Previous Article
Next Article
Table of Contents

Looking for QSOs the Multicolor Way (1Mar95)
(from NOAO HIGHLIGHTS!, NOAO Newsletter No. 41, 1 March 1995)

Pat Osmer (Ohio State), Pat Hall (Arizona), Richard Green and the late Alain
Porter (NOAO), and Steve Warren (Imperial College) have completed a
survey for QSO candidates with the KPNO 4-m telescope, exploiting the
unique combination of faint limiting sensitivity, areal coverage, and spectral
sensitivity from 0.3 to 0.9 um that the telescope and Tek 2048 X 2048 CCD
combination offers.  Their motivations were 1) to investigate the space
density of faint QSOs at z >  4 to see if the evidence for the decline observed
at high luminosities continued for the larger numbers of QSOs at lower
luminosity; and 2) to discriminate between QSO luminosity evolution and
density evolution at redshifts near 2, where the apparent space density of
QSOs reaches a peak.  They also realized that the survey would return
information on faint field galaxies and the halo of our galaxy.  They covered
six regions on the sky with |b| >  35 deg for a total area of 0.83 sq. deg.
The 5 sigma limiting magnitude for a single exposure in B was 23.8.  They used
six filters (U, B, V, Rū, I75, and I86) (I75 and I86 are interference filters
that divide the traditional I band roughly in two).

Osmer et al. used FOCAS to identify and classify objects on the data frames,
and also carried out astrometric determinations for all the fields. They now
have a catalog of positions, magnitudes, and error estimates in the six
photometric bands for 21,375 stellar objects. Osmer et al. also have a
provisional catalogue of approximately 9,200 field galaxies in the data set.
QSO candidates were selected via four different multicolor approaches.
Spectroscopic evaluation of the candidates was then done, using the Hydra
multifiber spectrograph, whose areal coverage is well matched their fields.
The goals of this stage of the work were to explore the multicolor space for
the brighter objects that could be observed with Hydra, and then optimize the
selection techniques so that fainter candidates would include an improved
yield of true QSOs. These are observed with smaller field but higher
throughput spectrographs.

                         [Figures not included]

         Figure 1.  The B-band image of the survey field.

         Figure 2.  A spectrum of a z = 4.3 QSO discovered in the survey.
                    L[alpha] is seen at ~6,800.

The multicolor techniques used to identify QSO candidates were:

1)   Looking for outliers in five-dimensional multicolor space.  This
approach is based on that of Warren, Hewett, and Osmer (1991, 1994).  It
identifies objects in low-density regions (away from the main stellar locus)
of multicolor space and is the most general of the 4 approaches used.

2)   Identifying objects with ultraviolet excess (U-B <  -0.3).  This is the
traditional and powerful method for identifying QSOs with z up to about 3.

3)   Locating outliers in U-V / V-R and/or B-V / V-R diagrams.
Observational results and numerical simulations show that this approach is
effective for finding QSOs with redshifts between 3 and 4.

4)   Selecting extremely red (B-R >  3) objects.  This approach is effective
for QSOs with redshifts larger than 4.  The combination of Lyman alpha
emission in the R band and the depression caused by Lyman alpha
absorption along the line of sight, depresses the flux in the B band, making
QSOs at such redshifts even redder than M stars.

To date, Osmer et al. have observed approximately 350 candidates with
Hydra, repeating exposures to achieve a typical total of 4 hours per field.
The spectra show 1) 41 confirmed QSOs, with 0.6 <  z <  4.3, 2) 41
confirmed emission-line galaxies, with 0.05 <  z <  0.8, and 3) 100 stars of
different types, including white dwarfs, metal-poor stars, and very late type
M stars.

The first results show that high redshift QSOs are being found in the survey
at brighter magnitudes in approximately the numbers expected.  Osmer et al.
also find that the survey is turning up many kinds of unusual objects, such as
compact, narrow emission-line galaxies at redshifts up to and exceeding 0.5,
and a wide variety of rare stars, as might be expected from the selection of
candidates that lie away from the main stellar locus in the various multicolor
spaces.

There is also preliminary evidence for a larger number of high-redshift
QSOs at fainter magnitudes than predicted from the results of Warren,
Hewett, and Osmer (1994) and from Schmidt, Schneider, and Gunn (1991, in
the Space Distribution of Quasars, ed. D. Crampton, p. 109).  If confirmed,
this will modify our view of the evolution of QSOs at lower luminosities.
Spectroscopic observations of these candidates is now the highest priority
task for this survey.

Previous Article
Next Article
Table of Contents