There are some systematics in the published SDSS QSO redshifts (see Figure 1 and Figure 2 and note the wiggles). A careful construction of a master QSO template makes these systematics go away. Very important for statistical analyses of QSO clustering for example. Download Table 4 from the paper here , and the new master QSO composite spectrum (cross-correlation template) here . NEW: Download the full DR7 QSO catalogue here and the README file here .

Figure 1: Redshift differences, Delta(z) /(1 + z), between SDSS and Princeton pipeline reductions. Spectra plotted possess SDSS final-redshifts with high-confidence (zConf>0.9) derived via cross-correlation (zstatus=3 or 4). Large differences between redshifts extend to amplitudes of +/- 5×10−3 (+/- 1500 km/s). Particularly striking is the sequence of apparent discontinuities in the behaviour of the redshift differences as a function of redshift. Data for 69 915 spectra are included (503 lie outside the y-axis range plotted).

Figure 2: Redshift differences, Delta(z) /(1 + z), between SDSS redshifts and redshifts derived from the SDSS-determined Mg II emission line locations. Spectra plotted possess Mg I I emission line SNR>10. The solid red line, calculated using a 2001-point running median of the data-points, shows the form of the systematic trends with redshift. Systematic redshift differences of ~500 km s −1 shifts over small redshift intervals are evident and, over a larger redshift interval, a prominent systematic trend of 2×10 −3 (600 km s −1 ) can be seen. Data for 60 190 spectra are included (2101 lie outside the y-axis range plotted).

Figure 3: Redshift differences, Delta(z) /(1 + z ), between HW redshifts and redshifts derived from the SDSS-determined Mg II emission line locations. Spectra plotted possess Mg I I emission line SNR>10. The solid red line, calculated using a 2001-point running median of the data-points, demon- strates the complete removal of detectable systematic trends with redshift. Contrast the behaviour with that shown in Fig. 2 . Data for 60 120 spectra are included (559 lie outside the y-axis range plotted).

---

Abstract: A systematic investigation of the relationship between different redshift estimation schemes for more than 91000 quasars in the Sloan Digital Sky Survey (SDSS) Data Release 6 is presented. The publicly available SDSS quasar redshifts are shown to possess systematic biases of (600kms-1) over both small and large redshift intervals. Empirical relationships between redshifts based on (i) CaII H&K host galaxy absorption, (ii) quasar [OII]3728, (iii) [OIII]4960, 5008 emission and (iv) cross-correlation (with a master-quasar template) that includes, at increasing quasar redshift, the prominent MgII2799, CIII]1908 and CIV1549 emission lines are established as a function of quasar redshift and luminosity. New redshifts in the resulting catalogue possess systematic biases, a factor of 20 lower compared to the SDSS redshift values; systematic effects are reduced to the level of (30km/s) per unit redshift or per unit absolute magnitude. Redshift errors, including components due both to internal reproducibility and to the intrinsic quasar-to-quasar variation among the population, are available for all quasars in the catalogue. The improved redshifts and their associated errors have wide applicability in areas such as quasar absorption outflows, quasar clustering, quasar-galaxy clustering and proximity-effect determinations.