Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 331, 925-933 (1998)

Previous Section Next Section Title Page Table of Contents

2. The data

2.1. X-ray observations

RX J0947.0+4721 was detected during an extensive search for AGN in the RASS by Bade et al. (1995). They found a very soft source with a count rate of [FORMULA] [FORMULA] in the total ROSAT band (0.1-2.4 keV).

Besides its presence in the RASS and the medium deep survey pointings, RX J0947.0+4721 is located in the outer part of a pointing on the Abell cluster A851. All observations were made with the PSPC detector (Pfeffermann et al. 1986) on board the ROSAT satellite (Trümper 1983), between April 1991 and October 1993. Each pointing was split into at least two observational intervals (OBIs) separated by intervals ranging from a few hours to more than a year. The OBIs are therefore treated separately for timing analysis. Table 1 lists date and length of the OBIs as well as number of counts, off-axis angle [FORMULA], count rates and hardness ratios of RX J0947.0+4721 found therein.


Table 1. Observation log of RX J0947.0+4721. Count rates are given in the 0.1-1.0 keV band. Partially shadowed OBIs are marked (-).

The entrance window of the PSPC is covered by a thin foil supported by aluminum struts (Pfeffermann et al. 1986). Depending on its position in the field of view (FOV), these struts can shadow a source. ROSAT is operated in the so-called `wobble mode' to lessen the imprints of the struts. As a consequence, sources close to the detector edge can be moved (partly) out of the FOV. RX J0947.0+4721 is obscured by the struts or moved partly out of the FOV in several OBIs. These are marked (-) in Table 1.

In pointing 700165, RX J0947.0+4721 was detected only in the second OBI, but not in the first, although it is unobscured in both. We determined the vignetting corrected number of counts between 0.1 and 1.0 keV in the source area of the first OBI, [FORMULA], and use the count rate [FORMULA] as an upper limit in the following.

In those OBIs where RX J0947.0+4721 is located closest to the detector centre ([FORMULA] [FORMULA]), a fainter and harder source was found at a distance of [FORMULA] from RX J0947.0+4721 (Fig. 1). The count rate of this neighbouring source, RX J0947.1+4721, is [FORMULA] in the total ROSAT band. With the adaptive hardness ratio method developed by Schartel et al. (1996), assuming a power law spectrum ([FORMULA]) with absorption fixed at the Galactic value, we found a [FORMULA] for RX J0947.1+4721 , and a flux of [FORMULA] in the total ROSAT band, which is about a factor 10 lower than the overall average value of RX J0947.0+4721 (see next section). The source might even be slightly fainter and harder than the values computed here, because the two sources are not fully separated.

[FIGURE] Fig. 1. Merged image in the total ROSAT band (0.1-2.4 keV) showing the four AGN RX J0947.0+4721, RX J0947.1+4721, HS0943 [FORMULA] 4725 ([FORMULA], Bade et al. 1995; Engels et al. 1997) and HS0944 [FORMULA] 4725 ([FORMULA], Engels et al. 1997). Only pointings with RX J0947.0+4721 located in the inner [FORMULA] of the PSPC FOV are merged.

In all other pointings the two sources are blended, due to the large point spread function (PSF) at larger off-axis angles, so that any extraction circle around RX J0947.0+4721 includes at least parts of the weaker source's photons. The contamination will be more significant at the higher energies where RX J0947.0+4721 is faintest. The analysis was therefore largely restricted to the 0.1-1.0 keV band (0.15-1.54 keV in the quasar's rest frame) which is called 'broad' in the following.

2.2. Optical observations

Fig. 2 shows a blue CCD image including both RX J0947.0+4721 and RX J0947.1+4721 . It was taken with the 3.5m telescope [FORMULA] focal reducer at Calar Alto in March 1994. The B-magnitudes for the two objects are [FORMULA] and [FORMULA], respectively. In both cases, optical and X-ray positions agree within [FORMULA], and no other optical sources are present in the X-ray error circles, so that the identifications can be considered as unambiguous.

[FIGURE] Fig. 2. 60s blue CCD (Tektronix, [FORMULA]) image including RX J0947.0+4721 and RX J0947.1+4721 . The circles represent the X-ray positional errors.

The identification spectrum of RX J0947.0+4721 was taken in February 1992 (Bade et al. 1995). It shows Mg II and narrow Balmer emission lines with a redshift of [FORMULA] and strong bumps due to the Fe II multiplets around [FORMULA] and [FORMULA].

An additional spectrum with [FORMULA] resolution taken at red wavelengths has the range from H [FORMULA] to the FeII [FORMULA] bump in the centre (Fig. 3 a). Due to oncoming clouds during the exposure, no absolute flux calibration is possible, only relative values are given. H [FORMULA] is of similar width and strength as in the first spectrum: [FORMULA] and [FORMULA]. (The instrumental profile has been subtracted. The quasar did not fill the slit, and so the FWHM might be slightly underestimated.) The rest frame equivalent widths for H [FORMULA] and FeII [FORMULA] are given in Table 2. The value for FeII [FORMULA] might be underestimated due to the presence of the sky absorption line at [FORMULA] at the blue side of the bump. Their ratio is FeII [FORMULA], placing RX J0947.0+4721 among the strong Fe II emitters, where strong means FeII [FORMULA] (Joly 1991).

[FIGURE] Fig. 3a and b. Spectra of a RX J0947.0+4721 and b RX J0947.1+4721. [FORMULA] is plotted against observed wavelength. Both spectra were taken in March 1994 with the 3.5m telescope at Calar Alto, equipped with focal reducer and two different grisms.


Table 2. Parameters of RX J0947.0+4721. [FORMULA]: own measurements (Molthagen et al. 1997); equivalent widths: in the QSO's rest frame.

RX J0947.1+4721 was observed with a low dispersion grism ([FORMULA]) to obtain high signal to noise in a short time. The spectrum is shown in Fig. 3 b. It contains one strong emission line at [FORMULA] which we identified with Mg II, leading to a redshift [FORMULA]. The broad bump between [FORMULA] and [FORMULA] would then contain H [FORMULA] and the [OIII] lines. No further emission lines can be seen in the spectrum, and since no strong lines are expected between MgII and H [FORMULA], we are confident that our identification is correct.

2.3. Radio observations

RX J0947.0+4721 was detected as a radio source with the Dominion Radio Astrophysical Observatory synthesis radio telescope 2 in Penticton, Canada. At 1420 MHz we found a source with a flux density [FORMULA], [FORMULA] apart from the optical position. This is within the expected uncertainty of the radio position. Measurements were also made at 408 MHz, but due to confusion with a nearby source only an upper limit of [FORMULA] was obtained.

With these two values, an upper limit for the radio spectral index ([FORMULA]), [FORMULA], can be estimated. It is thus not impossible that RX J0947.0+4721 has a flat radio spectrum, i.e. [FORMULA]. A limit for the radio-loudness [FORMULA] can be estimated from the upper limit of [FORMULA]. Wilkes & Elvis (1987) consider a quasar to be radio-loud if [FORMULA]. With [FORMULA] and [FORMULA] from Table 2, we get [FORMULA] and [FORMULA], leading to [FORMULA]. This is quite close to the dividing value, and depending on the true spectral shape, [FORMULA] can be larger than one.

The quasar is also found in the 1.4 GHz NRAO VLA Sky Survey (Condon et al. 1996). Its flux there is [FORMULA], consistent with our DRAO value.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1998

Online publication: March 3, 1998