2010MNRAS.406..720R

We present a catalogue of 179 hyperluminous infrared galaxies (HLIRGs) from the Imperial IRAS-FSS Redshift (IIFSCz) Catalogue. Of the 92 with detections in at least two far-infrared bands, 62 are dominated by an M82-like starburst, 22 by an Arp 220-like starburst and eight by an active galactic nucleus (AGN) dust torus. On the basis of previous gravitational lensing studies and an examination of Hubble Space Telescope (HST) archive images for a further five objects, we estimate the fraction of HLIRGs that are significantly lensed to be 10-30 per cent.

We show simple infrared template fits to the spectral energy distributions of 23 HLIRGs with spectroscopic redshifts and at least five photometric bands. Most can be fitted with a combination of two simple templates: an AGN dust torus and an M82-like starburst. In the optical, 17 of the objects are fitted with quasi-stellar object (QSO) templates, several with quite strong extinction. There are five objects fitted with galaxy templates in the optical, two of which show evidence for AGN dust tori and so presumably contain Type 2 (edge-on) QSOs. The remaining object is fitted with a galaxy template in the optical, but is of such high luminosity that this classification would be plausible only if there were very strong lensing. 20 of the 23 objects (87 per cent) show evidence of an AGN either from the optical continuum or from the signature of an AGN dust torus, but the starburst component is the dominant contribution to bolometric luminosity in 14 out of 23 objects (61 per cent). The implied star formation rates, even after correcting for lensing magnification, are in excess of .

We use infrared template-fitting models to predict fluxes for all HLIRGs at submillimetre wavelengths and show predictions at 350 and 850 µm. Most of them would have 850 µm fluxes brighter than 5 mJy, so they should be easily detectable with current submillimetre telescopes. At least 15 per cent should be detectable in the Planck all-sky survey at 350 µm, and all Planck all-sky survey sources with z < 0.9 should be IIFSCz sources.

From the luminosity-volume test, we find that HLIRGs show strong evolution. A simple exponential luminosity evolution applied to all HLIRGs would be consistent with the luminosity functions found in redshift bins 0.3-0.5, 0.5-1 and 1-2. The evolution-corrected luminosity function flattens towards higher luminosities, perhaps indicating that a different physical mechanism is at work compared to lower luminosity starbursts. In principle, this could be gravitational lensing though previous searches with HST have, perhaps surprisingly, not shown lensing to be widely prevalent in HLIRGs.