Identification of z ≳ 2 Herschel 500 μM Sources Using Color Deconfusion

We present a new method to search for candidate z ≳ 2 Herschel 500 μm sources in the Great Observatories Origins Deep Survey-North field using a S500 μm/S24 μm “color deconfusion” technique. Potential high-z sources are selected against low-redshift ones from their large 500 to 24 μm flux density ratios. By effectively reducing the contribution from low-redshift populations to the observed 500 μm emission, we are able to identify counterparts to high-z 500 μm sources whose 24 μm fluxes are relatively faint. The recovery of known z ≳ 4 starbursts confirms the efficiency of this approach in selecting high-z Herschel sources. The resulting sample consists of 34 dusty star-forming galaxies at z ≳ 2. The inferred infrared luminosities are in the range 1.5 × 1012-1.8 × 1013 L⊙, corresponding to dust-obscured star formation rates (SFRs) of ˜260-3100 M⊙ yr-1 for a Salpeter initial mass function. Comparison with previous SCUBA 850 μ {{m}}-selected galaxy samples shows that our method is more efficient at selecting high-z dusty galaxies, with a median redshift of z=3.07+/- 0.83 and with 10 of the sources at z ≳ 4. We find that at a fixed luminosity, the dust temperature is ˜5 K cooler than that expected from the {T}d-{L}{{IR}} relation at z\quad ≲ 1, though different temperature selection effects should be taken into account. The radio-detected subsample (excluding three strong active galactic nucleus) follows the far-infrared (far-IR)/radio correlation at lower redshifts, and no evolution with redshift is observed out to z˜ 5, suggesting that the far-IR emission is star formation dominated. The contribution of the high-z Herschel 500 μm sources to the cosmic SFR density is comparable to that of (sub)millimeter galaxy populations at z˜ 2.5 and at least 40% of the extinction-corrected UV samples at z˜ 4. Further investigation into the nature of these high-z dusty galaxies will be crucial for our understanding of the star formation histories and the buildup of stellar mass at the earliest cosmic epochs.