The Highest Redshift Quasars

(Dec 8, 1998)

Since the time the following was written, we have obtained spectra of two additional high-redshift quasars from the SDSS imaging data. One has z = 4.9; the other has z = 5.0. These both beat the previous record redshift of z=4.89 by Schneider, Schmidt, and Gunn (1991), and thus the SDSS now has found three of the top four highest redshift quasars known.

The SDSS will use the colors and morphologies of objects to identify quasar candidates from the photometric data: objects with stellar appearance and colors that lie well outside the stellar locus in color space will be flagged for spectroscopic investigation.

Fig.1. Observed color-color diagrams of 20 square degrees from the SDSS test data r* < 20). The positions of 22 newly discovered quasars (selected from 130 square degrees) are indicated. Already known quasars are not indicated in this figure.
Figure 1 shows the color-color diagram of stellar objects with r* < 20 from 20 square degrees of SDSS imaging commissioning data taken in September 1998¹. Notice the narrowness of the distribution: this is a tribute both to the quality of the data, and the pipeline used to reduce it. As the SDSS spectrographs have not been commissioned as of this writing, we are using the Double Imaging Spectrograph on the Apache Point 3.5m telescope to carry out spectroscopy of promising high-redshift quasar candidates. Superposed on Figure 1 are the places in color-color space where the 22 new quasars we have identified thus far lie, based on roughly 130 square degrees of imaging data.
These quasars do not by any means consistute a complete sample. We have recently concentrated on those objects which appeared from their broad-band colors to be high-redshift candidates. Out of 11 candidates, 9 are indeed quasars at z > 3.65 (the two high-redshift quasars previously known in the survey area also stood out cleanly in the color-color diagrams, and would have been selected as well). All are brighter than i* = 20. This success rate far surpasses the typical 10% found in the literature for high-redshift quasar surveys (Schneider etal 1994; Hall etal 1996; Kennefick etal 1995), although again, we do not have a complete sample to make this quantitative.

Fig 2. Spectra of a new SDSS quasars with z = 4.75. Clicking on the image will bring up 3 more spectra, 3 QSOs with z > 4, plus one with a broad absorption-line spectrum, obtained with the 3.5m ARC telescope and Double Imaging Spectrograph.
Figure 2 shows our spectra of the three highest-redshift quasars we have found thus far, plus one which shows strong associated absorption. The one at z = 4.75 is the second-highest redshift quasar known (the current redshift holder is z = 4.89; see Schneider etal 1991). These spectra are of quite low resolution, roughly 7Å/pixel, while the SDSS spectrographs will deliver 1-1.5Å/pixel over a similar wavelength coverage. These objects were selected from roughly 1% of the sky that the SDSS will image. We therefore suspect that there are enormously more high-redshift quasars to be discovered as part of the SDSS.

Hall, P.B., Osmer, P.S., Green, R.F., Porter, A.C., & Warren, S.J. 1996, AJ, 462, 614
Kennefick, J.D. et al. 1995, AJ, 110, 78
Schneider, D. P., Schmidt, M., & Gunn, J.E. 1991, AJ, 102, 837
Schneider, D. P., Schmidt, M., & Gunn, J.E. 1994, AJ, 107, 1245

¹The asterisk * indicates that the final SDSS photometric system has not yet been defined; this is preliminary photometry, accurate to perhaps 0.05 mag.