2010ApJ...716..556T

We present Swift UVOT ultraviolet (UV; 1600-3000 Å) data with complete three-band UV photometry for a sample of 41 galaxies in 11 nearby (<4500 km/s) Hickson Compact Groups (HCGs) of galaxies. We use UVOT uvw2-band (2000 Å) photometry to estimate the dust-unobscured component, SFR_UV, of the total star formation rate, SFR_TOTAL. We use Spitzer MIPS 24 µm photometry to estimate SFR_IR, the component of SFR_TOTAL that suffers dust extinction in the UV and is re-emitted in the IR. By combining the two components, we obtain SFR_TOTAL estimates for all HCG galaxies. We obtain total stellar mass, M_*, estimates by means of Two Micron All Sky Survey K_s-band luminosities, and use them to calculate specific star formation rates, SSFR ≡ SFR_TOTAL/M _*. SSFR values show a clear and significant bimodality, with a gap between low (≲3.2x10^–11/yr) and high-SSFR (≳1.2x10^–10/yr) systems. We compare this bimodality to the previously discovered bimodality in α_IRAC, the MIR activity index from a power-law fit to the Spitzer IRAC 4.5-8 µm data for these galaxies. We find that all galaxies with α_IRAC≤ 0 ( >0) are in the high- (low-) SSFR locus, as expected if high levels of star-forming activity power MIR emission from polycyclic aromatic hydrocarbon molecules and a hot dust continuum. Consistent with this finding, all elliptical/S0 galaxies are in the low-SSFR locus, while 22 out of 24 spirals/irregulars are in the high-SSFR locus, with two borderline cases. We further divide our sample into three subsamples (I, II, and III) according to decreasing H I richness of the parent galaxy group to which a galaxy belongs. Consistent with the SSFR and α_IRACbimodality, 12 out of 15 type I (11 out of 12 type III) galaxies are in the high- (low-) SSFR locus, while type II galaxies span almost the full range of SSFR values. We use the Spitzer Infrared Nearby Galaxy Survey (SINGS) to construct a comparison subsample of galaxies that (1) match HCG galaxies in J-band total galaxy luminosity and (2) are not strongly interacting and largely isolated. This selection eliminates mostly low-luminosity dwarfs and galaxies with some degree of peculiarity, providing a substantially improved, quiescent control sample. Unlike HCG galaxies, galaxies in the comparison SINGS subsample are continuously distributed both in SSFR and α_IRAC, although they show ranges in SFR_TOTALvalues, morphologies and stellar masses similar to those for HCG systems. We test the SSFR bimodality against a number of uncertainties, and find that these can only lead to its further enhancement. Excluding galaxies belonging to HCGs with three giant galaxies (triplets) leaves both the SSFR and the α_IRAC bimodality completely unaffected. We interpret these results as further evidence that an environment characterized by high galaxy number densities and low galaxy velocity dispersions, such as the one found in compact groups, plays a key role in accelerating galaxy evolution by enhancing star formation processes in galaxies and favoring a fast transition to quiescence.