The QDOT all-sky IRAS galaxy redshift survey

We describe the construction of the QDOT survey, which is publicly available from an anonymous FTP account. The catalogue consists of infrared properties and redshifts of an all-sky sample of 2387 IRAS galaxies brighter than the IRAS PSC 60-mu m completeness limit (S-60 > 0.6 Jy), sparsely sampled at a rate of one-in-six. At \b\ > 10 degrees, after removing a small number of Galactic sources, the redshift completeness is better than 98 per cent (2086/2127). New redshifts for 1401 IRAS sources were obtained to complete the catalogue; the measurement and reduction of these are described, and the new redshifts tabulated here. We also tabulate all sources at \b\ > 10 degrees with no redshift so far, and sources with conflicting alternative redshifts either from our own work, or from published velocities. A list of 95 ultraluminous galaxies (i.e. with L-60 mu m > 10(12) L.) is also provided. Of these, similar to 20 per cent are AGN of some kind; the broad-line objects typically show strong Fe II emission. Since the publication of the first QDOT papers, there have been several hundred velocity changes: some velocities are new, some QDOT velocities have been replaced by more accurate values, and some errors have been corrected. We also present a new analysis of the accuracy and linearity of IRAS 60-mu m fluxes, We find that the flux uncertainties are well described by a combination of 0.05-Jy fixed size uncertainty and 8 per cent fractional uncertainty, This is not enough to cause the large Malmquist-type errors in the rate of evolution postulated by Fisher et al, We do, however, find marginal evidence for non-linearity in the PSC 60-mu m flux scale, in the sense that faint sources may have fluxes overestimated by about 5 per cent compared with bright sources. We update some of the previous scientific analyses to assess the changes. The main new results are as follows. (1) The luminosity function is very well determined overall but is uncertain by a factor of several at the very highest luminosities (L-60 mu m > 5 x 10(12) L.), as this is where the remaining unidentified objects are almost certainly concentrated. (2) The best-fitting rate of evolution is somewhat lower than our previous estimate; expressed as pure density evolution with density varying as (1 + z)(P), we find p = 5.6 + 2.3, Making a rough correction for the possible (but very uncertain) non-linearity of fluxes, we find p = 4.5 + 2.3. (3) The dipole amplitude decreases a little, and the implied value of the density parameter, assuming that IRAS galaxies trace the mass; is Omega = 0.9(+0.45, -0.25). (4) Finally, the estimate of density variance on large scales changes negligibly, still indicating a significant discrepancy from the predictions of simple cold dark matter cosmogonies.