MID-IR LUMINOSITIES AND UV/OPTICAL STAR FORMATION RATES ATz< 1.4

Abstract

Ultraviolet non-ionizing continuum and mid-IR emission constitute the basis of two widely used star formation indicators at intermediate and high redshifts. We study 2430 galaxies with z < 1.4 in the Extended Groth Strip with deep MIPS 24 µm observations from FIDEL, spectroscopy from DEEP2, and UV, optical, and near-IR photometry from AEGIS. The data are coupled with dust-reddened stellar population models and Bayesian SED fitting to estimate dust-corrected SFRs. In order to probe the dust heating from stellar populations of various ages, the derived SFRs were averaged over various timescales-from 100 Myr for "current" SFR (corresponding to young stars) to 1-3 Gyr for long-timescale SFRs (corresponding to the light-weighted age of the dominant stellar populations). These SED-based UV/optical SFRs are compared to total infrared luminosities extrapolated from 24 µm observations, corresponding to 10-18 µm rest frame. The total IR luminosities are in the range of normal star forming galaxies and LIRGs (10 10 -10 12 L ⊙ ). We show that the IR luminosity can be estimated from the UV and optical photometry to within a factor of two, implying that most z < 1.4 galaxies are not optically thick. We find that for the blue, actively star forming galaxies the correlation between the IR luminosity and the UV/optical SFR shows a decrease in scatter when going from shorter to longer SFR-averaging timescales. We interpret this as the greater role of intermediate age stellar populations in heating the dust than what is typically assumed. Equivalently, we observe that the IR luminosity is better correlated with dust-corrected optical luminosity than with dust-corrected UV light. We find that this holds over the entire redshift range. Many so-called green valley galaxies are simply dust-obscured actively star-forming galaxies. However, there exist 24 µm-detected galaxies, some with L IR > 10 11 L ⊙ , yet with little current star formation. For them a reasonable amount of dust absorption of stellar light (but presumably higher than in nearby early-type galaxies) is sufficient to produce the observed levels of IR, which includes a large contribution from intermediate and old stellar populations. In our sample, which contains very few ULIRGs, optical and X-ray AGNs do not contribute on average more than ∼ 50% to the mid-IR luminosity, and we see no evidence for a large population of "IR excess" galaxies.

CONFIRMING THE EXISTENCE OF A QUIESCENT GALAXY POPULATION OUT TO z = 3: A STACKING ANALYSIS OF MID-, FAR-INFRARED, AND RADIO DATA

“…longer than 160 μm (Bendo et al 2012; see also Bell 2003;Salim et al 2009;Fumagalli et al 2014), which are most relevant for the Herschel SPIRE bands. We note in Figure 3 that the SFR 24 /SFR H ratio is less discrepant for massive IR-faint QGs candidates than less massive ones, which contradicts the expectations if old stellar populations are primarily responsible for the FIR emission.…”

Galex–sdss–wise Legacy Catalog (Gswlc): Star Formation Rates, Stellar Masses, and Dust Attenuations of 700,000 Low-Redshift Galaxies

“…These are nearly quiescent galaxies of LINER type (thus, potentially passive galaxies with non-AGN emission lines; Stasińska et al 2008). For such galaxies the IR emission, if detected, will be dominated by dust heating by relatively dusty old or intermediate-age (e.g., AGB) stellar populations (Bressan et al 2001;Villaume et al 2015) and will therefore not be indicative of the current SF (Cortese et al 2008;Salim et al 2009). When the IR luminosity of such galaxies is converted into SFRs using simple, fixed-coefficient formulae (Equation (8)), which assume that dust is heated by young stars, the SFR will be overestimated (Buat & Xu 1996;Kennicutt 1998;Boquien et al 2016).…”

Encoding of the infrared excess in theNUVrKcolor diagram for star-forming galaxies

“…Both measurements suffer from independent source of uncertainties, thus the bias in the IR luminosity may be related to either methods. However, the small scatter (less than a factor 2) and bias show that the SED fitting is a robust and reliable method to estimate the total IR luminosity, when accurate observations on a wide enough wavelength are available (e.g., Salim et al 2009). This is indeed the case for the COSMOS dataset used here (i.e., 31 passbands from Far-UV to Mid-IR: 0.15 ≤ λ ≤ 4.5 µm).…”