Breaking the 'redshift deadlock' - I. Constraining the star formation history of galaxies with submillimetre photometric redshifts

Future extragalactic submillimetre and millimetre surveys have the potential to provide a sensitive census of the level of obscured star formation in galaxies at all redshifts. While in general there is good agreement between the source counts from existing Submillimetre Common User Bolometer Array (SCUBA; 850-mum) and Max Planck Millimetre Bolometer Array (MAMBO; 1.25-mm) surveys of different depths and areas, it remains difficult to determine the redshift distribution and bolometric luminosities of the submillimetre and millimetre galaxy population. This is principally due to the ambiguity in identifying an individual submillimetre source with its optical, IR or radio counterpart which, in turn, prevents a confident measurement of the spectroscopic redshift. Additionally, the lack of data measuring the rest-frame FIR spectral peak of the submillimetre galaxies gives rise to poor constraints on their rest-frame FIR luminosities and star formation rates. In this paper we describe Monte Carlo simulations of ground-based, balloon-borne and satellite submillimetre surveys that demonstrate how the rest-frame FIR-submillimetre spectral energy distributions (250-850 mum) can be used to derive photometric redshifts with an rms accuracy of +/-0.4 over the range 0 < z < 6. This opportunity to break the redshift deadlock will provide an estimate of the global star formation history for luminous optically obscured galaxies (L-FIR > 3 x 10(12) L.) with an accuracy of similar to20 per cent.