integral oblate axisymmetric Jeans equation models for the observed kinematics. Fits to the data yield for each galaxy the degree of rotational support and the mass-to-light ratio M/L_Jeans. S0 and Sa galaxies are generally rotationally supported, whereas elliptical galaxies rotate less rapidly or not at all. Down to M(B)=-19.5 (corrected for luminosity evolution), we find no evidence for evolution in the fraction of rotating early-type (E+S0) galaxies between z=1 (63+/-11%) and the present (61+/-5%). We interpret this as evidence for little or no change in the field S0 fraction with redshift. We compare M/L_Jeans with M/L_vir inferred from the virial theorem and globally averaged quantities and assuming homologous evolution. There is good agreement for non-rotating (mostly E) galaxies. However, for rotationally supported galaxies (mostly S0) M/L_Jeans is on average ~40% higher than M/L_vir. We discuss possible explanations and the implications for the evolution of M/L between z=1 and the present and its dependence on mass.