LBT and Spitzer Spectroscopy of Star-Forming Galaxies at 1 < z < 3: Extinction and Star Formation Rate Indicators

TL;DR

This study uses spectroscopic data from gravitationally lensed star-forming galaxies at redshifts 1 to 3 to analyze dust extinction and evaluate star formation rate indicators, revealing significant extinction effects and variability in dust distribution.

Contribution

It provides new measurements of dust extinction and tests multiple SFR indicators in high-redshift galaxies, highlighting the importance of accounting for dust in SFR estimates.

Findings

Extinction values range from ~0 to 5.9 mag, larger than previously known.

H-alpha SFR underestimates true SFR without dust correction.

6.2 micron PAH luminosity correlates with SFR but with scatter.

Abstract

We present spectroscopic observations in the rest-frame optical and near- to mid-infrared wavelengths of four gravitationally lensed infrared (IR) luminous star-forming galaxies at redshift 1 < z < 3 from the LUCIFER instrument on the Large Binocular Telescope and the Infrared Spectrograph on Spitzer. The sample was selected to represent pure, actively star-forming systems, absent of active galactic nuclei. The large lensing magnifications result in high signal-to-noise spectra that can probe faint IR recombination lines, including Pa-alpha and Br-alpha at high redshifts. The sample was augmented by three lensed galaxies with similar suites of unpublished data and observations from the literature, resulting in the final sample of seven galaxies. We use the IR recombination lines in conjunction with H-alpha observations to probe the extinction, Av, of these systems, as well as testing star formation rate (SFR) indicators against the SFR measured by fitting spectral energy distributions to far-IR photometry. Our galaxies occupy a range of Av from ~0 to 5.9 mag, larger than previously known for a similar range of IR luminosities at these redshifts. Thus, estimates of SFR even at z ~ 2 must take careful count of extinction in the most IR luminous galaxies. We also measure extinction by comparing SFR estimates from optical emission lines with those from far-IR measurements. The comparison of results from these two independent methods indicates a large variety of dust distribution scenarios at 1 < z < 3. Without correcting for dust extinction, the H-alpha SFR indicator underestimates the SFR; the size of the necessary correction depends on the IR luminosity and dust distribution scenario. Individual SFR estimates based on the 6.2 micron PAH emission line luminosity do not show a systematic discrepancy with extinction, although a considerable, ~0.2 dex scatter is observed.