DUST EXTINCTION FROM BALMER DECREMENTS OF STAR-FORMING GALAXIES AT 0.75 ⩽z⩽ 1.5 WITHHUBBLE SPACE TELESCOPE/WIDE-FIELD-CAMERA 3 SPECTROSCOPY FROM THE WFC3 INFRARED SPECTROSCOPIC PARALLEL SURVEY

Abstract

Spectroscopic observations of Hα and Hβ emission lines of 128 star-forming galaxies in the redshift range 0.75 ≤ z ≤ 1.5 are presented. These data were taken with slitless spectroscopy using the G102 and G141 grisms of the Wide-Field-Camera 3 (WFC3) on board the Hubble Space Telescope as part of the WFC3 Infrared Spectroscopic Parallel (WISP) survey. Interstellar dust extinction is measured from stacked spectra that cover the Balmer decrement (Hα/Hβ). We present dust extinction as a function of Hα luminosity (down to 3 × 10 41 erg s −1 ), galaxy stellar mass (reaching 4 × 10 8 M ), and rest-frame Hα equivalent width. The faintest galaxies are two times fainter in Hα luminosity than galaxies previously studied at z ∼ 1.5. An evolution is observed where galaxies of the same Hα luminosity have lower extinction at higher redshifts, whereas no evolution is found within our error bars with stellar mass. The lower Hα luminosity galaxies in our sample are found to be consistent with no dust extinction. We find an anti-correlation of the [O III]λ5007/Hα flux ratio as a function of luminosity where galaxies with L Hα < 5 × 10 41 erg s −1 are brighter in [O III]λ5007 than Hα. This trend is evident even after extinction correction, suggesting that the increased [O III]λ5007/Hα ratio in low luminosity galaxies is likely due to lower metallicity and/or higher ionization parameters.

Predicting Hα emission-line galaxy counts for future galaxy redshift surveys

“…For instance, further calibration of the galaxy metallicities of bright galaxies may boost the optical depths pre- (2000) method, the lack of trend could suggest the need for further parametrisation. We note however that due to the size of uncertainties of their measurements, the WISP optical depths from Domínguez et al (2013) could also be consistent with a flat relation and no change in optical depth with luminosity.…”

“…For luminosities fainter than 10 42.5 erg s −1 all three methods are consistent within error with one another, whilst for the brightest luminosities there is growing disagreement between the Ferrara et al (1999) and Calzetti et al (2000) methods. From their measurements, Domínguez et al (2013) report an increase in the optical depth with increasing observed luminosity. A similar trend is also reported by Hopkins et al (2001) and Sobral et al (2012).…”

“…For GALACTICUS we show the mean and ±1σ dispersions in bins of deblended dust-attenuated Hα luminosity over the redshift range 0.75 z 1.5. For comparison we show the WISP results from Domínguez et al (2013). line wavelength according to,…”

“…In each case we assume the attenuation parameters that minimise χ 2 /ν, as reported in Table 1. For comparison we show the estimates from the WISP analysis of Domínguez et al (2013). Note that attenuation computed by the Calzetti et al (2000) method depends only on wavelength and so the attenuation is identical for each galaxy, leading to a constant mean optical depth with zero scatter.…”

Predicting dust extinction properties of star-forming galaxies from Hα/UV ratio

“…Using our own datasets, we also confirm that the scatter around the best-fit AHα-stellar mass relation is fairly small (0.32 mag): this is actually even smaller than the Hα/UV approach (∼0.38 mag), and is comparable to those achieved with our new recipe (see Table 1). Furthermore, the correlation between dust extinction and stellar mass is reported to be unchanged at least out to z ∼ 1.5 Sobral et al 2012;Ibar et al 2013;Domínguez et al 2013), and so stellar mass can be an useful indicator for a rough estimate of dust extinction of galaxies at all redshifts in the absence of the Balmer decrement or deep FIR photometry (but see Oteo 2014; Pannella et al 2015; Shimakawa et al 2015a). However, as demonstrated in this work, it might be too simplistic to blindly apply the M⋆-dependent dust -7), with the greyscale indicating the average aperture correction value for galaxies at each pixel.…”