2006ApJ...646..834X

We study two samples of local galaxies, one UV (GALEX) selected and the other FIR (IRAS) selected, to address the question of whether UV and FIR surveys see two sides (``bright'' and ``dark'') of the star formation of the same population of galaxies or two different populations of star-forming galaxies. No significant difference between the L_tot(=L₆₀+L_FUV) luminosity functions of the UV and FIR samples is found. In addition, after the correction for the ``Malmquist bias'' (bias for flux-limited samples), the FIR-to-UV ratio versus L<sub>tot</sub> relations of the two samples are consistent with each other. In the range of 9≲log(L<sub>tot</sub>/L<sub>☉</sub>)≲12, both can be approximated by a simple linear relation of log(L<sub>60</sub>/L<sub>FUV</sub>)=log(L<sub>tot</sub>/L<sub>☉</sub>)-9.66. These are consistent with the hypothesis that the two samples represent the same population of star-forming galaxies, and their well-documented differences in L<sub>tot</sub> and in FIR-to-UV ratio are due only to the selection effect. A comparison between the UV luminosity functions shows marginal evidence for a population of faint UV galaxies missing in the FIR-selected sample. The contribution from these ``FIR-quiet'' galaxies to the overall UV population is insignificant, given that the K-band luminosity functions (i.e., the stellar mass functions) of the two samples do not show any significant difference.