Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 323, L21-L24 (1997)

Previous Section Next Section Title Page Table of Contents

3. Discussion

The probable optical counterpart of RX J1028.6-0844 is the QSO. The star #B is too faint to be a likely counterpart. For F to G type stars the expected [FORMULA] is of the order of -4.6 to -2.5 (Stocke et al. 1991 , Maccacaro et al. 1988 ). With the B and R magnitudes given in Tab. 2 and the spectral type the estimated V magnitude is about [FORMULA]. With an energy conversion factor ECF from count rate to flux of 6 [FORMULA] cts cm2 erg-1 (Paper II) we obtained [FORMULA]. Taking the correction for the ROSAT energy band into account (cf. Paper II) this is about 2 dex above the limit for late F to early G-type stars. Likewise, the elliptical galaxy, object #C, is optically too faint to be a plausible counterpart. The estimated V magnitude is of the order of [FORMULA]. As discussed in Paper II, elliptical galaxies are expected to be brighter than [FORMULA] in order to have detectable X-ray emission in the RASS. The diffuse object #d to the East of #A also is a galaxy. A BL Lac object as bright as object #d would lead to a plausible [FORMULA] ratio. Spectroscopically we cannot rule out that #d is a BL Lac object. However, based upon its morphology in our direct images (cf. Fig. 3) we consider this possibility very unlikely. It is more probably a spiral galaxy which is optically by far too faint to be the counterpart of the X-ray source. Hence, the most likely optical counterpart for the X-ray source RX J1028.6-0844 is the QSO whose optical position from the COSMOS UKST data base (plate epoch 2000.0) is given in Tab. 1. The positional uncertainty is [FORMULA] in right ascension and declination, respectively.

Assuming a photon index of [FORMULA] and a galactic absorption [FORMULA] = 4.59 1020 cm-2 (Dickey & Lockman 1990 ) the observed count rate of [FORMULA] cts s-1 yields an unabsorbed flux in the 0.1-2.4 keV energy band of [FORMULA] = 8.33 10-13 erg s-1 cm-2 (7.0 10-13 erg s-1 cm-2). Due to the high redshift the observed energy range corresponds to 0.53-12.66 keV in the quasar rest frame. The given flux yields a luminosity in the standard 2-10 keV rest frame energy range of [FORMULA] = 6.4 [FORMULA] erg s-1 ([FORMULA] erg s-1). Likewise, the luminosity in the 0.1-2.4 keV energy range is [FORMULA] = 1.3 [FORMULA] erg s-1 ([FORMULA] erg s-1).

By cross-correlation with the NED and SIMBAD data bases we found a radio source, PKS B1026-084, at a distance of 23 [FORMULA] from the RASS X-ray position. The distance of the radio source to the QSO which very likely is its optical counterpart, is only 5 [FORMULA]. The radio source was also detected at 4.85 GHZ in the Parkes-MIT-NRAO (PMN) Survey (Griffith et al. 1995 ) at about 22 [FORMULA] distance to the QSO. The radio fluxes at 2.7 GHz and 5 GHz in the Parkes catalog of radio sources (Parkes Catalogue 1990 , cf. Otrupcek & Wright 1991 ) are 270 mJy and 220 mJy, respectively, implying a spectral energy index of [FORMULA]. In the PMN survey a 4.85 GHz flux of [FORMULA] mJy was measured. The positions of the radio sources are indicated in Fig. 1. The continuum coefficients [FORMULA] and [FORMULA] (Tananbaum et al. 1979 ) were calculated using [FORMULA] (in erg s-1 cm-2 Hz-1) derived from the Cousins R magnitude assuming an energy index [FORMULA] and the absolute flux calibration for Cousins R given in Lamla (1982 ):


The 4.85GHz radio flux [FORMULA] was k-corrected using [FORMULA]. We obtained [FORMULA] and [FORMULA] which places RX J1028.6-0844 among the radio-loud QSOs in the [FORMULA] diagram (e.g. Stocke et al. 1991 ).

An important finding of pointed ROSAT observations of high redshift QSOs at [FORMULA] was that radio-loud QSOs show a low energy cut-off in the X-ray spectra which cannot be explained by absorption due to intervening gas in our galaxy but rather seems to be due to intrinsic absorption in these QSOs (Elvis 1996 ). The fraction of absorbed quasars seems to increase with redshift. Interestingly, RX J1028.6-0844 exhibits a very hard X-ray spectrum indicating strong absorption of the soft X-ray photons. Although the number of detected photons is too small to directly fit a model spectrum, we can make use of the observed hardness ratios [FORMULA] and [FORMULA] to study the spectral energy distribution. Fitting a power law energy distribution to the observed hardness ratios with the galactic value of [FORMULA] formally yields a photon index of [FORMULA], i.e. an increasing flux distribution with increasing energy, however, with a poor fit only. (This [FORMULA] value would lead to practically the same flux [FORMULA] as given above.) This result can be understood in terms of intrinsic absorption in the QSO leading to the hard observed spectrum. In order to investigate this qualitatively we assumed a warm absorber model. Fixing the photon index at [FORMULA] and [FORMULA] at the galactic value, we additionally assumed the presence of absorption edges at rest energies [FORMULA] with an optical depth [FORMULA]. The absorption was assumed to be proportional to [FORMULA] for [FORMULA] with E being the energy, and [FORMULA] given in units of keV3 (cf. Zimmermann et al. 1994 ) Due to the high redshift of RX J1028.6-0844 the oxygen absorption edge at 0.8 keV has an effect on the observed spectrum at low energies. [FORMULA] is, however, mainly affected by the galactic absorption. For the galactic [FORMULA] a [FORMULA] at the oxygen absorption edge energy is sufficient to reproduce the observed [FORMULA]. In order to reproduce the observed hardness ratio [FORMULA] a second absorption edge at [FORMULA] keV is needed with a rather large optical depth of the order of [FORMULA]. This absorption edge is probably due to silicon and sulfur. It indicates the presence of gas with a very high column density on the order of [FORMULA] [FORMULA] cm-2 at a temperature around [FORMULA] K (assuming an incident power law energy distribution with an energy index of 1, Krolik & Kallman 1984 ). RX J1028.6-0844 thus appears to be a further example of an absorbed high-redshift quasar.

Based on their deep pointed ROSAT survey Henry et al. (1994 ) estimated a surface density for [FORMULA] QSOs of [FORMULA] deg-2. We found one such object in our complete X-ray flux-limited sample covering an area of 685 deg2. The average flux limit of our survey is on the order of 50 (20 to 100) higher than in the survey of Henry et al. (cf. Paper II). Assuming a logarithmic slope of 1.6 for the [FORMULA] distribution (cf. Boyle et al. 1993 ) we hence would expect on the order of [FORMULA] times less [FORMULA] QSOs, i.e. [FORMULA] deg-2. Actually we found [FORMULA] deg-2 which is in agreement with Henry et al.. It is, however, clear that a considerably larger data base than presently available is needed in order to enable statistically significant conclusions on the number density of X-ray selected high-redshift QSOs.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998