Increasing the accuracy of galaxy metallicity measurements over the past 12 Gyr of cosmic history

Ryan Sanders (UCLA)
October 20, 2017 - 12:15pm


The baryonic content of galaxies is thought to grow through the interplay of accretion of gas from the intergalactic medium, processing of that gas through star formation, and feedback processes such as supernovae that heat gas and drive outflows, known collectively as the "cycle of baryons."  Measuring the gas-phase metallicities of galaxies across multiple epochs in cosmic history is one of the most promising avenues to gain insight into the cycle of baryons controlling galaxy growth.  Obtaining gas-phase metallicity estimates at any redshift requires the measurement of multiple emission lines from different ionic species and an understanding of the systematic biases associated with measuring emission-line properties from global galaxy spectra.  To study the redshift evolution of scaling relationships such as the stellar mass-gas-phase metallicity relation (MZR) and the mass-SFR-metallicity (M*-SFR-Z) relation, the effects of chemical evolution on line ratios must be disentangled from changes in other parameters, such as electron density, ionization parameter, and ionizing spectrum.  We present results from the MOSDEF survey on the electron density, ionization state, and metallicity of star-forming galaxies at z~2-4, and investigate the evolution of metallicity scaling relations over the past 12 Gyr.  In order to clearly break degeneracies between metallicity and other parameters, measurements of electron temperature-sensitive auroral lines, such as [OIII]4363, at high redshifts are needed to obtain direct-method metallicities without relying on strong-line calibrations.  We present measurements of [OIII]4363 at z>2, and discuss implications for the evolution of metallicity calibrations with redshift.  Future observations from the James Webb Space Telescope will provide more auroral line detections at high redshifts, leading to a clearer picture of the chemical evolution of galaxies than has been possible before.  Systematic biases affecting measurements at z=0 must also be carefully considered.  We use a set of empirically-motivated models to consider systematic biases in properties derived from global galaxy spectra at both low and high redshifts, including emission from diffuse ionized gas (DIG) and flux-weighting effects from multiple line emitting components.  Using these models, we demonstrate the effects that DIG contamination can have on z~0 galaxy metallicity estimates and evolutionary studies.