EDP Sciences
Free access
Volume 522, November 2010
Article Number A109
Number of page(s) 5
Section Extragalactic astronomy
DOI http://dx.doi.org/10.1051/0004-6361/201014953
Published online 09 November 2010

© ESO, 2010

1. Introduction

PKS 0537 − 441 is one of the best studied blazars of the southern hemisphere. The source is strongly variable at all wavelengths. The overall spectrum in different intensity states has been discussed in detail by Pian et al. (2002). Recently, this source was the target of simultaneous long-term optical and near-infrared monitoring with the Rapid Eye Mounting (REM) telescope and UV and X-ray monitoring with the Swift satellite between 2004 December and 2005 November (Dolcini et al. 2005; Pian et al. 2007). This systematic monitoring in the optical and near-infrared showed an excursion in all bands of  ~5 mag.

According to its spectral energy distribution (SED) the source belongs to the low-energy-peaked BL Lac (LBL) or low-synchrotron-peaked (LSP; see Abdo et al. 2010) class. BL Lacs usually exhibit featureless spectra, although weak emission lines are sometimes seen during quiescent periods in some objects. Instead, a prominent emission feature at 5304 Å, likely associated to the Mg II λ2798 line, was observed in the optical spectrum of PKS 0537 − 441, which indicates a redshift of 0.894 (Peterson et al. 1976; Stickel et al. 1993). Furthermore strong, broad emission lines of Lyα and C IV were observed by Hubble Space Telescope and, on the basis of their measured central wavelengths, Pian et al. (2002) derived z = 0.896 ± 0.001. The detection of these broad emission lines suggests a mistaken classification of this object as BL Lac. At this redshift the inferred properties of PKS 0537–441 places it among the most luminous blazars.

PKS 0537 − 441 was detected in the γ-ray band by the EGRET telescope onboard the Compton Gamma-Ray Observatory (Hartman et al. 1999). Using this instrument, variability up to a factor 3 was found with a time scale of several years (Pian et al. 2002). More than ten years after the EGRET era this blazar appears also in the First AGILE-GRID Catalog of High Confidence Gamma-Ray Sources (Pittori et al. 2009) and in the First Catalog of Active Galactic Nuclei detected by the Fermi-Large Area Telescope (LAT) (Abdo et al. 2010), confirming its γ-ray activity over timescales of at least two decades. In particular, between 2008 September 15 and October 3, Fermi-LAT detected an increase of the γ-ray activity from the source (Tosti 2008) and this alert triggered multiwavelength observations by REM, Swift and AGILE observatories.

In this paper, we present the analysis of the AGILE data obtained during a ToO on PKS 0537 − 441 between 2008 October 10 and 17. In the same period the source was monitored in optical/UV and X-rays bands by Swift through 6 ToOs, performed between 2008 October 8 and 17. Moreover, the source was observed in near-infrared and optical bands by the REM telescope on October 7, 8, and 9 immediately before the AGILE observation. Throughout this paper, the quoted uncertainties are given at the 1-σ significance level, unless otherwise stated.

2. AGILE data

2.1. AGILE observations

AGILE (Astrorivelatore Gamma a Immagini LEggero) is an Italian Space Agency (ASI) mission devoted to high-energy astrophysics studies. The AGILE scientific Instrument (Tavani et al. 2008, 2009) is very compact and combines four active detectors that provide broad-band coverage from hard X-rays to γ-rays: a silicon tracker sensitive in the 30 MeV − 30 GeV energy band (ST; Prest et al. 2003; Barbiellini et al. 2001); a co-aligned coded-mask X-ray imager sensitive in the 18 − 60 keV energy band (SuperAGILE; Feroci et al. 2007; Costa et al. 2001); a non-imaging cesium iodide mini-calorimeter placed below the ST and sensitive in the 0.3 − 100 MeV energy band (MCAL; Labanti et al. 2009); and a segmented anti-coincidence system (ACS; Perotti et al. 2006). The combination of silicon tracker, mini-calorimeter, and anti-coincidence system forms the Gamma-Ray Imaging Detector (GRID).

In October 2008, during a dedicated satellite repointing on the source, the AGILE satellite devoted one week to observing the blazar PKS 0537 − 441 between 2008 October 10 11:50 UT and October 17 10:16 UT. SuperAGILE observed PKS 0537 − 441 for a total onsource net exposure time of about 308 ks.

Table 1

Swift/XRT observations and results of the spectral fits.

2.2. Data analysis and results

Level–1 AGILE-GRID data were analysed using the AGILE Standard Analysis Pipeline. After the alignment of all data times to Terrestrial Time (TT), an ad-hoc implementation of the Kalman Filter technique was used to achieve track identification and event direction reconstruction. Subsequently, the data were filtered and a quality flag was assigned to each GRID event. We selected only events flagged as confirmed γ-ray events, while all events collected during the South Atlantic Anomaly were rejected. We also rejected all γ-ray events with reconstructed directions that formed angles with the satellite-Earth vector smaller than 80°, and this reduced the γ-ray Earth Albedo contamination by excluding regions within  ~10° from the Earth limb.

Counts, exposure, and Galactic-background γ-ray maps, the last based on the diffuse emission model developed for AGILE (Giuliani et al. 2004), were created with a binsize of 0.25° × 0.25° for photons of energies higher than 100 MeV.

During the observation of PKS 0537 − 441 an unidentified source at  ~2° from the blazar showed weak activity in γ-rays. Thus, in order to separate the contributions of the two sources, the γ-ray flux of PKS 0537–441 was derived with the ALIKEMULTI2 task of the AGILE Scientific Analysis Pipeline, which allows multi-source maximum likelihood analysis considering two or more sources simultaneously.

During the period 2008 October 10–17, PKS 0537 − 441 was detected by AGILE-GRID at a significant level of 5.3-σ. The AGILE 95% maximum likelihood contour level barycentre of the source is l = 249.78°, b =  − 31.23°. The distance between this position and the PKS 0537 − 441 radio position (l = 250.08°, b =  − 31.09°) is 0.29°. The overall AGILE error circle, also taking systematic errors into account, has a radius r = 0.35°. The average γ-ray flux above 100 MeV during the entire observation period is FE > 100      MeV = (42 ± 11) × 10-8 photons cm-2 s-1. To investigate the possible variation in γ-ray activity during the AGILE observation, we divided the whole period into two intervals of equal length: P1 (2008-10-10 11:50 UT – 2008-10-13 23:03 UT) and P2 (2008-10-13 23:03 UT – 2008-10-17 10:16 UT). During the interval P1 the source was marginally detected at a significance level of 2.5-σ, with an average flux of FE > 100      MeV = (24 ± 13) × 10-8 photons cm-2 s-1, corresponding to a 2-σ upper limit of 54 × 10-8 photons cm-2 s-1 (95% confidence level). During the interval P2, however, the source was detected at a significance level of 5.7-σ, with an average flux of FE > 100      MeV = (72 ± 19) × 10-8 photons cm-2 s-1, showing a clear rise of activity of the source in the second half of the observing period.

SuperAGILE observed PKS 0537 − 441 during 2008 October 10–17 for a total onsource net exposure time of 308 ks. The source was not detected (above 5-σ) by the SuperAGILE iterative removal of sources (IROS) algorithm, which was applied to the image in the 20 − 60 keV energy range. A 3-σ upper limit of 6 mCrab was derived from the observed count rate by a study of the background fluctuations at the position of the source and a simulation of the source and background contributions with IROS.

3. Swift observations

The NASA Swift γ-ray Burst Mission (Gehrels et al. 2004) data on PKS 0537 − 441 were collected as a ToO observing campaign carried out between 2008 October 8 and 17 for a total onsource exposure of about 12 ks. All the observations were performed using all three onboard experiments: the X-ray Telescope (XRT; Burrows et al. 2005, 0.2 − 10 keV), the UV and Optical Telescope (UVOT; Roming et al. 2005, 170 − 600 nm), and the coded-mask Burst Alert Telescope (BAT; Barthelmy et al. 2005, 15 − 150 keV). The hard X-ray flux of this source is below the sensitivity of the BAT instrument for this short exposure. A 3-σ upper limit of 1.2 mCrab in the 15–50 keV energy range was derived over the whole observing period.

3.1. Swift/XRT

The XRT data were processed with standard procedures (xrtpipeline v0.12.4), filtering, and screening criteria by using the Heasoft package (v.6.8). The source count rate was low during the whole campaign (mean count rate  < 0.5 counts s-1), so we only considered photon counting (PC) data and further selected XRT event grades 0 − 12. Pile-up correction was not required. Source events were extracted from a circular region (radius of 20 pixels; 1 pixel  ~ ), while background events were extracted from an annular region centred on the source and with radii of 55 and 95 pixels. Ancillary response files were generated with xrtmkarf, and they account for different extraction regions, vignetting and PSF corrections. We used the spectral redistribution matrices v011 in the calibration database maintained by HEASARC.

A mean spectrum was extracted from the combined observations and was fit with XSPEC (v11.3.2) by adopting an absorbed power-law model with free photon index Γ and using the photoelectric absorption model tbabs (Wilms 2000) with a neutral hydrogen column fixed to its Galactic value (2.91 × 1020 cm-2; Murphy et al. 1996), consistently with Pian et al. (2007).

We also extracted spectra for two different subperiods, one before and one during the AGILE observation, and performed fits adopting the same model described above. No evidence of variability is detected in the X-ray during the Swift/XRT observations. Data were rebinned to have at least 20 counts per energy bin to allow the χ2 minimization. The fit results are reported in Table 1.

Table 2

Swift/UVOT observations and magnitudes.

3.2. Swift/UVOT

During the six Swift pointings, the UVOT (Poole et al. 2008) instrument observed PKS 0537 − 441 in the v, b, u, and the uvw1, uvm2, and uvw2 photometric bands. The uvotsource tool was used to extract counts, correct for coincidence losses, apply background substraction, and calculate the source flux. We applied a standard 5 arcsec radius source aperture, and a 20 arcsec background region. The source fluxes were dereddened using the interstellar extinction curve in Fitzpatrick (1999). The magnitudes in the different filters are reported in Table 2.

4. REM observations

Since 2004 PKS 0537 − 441 was observed routinely by the REM robotic telescope (Zerbi et al. 2001; Covino et al. 2004). The REM telescope hosts two instruments: REMIR for near-IR (Conconi et al. 2004) and ROSS for the optical (Tosti et al. 2004), used to obtain nearly simultaneous data.

Observations in the VRIJHK bands from 2004 December to 2007 July are reported in Dolcini et al. (2005) and in Pian et al. (2007), while the 2007 November – 2008 February data appear in Impiombato et al. (2008). The entire set of data from 2004 to 2009 will be published in a forthcoming paper (Impiombato et al. 2010), where details on calibration and analysis procedure can be found.

After the alert on the high γ-ray state of PKS 0537-441 detected by Fermi-LAT (Tosti 2008), REM observed the source on three consecutive days, the first occurring on 2008 October 7. The results for all the bands are reported in Table 3. No significant variability is shown during the three nights. The state of the source was relatively low. In fact the overall 2004–2009 light curve in V-band ranges between 13.3 and 16.7 (see Impiombato et al. 2010, for more details).

Table 3

REM observations and magnitudes.

5. Discussion

PKS 0537 − 441 was detected for the first time in the γ-ray band by EGRET in 1991 (Michelson et al. 1992). Subsequently it was observed in several different states (Treves et al. 1993; Hartman et al. 1999; Pian et al. 2002) with an integrated flux above 100 MeV between (16.5 ± 4.5) and (91.1 ± 14.6) × 10-8 photons cm-2 s-1. The average flux over all the EGRET observations was (25.3 ± 3.1) × 10-8 photons cm-2 s-1. However, the peak of the γ-ray emission was observed by EGRET in January 1995 at (200 ± 50) × 10-8 photons cm-2 s-1 (Pian et al. 2002), considering a temporal binning of 2 days.

During its first eleven months of observations (2008 August – 2009 July) Fermi-LAT observed an average flux of FE > 100      MeV = (37.77 ± 1.06) × 10-8 photons cm-2 s-1 (Abdo et al. 2010). This blazar is included also in the First AGILE-GRID Catalog of High Confidence Gamma-Ray Sources with an average flux of FE > 100      MeV = (42 ± 10) × 10-8 photons cm-2 s-1 (Pittori et al. 2009), confirming, also in the period 2007 July − 2008 June, the average flux level of the source observed by Fermi-LAT.

Since 2008 September 15, Fermi-LAT observed an increase in the γ-ray flux of the source. On October 3 the flux almost reached the maximum observed by EGRET in the several viewing periods. A similar level of activity was also detected by Fermi-LAT on 2008 July 8 (Bastieri 2009) and AGILE on 2010 mid-February (Lucarelli et al. 2010).

AGILE observed γ-ray activity from PKS 0537 − 441 between 2008 October 10 and 17. A multi-source maximum likelihood analysis (which also takes into account the weak emission coming from an unidentified source at  ~ 2° from the blazar) yields a source flux of FE > 100      MeV = (42 ± 11) × 10-8 photons cm-2 s-1 for photon energies above 100 MeV. This is a factor of five lower than the maximum flux detected by EGRET but comparable to the averaged flux detected by Fermi-LAT during the first eleven months of observations and practically coincident with the average flux detected by AGILE in the first year of operation. An increase by a factor of three between the first and the second halves of the observing period was observed, with a γ-ray flux value rising to (72 ± 19) × 10-8 photons cm-2 s-1. This flux is a factor of three lower than the maximum flux observed by EGRET.

During 2004–2005 the source was intensively monitored by REM and Swift and these long-term observations showed that the optical V-band and X-ray light curves are highly correlated, even if the V-band varies with much higher amplitude: an optical flux variation of a factor of 60 was observed between 2004 December and 2005 February, whereas in the X-ray band Swift/XRT observed a variation only of a factor 2.

thumbnail Fig. 1

The SED of PKS 0537 − 441, including near-infrared and optical REM data, UV, soft X-ray and hard X-ray Swift data, and γ-ray AGILE data. Blue symbols refer to the period 2008 October 7–13, red symbols refer to the period 2008 October 13–17. The Swift/XRT and Swift/BAT data (in green) are averaged over the whole observing period. For Swift/UVOT are represented the data collected in the u, b, uvw1 filters, with the data of the first period averaged over all the available observations. The black, blue and orange lines represent the one-component SSC, two-component SSC, and EC model, respectively. The dashed orange line represents the emission from the disk in the EC model.

Open with DEXTER

The average flux observed by Swift during 2008 October obtained by combining the 6 ToO observations, F(0.3 − 10 keV) = (7.7 ± 0.6) × 10-12 erg cm-2 s-1, seems to indicate an intermediate activity state of the source, with respect to the range of flux detected during the observations carried out by Swift in 2005, (3.9 − 14.9) × 10-12 erg cm-2 s-1. No significant variability in X-rays is noted between the different observations.

During the period October 10–17 the source was also observed in the 2 − 10 keV energy band by the All Sky Monitor (ASM) onboard the Rossi X-ray Timing Explorer (RXTE), and on October 13 there is a 3-σ detection with a flux of the order of 2 × 10-10 erg cm-2 s-1, which, while considering the low significance of the detection and the relative high uncertainties, could be a hint of the increase in activity of the source in X-rays, also observed in the γ-ray band. Moreover, the fluxes observed by Swift/UVOT in u, b, uvw1 bands increased by about 0.5 mag between the observations in October 12 and 17, possibly indicating a rise in the activity of PKS 0537 − 441.

The SED for the AGILE, Swift, and REM data of 2008 mid-October is shown in Fig. 1. Our modelling of the spectrum starts with an assumed electron distribution in the form of a broken power-law (1)where ζ1 and ζ2 are the spectral indices before and after the break Lorentz factor γb. These electrons emit a primary synchrotron spectrum, and a second contribution is then produced by inverse Compton as the primary synchrotron photons scatter off the same electron population. If we assume that the broad optical/UV peak is due to synchrotron emission from a population of high-energy electrons in a random magnetic field B, then the synchrotron self-Compton (SSC) emission can account for the strong γ-ray flux observed by GRID.

In this case, SSC contribution by a second population of low-energy electrons is required to account for the stable and moderate emission in X-rays, so we adopt a two-component SSC model. For the first component the parameters are: γb = 500; γmin = 100; ζ1 = 2; ζ2 = 5; K = 70cm-3 and B = 1.0G. For the second one, γb = 6500; γmin = 3000; ζ1 = 2; ζ2 = 5; K = 20cm-3 and B = 0.15G. In this way, the total jet power is Lj = 5 × 1045erg cm-2s-1 (see Celotti & Ghisellini 2008).

A standard one-component SSC model can only marginally reproduce the optical bump and the X-ray spectrum simultaneously. Moreover, the very strong jet power involved in such a model, up to Lj = 9 × 1045erg s-1, is consistent with the Blandford-Znajek (1977) mechanism of power extraction from a rotating black hole (BH) only for BH masses higher than 2 × 109   M. For this model the parameters are γb = 6500; γmin = 230; ζ1 = 2; ζ2 = 5; K = 10cm-3 and B = 0.2G.

We also propose an alternative model where the optical bump comes from the accretion disk emission, with a subsequent significant contribution in the high-energy bump of the external Compton (EC) scattering of both direct-disk radiation and photons reprocessed by the broad line region. In this case remarkable disk power of Ld ~ 3 × 1046 erg s-1 would be required with the following jet parameters: γb = 400; γmin = 100; ζ1 = 2; ζ2 = 5 K = 70cm-3, and B = 1.6G. In all models the emission region is assumed to have radius R = 2.3 × 1016 cm and to move with bulk Lorentz factor Γ = 15 at an angle  ~ 2° with respect to the observer. BL Lacs are usually associated with low photon-density ambients; however, past observations of broad emission lines in the optical/UV spectrum of this object (Pian et al. 2002) suggest that, at least during significant activity states, the EC contribution could be important for the radiation emission in the high-energy part of the spectrum. Only the long-term monitoring in γ-rays by the AGILE and Fermi satellites, together with suitable multifrequency coverage, can firmly conclude on the EC contribution in this object and on the real nature of this peculiar blazar.


We thank the referee, R. Hartman, for his very useful suggestions and comments. The AGILE Mission is funded by the Italian Space Agency (ASI) with scientific and programmatic participation by the Italian Institute of Astrophysics (INAF) and the Italian Institute of Nuclear Physics (INFN). We wish to express our gratitude to the Carlo Gavazzi Space, Thales Alenia Space, Telespazio and ASDC/Dataspazio Teams that implemented the necessary procedures to carry out the AGILE re-pointing. We thank the Swift Team for making these observations possible, particularly the duty scientists and science planners. We thank A. Beardmore for useful discussions. This investigation was carried out with partial support under ASI contract N. I/089/06/1. We acknowledge financial support by the Italian Space Agency through contract ASI-INAF I/088/06/0 for the Study of High-Energy Astrophysics.

Facilities: AGILE, REM, Swift.


All Tables

Table 1

Swift/XRT observations and results of the spectral fits.

Table 2

Swift/UVOT observations and magnitudes.

Table 3

REM observations and magnitudes.

All Figures

thumbnail Fig. 1

The SED of PKS 0537 − 441, including near-infrared and optical REM data, UV, soft X-ray and hard X-ray Swift data, and γ-ray AGILE data. Blue symbols refer to the period 2008 October 7–13, red symbols refer to the period 2008 October 13–17. The Swift/XRT and Swift/BAT data (in green) are averaged over the whole observing period. For Swift/UVOT are represented the data collected in the u, b, uvw1 filters, with the data of the first period averaged over all the available observations. The black, blue and orange lines represent the one-component SSC, two-component SSC, and EC model, respectively. The dashed orange line represents the emission from the disk in the EC model.

Open with DEXTER
In the text