A multi-wavelength look at dust attenuation in star forming galaxies

Astronomers can estimate how many stars form per unit time using the X-ray, ultraviolet, infrared and radio continua, or a variety of Hydrogen recombination lines or heavier-atomic forbidden lines. However, there are known discrepancies between these different star formation rate indicators. It is highly likely that the derived star formation rates may disagree due to factors such as dust attenuation within a galaxy and its variable impact with wavelength, old stars dominating the integrated light across a galaxy, an incorrect Initial Mass Function, metallicity, elemental abundances and the ionization state of the gas. It is possible to quantify the impact of these effects (specifically the impact of dust attenuation) by comparing Halpha fluxes with the 24mum data and the ultraviolet continuum. Astronomers have devised 'robust' star formation rate measures via infrared + ultraviolet data or 24mum + Halpha data or just ultraviolet data. The results presented here are from the Wyoming Survey for Halpha (WySH), a comprehensive four-square degree survey to probe the evolution of star-forming galaxies over 0.16 ≲z≲ 0.40 and the NEWFIRM Narrowband Halpha Survey (NNHS), a similar one-square degree survey designed to probe the evolution of star-forming galaxies over 0.81 ≲z≲ 2.2. We have supplemented the Halpha data from WySH and NNHS with infrared data from the Spitzer Wide-area Infrared Extragalactic (SWIRE) Survey and the Cosmic Evolution Survey (COSMOS) and ultraviolet data from the Galaxy Evolution Explorer (GALEX) Deep Imaging Survey. This dataset provides a multi-wavelength look at the evolution of the dust attenuation, and here we compare a traditional measure of dust attenuation (L(TIR)/L(FUV)) to a diagnostic based on a recently-developed robust star formation rate (SFR) indicator, [Halphaobs + 24mum]/Halphaobs. With such data over multiple epochs, the evolution in the dust attenuation with redshift can be assessed. We present results from the ELAIS-N1 and Lockman Hole regions at z ∼ 0.16, 0.24, 0.32 and 0.40 and the COSMOS and SXDS regions at z ∼ 0.81. While the ensemble averages of both diagnostics are relatively constant from epoch to epoch, each epoch individually exhibits a larger attenuation by dust for higher star formation rates. Hence, an epoch to epoch comparison at a fixed star formation rate suggests a mild decrease in dust attenuation with redshift.