Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 348, 63-70 (1999)

Previous Section Next Section Title Page Table of Contents

4. Results

4.1. Light curves

We have extracted light curves from the data of the three MECS units (merged together) using XSELECT and considering a circle of radius [FORMULA] centered around the target. In Fig. 1 we show the total MECS count rates obtained with the XIMAGE program corresponding to the 8 observations. This light curve is consistent with a constant. We can set an upper limit on the amplitude of possible variations of [FORMULA]40%. No significant variations are present within the single observations either.

[FIGURE] Fig. 1. The top panel shows the MECS light curve, corresponding to the 8 observations. In the middle panel we report the optical light curve in the same period, in the [FORMULA] filter. The bottom panel shows the light curve at 8.3 and 2.25 GHz, by the Green Bank monitoring campaign.

Simultaneously with the Beppo SAX observations, the source was observed in the optical, with the REOSC 1.05 meter telescope of Pino Torinese (see Villata et al. 1997 for a description of the instrument and the data analysis procedure), and in the radio band with the Green Bank Interferometer (GBI). As shown by the middle and lower panels of Fig. 1, the source was varying significantly in the optical (with a probability [FORMULA]) and at 8.3 GHz.

We also observed the source in the near infrared with the 1.9 m telescope at Sutherland using the MkIII photometer, obtaining [FORMULA] on Mar 2, 1998 (JD=2450510), using a 12" diameter aperture.

4.2. Spectral fitting

4.2.1. LECS and MECS

Spectra of the LECS and MECS detectors have been extracted using XSELECT and a circle of radius 4´ around the target for both the instruments (given the weak flux of the source and the heavy absorption under 0.7 keV, we have not used the [FORMULA] radius recommended for the LECS for brighter sources, to minimize the effects of the background, especially at low energies). As suggested by the Science Data Center of Beppo SAX (SDC), after checking that the background was constant during the observations, we have subtracted the background using the blank sky fields selecting the same detector regions where the target is located. We used the response matrices released by the SDC in September 1997. The LECS and MECS spectra have been jointly fit after allowing for a constant rescaling factor of 0.9 for the LECS data (this factor accounts for uncertainties in the inter-calibration of the instruments; Fiore, priv. comm.).

Fitting the 8 observations separately with a single power law and free hydrogen column density, [FORMULA], gives consistent values, within the (rather large) uncertainties. We therefore decided to combine the 8 spectra, in order to increase the signal to noise ratio. A single power law fit to the LECS+MECS dataset (see Fig. 2) yields an energy spectral index [FORMULA] with [FORMULA] cm-2 (unless otherwise indicated, the errors from the fits are at 90% confidence level, for two parameters of interest). No spectral feature is required by the fit and in particular an upper limit of [FORMULA] eV can be set to the equivalent width (in the source frame) of the 6.4 keV fluorescence [FORMULA] iron line (see Sect. 5.1 for more discussion). The results of the spectral fitting are reported in Table 2.

[FIGURE] Fig. 2. Fit to the LECS and MECS Beppo SAX spectrum. PDS data are not used for the fit, and appear to lie above the extrapolation of the power law from lower energies.


Table 2. Fits to Beppo SAX Data: single power law

As mentioned in the Introduction, the value of the [FORMULA] column density is uncertain, yet its determination is important, since the knowledge of the corresponding optical extinction allows to determine the location of the synchrotron peak. Beppo SAX data, alone, cannot well constrain [FORMULA], due to the faint level of the source. PKS 0528+134 was instead much brighter during the ASCA observations of March 1995, allowing a better determination of the [FORMULA] column [[FORMULA] cm-2], as described in Sect. 5.1.

Since the Beppo SAX measurement of the [FORMULA] column is consistent with the value found by ASCA, we fixed the [FORMULA] at the latter value. In this case the LECS+MECS data can be modeled by a single power law with an energy index [FORMULA] (90% confidence level for one parameter of interest).

4.2.2. PDS

Source visibility windows were selected following the criteria of no earth occultation and high voltages stability during the exposures. In addition, the observations closest to the South Atlantic anomaly were discarded from the analysis. From collimators positions the ON and OFF time windows were also created and merged with the source visibility window to create the final time windows on which the source+background and background spectra were accumulated for each of the four PDS units, using the XAS software package. A filtering for the temperature and energy dependence of the pulse rise time was used.

The source was significantly detected up to [FORMULA]90 keV. For each pointing, the grouped spectra from the four units were coadded. The net count rates are reported in Table 1.

To increase the S/N, we have averaged the eight PDS spectra and binned the resulting spectrum in 4 energy intervals. For the fitting, we used the response matrix available at SDC, keeping the rescaling factor with respect to the MECS constant and equal to 0.85 (this factor allows for uncertainties in the inter-calibration of the instruments).

As shown in Fig. 2, the PDS data points lie above the model fit for the LECS and MECS datasets. In particular, including the PDS data and using a single power law fit with [FORMULA] fixed to the value found by ASCA, we obtain [FORMULA], with the PDS points still lying above the fit (see Table 2 for details).

The same three datasets (LECS+MECS+PDS) were also fitted with a broken power law model with [FORMULA] and [FORMULA] and break energy [FORMULA] keV, leading to a [FORMULA] for 57 d.o.f. This model significantly improves the fit, according to the F-test ([FORMULA]), but the [FORMULA]-value yields a probability less than 5% that the model is correct. A power law fit to the PDS data alone gives a spectral index [FORMULA].

As can be seen, the high energy flux detected by the PDS shows a clear discontinuity with respect to the LECS/MECS data. A possible explanation can be that the PDS data are contaminated by the presence of another source in its field of view. The PDS field of view is [FORMULA]1.3 degree, with no imaging capability. The response matrix of this instrument is triangular, with a flat top of 3´ and a reduction of a factor 2 in sensitivity at 38´ from the center (Frontera et al. 1997). The MECS have a field of view of radius [FORMULA], and there are no sources other than PKS 0528+134 detected at more than 3[FORMULA].

We checked the ROSAT PSPC WGA catalog (White et al. 1994) for sources at an angular distance [FORMULA] degree from PKS 0528+134. There are two unidentified sources (1WGA J0532.9+134, at a distance of 31 arcmin from PKS 0528+134, and 1WGA J0533+135, at 41 arcmin) with a count rate in the ROSAT band comparable to that of PKS 0528+134. No sources were instead found in the EXOSAT and ASCA public archives. Given the low galactic latitude of PKS 0528+134 and the large offset of the PDS data points with respect to the extrapolation of the MECS spectrum, we think that the PDS data might be contaminated by other sources. For this reason, in the discussion of the SED of PKS 0528+134, we will not consider the PDS spectrum.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: July 16, 1999