EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 543 (July 2012) A&A, 543 (2012) A130 Full HTML
Free Access
Issue
A&A
Volume 543, July 2012
Article Number A130
Number of page(s) 9
Section Stellar structure and evolution
DOI https://doi.org/10.1051/0004-6361/201218869
Published online 10 July 2012

© ESO, 2012

1. Introduction

The launch of NASA’s Fermi Gamma-ray Space Telescope in June 2008 marked a breakthrough in γ-ray observations of pulsars. The Large Area Telescope (LAT; Atwood et al. 2009) has now detected more than 100 γ-ray pulsars, to be compared with the 7 detected by the Compton Gamma Ray Observatory (CGRO) in 10 years of operation (see, e.g. Thompson 2008, for a review). Some of them are associated with known, or recently discovered, radio pulsars, while others are yet undetected in the radio despite some deep searches and are, thus, members of the long-hypothesised class of radio-silent γ-ray pulsars, of which Geminga was the first example (e.g., Bignami & Caraveo 1996; Abdo et al. 2010a). In both cases, multi-wavelength follow-ups are important to study the pulsar radiation processes over several decades in energy, to constrain the source distance and position, and to search for pulsar wind nebulae (PWNe), still challenging targets in γ-rays despite the unprecedented spatial resolution of the Fermi-LAT (e.g., Ackermann et al. 2011). In the X-rays, many Fermi pulsars had already been observed prior to their γ-ray detection, while for the others systematic follow-up observations have been carried out with Chandra, XMM-Newton, and Swift (see, e.g. Marelli et al. 2011, for a summary of the X-ray observations of Fermi pulsars). In the optical, the source coverage with 8 m-class telescopes is still sparse with only a few Fermi pulsars having been observed beforehand and only a few deep follow-up observations on well-selected targets having been carried out so far (Mignani et al. 2011; Shearer et al., in prep.; De Luca et al., in prep.).

The radio pulsar PSR  J1028 − 5819 (P = 91.4 ms) was discovered by Keith et al. (2008) during a 3.1 GHz survey of a sample of low-latitude unidentified EGRET error boxes performed with the Parkes radio telescope and the Australia Telescope Compact Array (ATCA), prior to the launch of Fermi. The pulsar’s period derivative Ṗ = 1.61 × 10-14 s s-1 yields a spin-down age of  ~90 kyr, a spin-down power Ė ~ 8.3 × 1035 erg s-1, and a dipolar magnetic field B ~ 1.2 × 1012 G. Thus, PSR J1028−5819 may be considered a transition object between the young, Vela-like pulsars and the middle-aged ones. The dispersion measure (DM = 96.525  ±  0.002 cm-3 pc) implies a distance of  ~2.3 kpc, according to the Galactic electron-density model of Cordes & Lazio (2002). Initially associated with the EGRET source 3EG 1027−5817 upon its coincidence with the γ-ray error box, PSR J1028−5819 was identified as a γ-ray pulsar by Fermi (Abdo et al. 2009a) and included in both the Fermi catalogue Bright Source List (Abdo et al. 2009b) and the first Fermi-LAT catalogue of γ-ray pulsars (Abdo et al. 2010b). As for many Fermi pulsars, the γ-ray light curve is characterised by a double-peak profile, with a phase separation of  ~0.46, which trails the single radio peak by  ~0.2 in phase. The pulsar’s γ-ray spectrum is fit by a power law (PL) with an exponential cut-off, a photon index Γγ = 1.25, and a cut-off energy Ec = 1.9 GeV. Possible evidence of a PWN was also found in the Fermi data (Ackermann et al. 2011). Recently, PSR J1028 − 5819 was found to be spatially coincident with a TeV γ-ray source HESS J1023−575 (Abramowski et al. 2011), whose emission is probably associated with the PWN. In the X-rays, PSR J1028−5819 was identified with a faint X-ray source detected by the Swift X-ray Telescope (XRT) in a short follow-up observation (Abdo et al. 2009a; Marelli et al. 2011) of the γ-ray source error box. Unfortunately, the low detection significance ( ≲ 5σ) hampered the characterisation of the pulsar’s X-ray spectrum.

While deeper X-ray observations of PSR  J1028−5819 are now in progress with Chandra, no optical observations have yet been performed. Here, we present the first optical observations of PSR J1028−5819, performed with the Very Large Telescope (VLT) as a part of a planned survey to optically identify Fermi pulsars. To complement our multi-wavelength analysis, we also present a re-analysis of the published Swift/XRT observation of the pulsar and the first results of a more recent, yet unpublished, Suzaku observation available in the public science data archive.

This paper is organised as follows: observations, data reduction and analysis are described in Sect. 2, while results are presented and discussed in Sects. 3 and 4, respectively. Our conclusions then follow.

2. Observations and data reduction

2.1. VLT observation

PSR  J1028 − 5819 was observed in service mode on February 10 and 11, 2011 with the VLT Antu telescope at the ESO Paranal and the FOcal Reducer/low dispersion Spectrograph (FORS2), which is a multi-mode camera for imaging and long-slit/multi-object spectroscopy (Appenzeller et al. 1998). We used FORS2 used in imaging mode equipped with its standard MIT detector, a mosaic of two 2k  ×  4k CCDs optimised for wavelengths longer than 6000 Å. With its high-resolution colimator, the detector has a pixel scale of 0.″\hbox{$\farcs$}125 (2  ×  2 binning), which corresponds to a projected field-of-view (FOV) of 4.′15×4.′15\hbox{$\farcm15 \times 4\farcm15$} over the CCD mosaic. However, owing to vignetting, the effective sky coverage is smaller, and is larger for the upper CCD chip.

Observations were performed with the standard low-gain and fast read-out mode, with the high-resolution collimator, and through the high-throughput bHIGH (λ = 4400 Å; Δλ = 1035 Å) and vHIGH (λ = 5570 Å; Δλ = 1235 Å) filters. To allow for cosmic ray removal, we obtained sequences of short exposures (100 s) for a total integration time of 4800 s in both the bHIGH and vHIGH filters. We set such a short exposure time to minimise the saturation of a bright V = 13.12 star close to the pulsar position, that had been identified in the Digitised Sky Survey (DSS) images. Exposures were taken in dark time and under photometric conditions, as recorded by the ESO ambient condition monitor1, with an airmass of  ~1.2 and image quality of  ~0.″8\hbox{$ 0\farcs8$}, measured directly on the images by fitting the full-width at half maximum (FWHM) of unsaturated field stars.

Data were reduced through the ESO FORS2 pipeline2 for bias subtraction and flat-field correction. Per each band, we aligned and average-stacked the reduced science images using the swarp programme (Bertin et al. 2002), applying a 3σ filter to the single pixel average to filter out residual hot and cold pixels and cosmic ray hits. Since all exposures were taken with sub-arcsec image quality, we did not perform any selection prior to the image stacking. We applied the photometric calibration by using the extinction-corrected night zero points computed by the FORS2 pipeline, which are available through the instrument data quality control database3.

2.2. Swift observations

PSR J1028 − 5819 was pointed by the Swift/XRT on November 23rd 2008, during a Target of Opportunity (ToO) observation (Obs ID 00031298001) requested soon after the pulsar detection by Fermi. The XRT (Burrows et al. 2005) is a grazing incidence Wolter I telescope equipped with an X-ray CCD imaging spectrometer, yielding a 110 cm2 effective area at 1.5 keV, a 23.′6×23.′6\hbox{$23\farcm6 \times 23\farcm6$} FOV (2.″\hbox{$\farcs$}36/pixel), and an half energy width of 15′′ in the 0.2–10 keV energy range. The integration time was 9.6 ks (7.9 ks accounting for the exposure map correction). The observations were performed in PC mode. Data reduction and analysis were performed as in Marelli et al. (2011) but using the newest version (6.11) of the HEAsoft software package4. On the same visit, the PSR J1028 − 5819 field was also observed in the near-UV with the Ultra-Violet Optical Telescope (UVOT) on board Swift. The UVOT (Roming et al. 2005) is a 30 cm, f/12.7 Ritchey-Chrétien telescope using a microchannel-intensified CCD detector that operates in photon counting mode, covering the 1600–8000 Å range over a 17′ × 17′ FOV (0.″\hbox{$\farcs$}5/pixel). Observations were performed with the UVM2 (λ = 2246 Å; Δλ = 498 Å) and the UVW2 (λ = 1928 Å; Δλ = 657 Å) filters (Poole et al. 2008) for a total integration time of 4115 s and 5701 s, respectively. We retrieved the data from the Swift science data archive and combined the single exposures to obtain the best signal-to-noise ratio (S/N). We used the UVOT data analysis software available in HEAsoft and the version 20110131 of the UVOT calibration files.

2.3. Suzaku observations

The field of PSR J1028–5819 was also observed by Suzaku (Mitsuda et al. 2007) on July 22, 2009. The observation (Obs ID 504045010) was performed with the pulsar located at the nominal aim-point position of the X-ray Imaging Spectrometer (XIS; Koyama et al. 2007). The XIS consists of three front-illuminated (FI) CCD cameras (XIS0, XIS2, and XIS3), and one that is back-illuminated (BI; XIS1). Each CCD camera is combined with a single X-ray telescope (Serlemitsos et al. 2007) and has a single chip, yielding a 330 cm2 (FI) or 370 cm2 (BI) effective area at 1.5 keV, 18′ × 18′ field of view (1.″05/\hbox{$1\farcs05/$}pixel) and a 1.′8\hbox{$1\farcm8$} spatial resolution in the 0.2–12 keV energy range. One of the front-illuminated CCDs, XIS2, was unavailable at the time of our observation, since it suffered from a damage on 2006 November 9 and has been unusable since then. The observations were performed in full-frame mode.

We downloaded the data from the public Suzaku science data archive through the DARTS interface5. Data reduction and analysis of the XIS was performed with HEAsoft following the procedures described in the Suzaku ABC Guide6 version 2. We used the task xispi to convert the unfiltered XIS event files to pulse-invariant channels and then xisrepro to obtain cleaned of event files. For each XIS, both 3 × 3 and 5 × 5 edit modes cleaned event data were combined using xis5x5to3x3 and xselect. Following standard practices, we excluded times within 436 s of Suzaku passing through the South Atlantic Anomaly and we also excluded the data when the line of sight was elevated above the Earth’s limb by less than 5°, or is less than 20° from the bright-Earth terminator. Moreover, we excluded time windows during which the spacecraft was passing through the low cut-off rigidity of below 6 GV. Finally, we removed hot and flickering pixels using sisclean. The nominal on-source time is 23 ks. With all the aforementioned data selection criteria applied, the resulting total effective integration time was 20.2 ks. For the spectral analysis, the response matrices were generated with xisrmfgen and using ray-tracing simulations with xissimarfgen.

3. Results

3.1. VLT astrometry

The radio position of PSR J1028 − 5819 was obtained through interferometry observations with the ATCA (Keith et al. 2008) and is α=10h28m27.s95\hbox{$\alpha =10^{\rm h} 28^{\rm m} 27\fs95$} (±0.s1\hbox{$\pm 0\fs1$}) and δ=−58◦19′05.″22\hbox{$\delta = -58^\circ 19\arcmin 05\farcs22$} (±1.″5\hbox{$\pm 1\farcs5$}), at MJD = 54 562. Owing to the relatively small difference of  ~3 yrs between the epoch of the reference radio position and that of our VLT observation, we can neglect any additional uncertainty in the pulsar position due to its unknown proper motion. If we assumed a DM pulsar distance of 2.3 kpc and a transverse velocity three times as large as the peak value of the radio pulsar velocity distribution (~400 km s-1; Hobbs et al. 2005), the time difference would correspond to an additional uncertainty (in the radial direction) of  ~0.″\hbox{$\farcs$}3, which is much smaller than the nominal error in the radio position. No more recent (and more accurate) radio position for PSR J1028 − 5819 has been obtained by the Fermi Pulsar Timing Consortium (Smith et al. 2008), with Weltevrede et al. (2010) assuming the original ATCA position of Keith et al. (2008), while the measured Swift/XRT position of the pulsar candidate counterpart, α=10h28m27.s69\hbox{$\alpha =10^{\rm h} 28^{\rm m} 27\fs69$} and δ=−58◦19′05.″16\hbox{$\delta = -58^\circ 19\arcmin 05\farcs16$} (Marelli et al. 2011), has an estimated radial uncertainty of  ~5″. Thus, we assume the ATCA radio position as a reference. We note that Simbad reports a quite different position, α=10h28m39.s8\hbox{$\alpha =10^{\rm h} 28^{\rm m} 39\fs8$} and δ=−58◦17′31.″0\hbox{$\delta = -58^\circ 17\arcmin 31\farcs0$}, which is still that given in the Fermi-LAT Bright Source List (Abdo et al. 2009b).

To register the pulsar position in the FORS2 frames as precisely as possible, we re-computed their astrometric solution, which is, by default, based on the coordinates of the guide star used for the telescope pointing. As a reference, we used stars selected from the Guide Star Catalogue 2 (GSC-2; Lasker et al. 2008). Although the UCAC-3 catalogue (USNO CCD Astrograph Catalog; Zacharias et al. 2010) has a better astrometric accuracy then the GSC-2, it has a cut-off at brighter magnitudes (V ≲ 16) and most of the reference UCAC-3 stars are saturated even in the short-exposures FORS2 images. We selected GSC-2 stars evenly distributed in the field-of-view but far from the CCD edges and the vignetted regions, where geometric distortions appear to be larger than expected. We noticed this effect for the first time in our data set and, apparently, it affects all FORS2 images taken with the high-resolution collimator after 2009 (van den Ancker, priv. comm.). Its origin is still under investigation. We measured the star centroids by means of Gaussian fitting using the Graphical Astronomy and Image Analysis (GAIA) tool7 and used the code astrom8 to compute the pixel-to-sky coordinate transformation using an high-order polynomial, which accounts for the CCD distortions. The root mean square (rms) of the astrometric fits was σr~0.″1\hbox{$\sigma_{\rm r} \sim 0\farcs1$} in the radial direction. To this value, we added in quadrature the uncertainty σtr=0.″1\hbox{$\sigma_{\rm tr}=0\farcs1$} in the registration of the FORS2 image in the GSC2 reference frames, σtr=(3/Ns)σGSC2\hbox{$\sigma_{\rm tr}= \sqrt{(3/N_{\rm s})} \sigma_{\rm GSC2}$}, where σGSC2=0.″3\hbox{$\sigma_{\rm GSC2}=0\farcs3$} is the mean positional error of the GSC2 coordinates and Ns is the number of stars used to compute the astrometric solution (Lattanzi et al. 1997). After accounting for the  ~ 0.″15\hbox{$0\farcs15$} accuracy of the link of the GSC2 coordinates to the International Celestial Reference Frame (ICRF), we thus estimated that the overall (1σ) uncertainty in our FORS2 astrometry is δr~0.″2\hbox{$\delta_{\rm r}\sim0\farcs2$}. This is negligible compared with the absolute error in the reference pulsar position.

thumbnail Fig. 1

The 40″ × 40″ VLT/FORS2 image of PSR J1028 − 5819 in the vHIGH filter (4800 s). North to the top, east to the left. The cross marks the pulsar radio-interferometry position (Keith et al. 2008), while the circle (~5″ radius) marks the position of the candidate X-ray counterpart detected by Swift/XRT (Abdo et al. 2009a; Marelli et al. 2011). Stars close to the radio position are labelled (A–D). The bright star A (V = 13.12) is saturated in the FORS2 image. The intensity scale in the image has been stretched to help visualise the fainter stars (B–D).

3.2. The search for the pulsar’s optical counterpart

The computed PSR J1028 − 5819 position is shown in Fig. 1, overlaid on the FORS2 V-band image. As mentioned in Sect. 2.1, it falls  ≲ 4″ from a bright (V = 13.12) field star (star A) west of it. Unfortunately, since its halo extends to  ≲ 1″ from the pulsar position we could not mask star A without the risk of severely hampering our observations. Since the angular separation between star A’s halo and the pulsar is indeed compatible with the telescope pointing accuracy and smaller than the absolute uncertainty in the pulsar’s position (see above), this could have been accidentally covered by one of the FORS2 occulting bars (~12.″5\hbox{$12\farcs5$} width with the high-resolution collimator). Moreover, star A is part of a triplet (with star B and C) of relatively bright stars that was not resolved in the DSS-2 images, the only ones available for a quick look of the pulsar field, owing to their coarse spatial resolution. This would have caused the mask to be put even closer to the pulsar position, thus increasing the risk of an accidental occultation of our target. We note that the northern star of the triplet (star B) overlaps the pulsar position. However, its relatively high flux (B ~ 20.3 ± 0.1; V ~ 17.5 ± 0.1) obviously rules it out as a possible candidate counterpart to the pulsar (e.g., Mignani 2011). The same argument also applies to the southern star of the triplet (star C), which is of comparable brightness (B = 20.7 ± 0.1; V ~ 17.5 ± 0.1).

A fourth, fainter star (star D) is visible in the V-band image at  ~2″ from the nominal radio pulsar position, superimposed on the halo of the much brighter star A, while in the B-band image it cannot be clearly resolved, partially because of the slightly poorer image quality. The offset is comparable with the 1σ uncertainty in the pulsar position, which, in principle, makes it a possible candidate counterpart. We computed the flux of star D through customised aperture photometry using a small aperture of four pixel radius (0.″\hbox{$\farcs$}5) to minimise the contribution of the bright star A’s halo and optimise the S/N. We sampled the local background within circular areas of the same radius as the photometry aperture placed below and above the star D’s position. We then computed the aperture correction using the growth curves of a few unsaturated field stars. The computed magnitudes are B ~ 22.5 ± 0.2 and V ~ 20.5 ± 0.2, where the photometric errors are dominated by the uncertainty in the subtraction of the background of the bright star A’s halo, which also makes it an improbable candidate pulsar counterpart.

The positions and fluxes of all stars detected close to the PSR J1028 − 5819 position are summarised in Table 1. No other object is clearly visible at, or close to, the radio pulsar position. Although our images were taken under sub-arcsec seeing conditions and with short exposure times to minimise the flux contamination of star A, its bright halo still significantly affects the background close to the pulsar position, hampering the detection of its faint counterpart. We tried to subtract star A using the model image point spread function (PSF), computed by fitting the intensity profile of several unsaturated field stars. To this aim, we used the tools daofind, nstar, peak, and substar available in the IRAF package daophot. Since star A is partially saturated, this procedure obviously leaves some residuals at the PSF core, but can account for the subtraction of most of the PSF and its wings. Similarly, we also PSF-subtracted the stars B, C, and D.

Table 1

Coordinates and fluxes of the stars detected close to the PSR J1028 − 5819 position (see Fig. 1).

thumbnail Fig. 2

Left panel: Swift/XRT image (20′ × 20′) of the field of PSR J1028–5819 in the 0.3–10 keV energy range. The source highlighted by the green circle is the pulsar’s X-ray counterpart (Abdo et al. 2009a; Marelli et al. 2011). Right panel: zoom-in of the Swift/XRT image (2.′5×2.′5)\hbox{$2\farcm5 \times 2\farcm5)$} on the source position. The red cross marks the ATCA position of PSR J1028–5819 (Keith et al. 2008).

We did not detect any excess signal in the PSF-subtracted images, which can be associated with emission from a point-like source at the pulsar position. Thus, we concluded that PSR J1028 − 5819 was not detected in our VLT images. We computed the optical flux upper limits in the PSF-subtracted B and V bands images from the rms of the background sampled in the pulsar radio error box within cells of 7 × 7 pixels, i.e. of size equal to the computed image FWHM, and we applied the computed aperture correction. We then derived 3σ upper limits of V ~ 25.3 and B ~ 25.4. Unfortunately, these limits are affected by the residuals of the PSF subtraction of the four stars, to which star A and B are contributing the most, and which increase the local sky background by a factor of four with respect to other regions in the field, where the limiting magnitudes are about 1.5 mag deeper.

3.3. The pulsar’s X-ray counterpart: Swift observation

It is interesting to compare our derived optical flux upper limits on PSR J1028 − 5819, the deepest obtained so far, with the extrapolation of the best-fit to the X-ray spectrum of the pulsar’s counterpart. Unfortunately, this is poorly constrained by the available Swift/XRT observation (Fig. 2), owing to the short integration time (7.9 ks) and the low count statistics. We extracted all counts within a 20′′ radius using a similarly sized background region. This yields (2.41 ± 0.69) × 10-3 counts s-1 in the 0.3–10 keV energy range. We reanalysed the XRT spectral data using the newest version (12.7) of the xspec fitting package (Arnaud et al. 1996). As in Marelli et al. (2011), we found a satisfactory fit with a PL of fixed photon index ΓX = 2, assuming a fixed hydrogen column-density NH ~ 5 × 1021 cm-2 inferred from the Galactic value along the line of sight using the Heasarc nH tool9 scaled to the pulsar distance of 2.3 kpc assuming a constant distribution. We note that this value of the NH is about a factor of three lower than assumed by Abdo et al. (2009a), NH = 1.59 × 1022 cm-2, whose determination, however, is not explained. For these spectral parameters, the Swift/XRT spectrum corresponds to an unabsorbed 0.5–10 keV X-ray flux FX = (1.5 ± 0.5) × 10-13 erg cm-2 s-1.

3.4. The pulsar’s X-ray identification revisited

The uncertainty in the pulsar’s X-ray spectrum obviously affects the comparison with our optical flux measurements. Moreover, the positional coincidence between the Swift/XRT error circle of the pulsar candidate counterpart and star A (Fig. 1) suggests that, in principle, this might contribute to the observed X-ray emission, which would make the characterisation of the pulsar X-ray spectrum even more uncertain. As a most extreme possibility, it cannot be ruled out that the Swift/XRT source is, actually, the X-ray counterpart to star A and not to the pulsar. The star was clearly detected by the Swift/UVOT (see Fig. 3) but, admittedly, its possible contribution to the emission of the X-ray source was not accounted for by Marelli et al. (2011). Thus, we tried to evaluate such a contribution by determining the star’s spectral type via multi-band photometry.

For the Swift/UVOT observations, we used the ftool uvotsource to determine the count rate of the star in both the UVM2 and UVW2 filters. We used a 2′′ aperture to minimise the contamination from the nearby star north-west of it. The count rate was aperture-corrected to a radius of 5′′ to apply the count-rate to magnitude conversion using the zero pointsof the Swift/UVOT calibration (Breeveld et al. 2010, 2011). The computed magnitudes (in the Vega system) are mUVM2 = 16.29 ± 0.10 and mUVW2 = 16.51 ± 0.11, where the associated errors are purely statistical (systematic errors are  ~ 0.03 mag). The star’s optical magnitudes are RF = 12.87 and V = 13.12, according to the GSC-2 (Lasker et al. 2008), while the near-infrared magnitudes, derived from 2MASS (Skrutskie et al. 2006), are J = 12.38, H = 12.16, and Ks = 12.09. Unfortunately, the UVOT, DSS, and 2MASS, images have insufficient angular resolution to resolve the triplet of stars around star A, while the star  ≈ 7″ north-west of it (Fig. 1) is clearly resolved in all of them (see, e.g. Fig. 3). Thus, the above magnitudes must be interpreted with due care. However, from our VLT data we see that star B and C are at least a factor of  ~ 50 fainter than star A in both the B and V bands. Thus, their contribution only marginally affects the estimate of the star colours.

thumbnail Fig. 3

Same field as in Fig. 1 (40″ × 40″) but obtained by the Swift/UVOT through the UVM2 filter (4115 s). North to the top, east to the left. The group of stars around star A (see Fig. 1) is unresolved at the UVOT spatial resolution.

thumbnail Fig. 4

Left panel: Suzaku/XIS image (25′ × 25′) of the field of PSR J1028–5819 in the 0.2–10 keV energy range. The pulsar’s X-ray counterpart, also detected by the Swift/XRT (Abdo et al. 2009a; Marelli et al. 2011; Fig. 2), is at the centre of the field-of-view. Right panel: zoom-in of the Suzaku/XIS image (6.′5×6.′5)\hbox{$6\farcm5 \times 6\farcm5)$} at the source position. As in Fig. 2, the red cross marks the ATCA position of PSR J1028–5819 (Keith et al. 2008).

The colours of star A are compatible with those of an early F-type main-sequence star, assuming an interstellar extinction corresponding to an E(B − V) ~ 0.09. According to the relation of Predhel & Schmitt (1995), this value would correspond to an NH ~ 0.5 × 1021 cm-2, i.e. lower than estimated by scaling the Galactic value along the line of sight for the PSR J1028 − 5819 distance, suggesting that star A is, possibly, in the foreground to the pulsar, which would be consistent with a distance of  ≈ 1 kpc derived from its distance modulus. Usually, F-type stars have an X-ray-to-optical flux ratio FX/FV ≈ 10-3–10-2 (e.g. Krautter et al. 1999), which might yield an X-ray flux comparable with that of the Swift/XRT source. Thus, it is possible that this is, indeed, the X-ray counterpart to Star A and not to the pulsar. However, any conclusion may still be altered by a more accurate characterisation of the X-ray source spectrum and flux. To achieve this aim, we used an archival Suzaku observation (see Sect. 2.3).

3.5. The pulsar’s X-ray counterpart: Suzaku observation

The Suzaku image of the PSR J1028–5819 field obtained combining the data from the three XIS cameras is shown in Fig. 4. Using the sliding-cells source detection algorithm in the ximage package, a source is detected at the centre of the XIS FOV with a S/N of 18 in the 0.2–10 keV energy range. The source best-fit position, α=10h28m28.s7\hbox{$\alpha = 10^{\rm h}28^{\rm m} 28\fs7$}, δ=−58°18′59.″3\hbox{$\delta =-58\degr18\arcmin 59\farcs3$} (J2000), is consistent, within the absolute astrometry accuracy of Suzaku (~20″; Uchiyama et al. 2008), with both the ATCA coordinates of PSR J1028–5819 and the coordinates of the source detected by Swift.

thumbnail Fig. 5

Upper panel: Suzaku/XIS spectrum of the candidate PSR J1028 − 5819 X-ray counterpart obtained using data from the three XIS CCDs. The background level is marked in blue and the best-fit PL model in red. Lower panel: fit residuals. The residuals show some (rather small) structures. In the 1.5–2 keV interval, they are probably associated with the Si K-edge remain, while around 5 keV they likely arise from the different responses of the BI and FI detectors around this energy. However, we note that most residuals are within 1σ from the best-fit PL model.

Table 2

Summary of the Suzaku/XIS spectral analysis.

We extracted the source counts from each XIS detector using a circular region of 0.′7\hbox{$0\farcm7$} radius centred on the best-fit ximage position. This comparatively small region, corresponding to  ~50% of the encircled energy fraction, was chosen to improve the S/N of the spectra and reduce the possible contamination from the diffuse emission apparent in Fig. 4. The background was extracted from annuli with radii 3′ and 5′ centred on the X-ray source. Accounting for the counts of all the XIS detectors, the 0.5–10 keV net source count rate is (2.7 ± 0.3) × 10-3 counts s-1, which is not a very reliable statistical measurement for a detailed spectral analysis. To increase the quality of the photon statistics, we merged the source spectral files and response matrices using addascaspec. The 3-XIS spectrum was binned so that each bin contained a minimum of 15 counts. As done for the Swift/XRT spectrum, we performed our spectral fits using the XSPEC fitting package (version 12.7). The fit to an absorbed PL model (Fig. 5) yields a spectral index ΓX = 1.7 ± 0.2 and an absorption that is poorly constrained at NH < 1021 cm-2 at (1σ), with a reduced χν2=1.09\hbox{$\chi^2_\nu = 1.09$} for 21 degrees of freedom. This corresponds to an unabsorbed 0.5–10 keV X-ray flux FX = (1.5 ± 0.2) × 10-13 erg cm-2 s-1, which is strongly consistent with the value derived by Marelli et al. (2011) from the Swift/XRT observation across a similar energy range. We note that our upper limit to the NH is lower than the value inferred by scaling the Galactic value along the line of sight (Marelli et al. 2011), although it is not incompatible with the latter once the associated uncertainties are taken into account. We also tried other single-component models, such as a blackbody (BB), a bremsstrahlung, and a Raymond-Smith plasma (Raymond & Smith 1977). However, they all give higher χ2 values than the PL, although the fits obtained using the bremsstrahlung and Raymond-Smith plasma models are only slightly worse than obtained using the PL (see Table 2 for a summary). To improve our results, we tried to fit the joint Swift/XRT and Suzaku/XIS spectra, but owing to the low quality statistics of the former observation our results did not change significantly.

Our spectral fits confirm the qualitative results of Marelli et al. (2011), which were based on a lower quality statistics. In particular, the good fit with a PL (χν2=1.09\hbox{$\chi^2_\nu = 1.09$}) suggests that the spectrum of the X-ray source is non-thermal. Thus, it would not be produced by coronal emission from a main-sequence star, such as Star A. This would confirm that the X-ray source detected by both the Swift/XRT and Suzaku is, indeed, the pulsar’s X-ray counterpart. Owing to the still limited quality of the statistics, we were unable to attempt spectral fits with two-component models, such as a PL+BB. Thus, we cannot rule out that star A might, indeed, contribute at some level to the observed X-ray emission. Deeper observations with XMM-Newton will allow one to obtain a more precise characterisation of the X-ray spectrum of the PSR J1028 − 5819 counterpart, while optical spectroscopy of star A will provide a far more precise classification of its spectral type, allowing one to better quantify its expected X-ray flux. In addition, high-resolution Chandra images will allow one to spatially resolve a possible X-ray source associated with star A from the pulsar’s X-ray counterpart and secure its identification through the detection of X-ray pulsations10.

We also inspected the Suzaku data to search for possible evidence of a PWN associated with PSR J1028 − 5819. As mentioned above, diffuse emission on a scale of a few arcminutes is observed around the pulsar’s X-ray counterpart (Fig. 4), with an apparent elongation in the south-west direction. The count rate in this region is (5.3 ± 0.5) × 10-4 counts s-1 arcmin-2 in the 0.2–10 keV energy range, computed after summing the counts collected with the three XIS detectors. However, since each XRT/XIS detector has a different position-dependent and broad PSF, with a complicated (roughly cross-shaped) and asymmetric profile (see, e.g. Sugizaki et al. 2009), it is difficult to determine whether the diffuse emission observed around the pulsar’s X-ray counterpart is real. As a test, we compared its surface brightness profile with that of a reference X-ray source. Since no suitable sources were detected in the PSR J1028 − 5819 field-of-view (Fig. 4), we used as a reference an X-ray source detected in another Suzaku observation (Obs ID 804017020) and identified this with the ms-pulsar PSR  J1231 − 1411. Since both sources were observed on-axis, our comparison is unaffected by the position dependance of the XIS PSF. In both cases, we selected photons in the 0.3–10 keV energy range in all the three XIS detectors and used the xissim tool to simulate and subtract the background, accounting for the detector vignetting. Then, we extracted the radial intensity profiles, averaged over a 360° angle, for both the PSR J1028 − 5819 counterpart and the reference source. Finally, we normalised the counts of the reference source, computed within a radial distance bin of 10″, to that of the PSR J1028 − 5819 counterpart. As seen from Fig. 6, the latter seems to display an excess of counts with respect to the reference X-ray source at radial distances  ~ 80″–200″, although this excess is much less evident at larger distances. This, coupled with the low count-rate of the diffuse emission, makes it difficult to determine whether the source detection is real, hence possibly associated with either a PWN or an unresolved field source, and to quantify its contribution to the observed emission of the PSR J1028 − 5819 X-ray counterpart, in particular at angular distances smaller than 80″. Chandra observations will be crucial to confirm the presence of the diffuse emission observed by Suzaku. If the source is real, future measurements of the PSR J1028 − 5819 radio proper motion will, then, allow one to search for its possible alignment with the semi-major axis of the candidate PWN.

thumbnail Fig. 6

Surface brightness profiles of the PSR J1028 − 5819 X-ray counterpart (black) detected in the Suzaku/XIS data and of a reference X-ray source (red). The excess with respect to the reference source differs from zero at the  ~ 5σ level (accounting for statistical errors only). This suggests that diffuse X-ray emission surrounds PSR J1028 − 5819 on angular scales of a few arcminutes.

4. Discussion

4.1. The pulsar luminosity

We used our magnitude upper limits in the B-band11 to compute the corresponding extinction-corrected optical flux and luminosity of PSR J1028 − 5819. As a reference for the interstellar extinction correction, we assumed the upper limit to the hydrogen column density (NH = 1021 cm-2) derived from the fits to the Suzaku X-ray spectrum. According to the relation of Predhel & Schmitt (1995), this corresponds to an interstellar extinction AV ~ 0.55. Using the extinction coefficients of Fitzpatrick (1999), the extinction-corrected B-band upper limit is, thus, FB ≲ 9 × 10-16 erg cm-2 s-1. If we assumed for the pulsar the unabsorbed Suzaku X-ray flux derived from the best-fit PL model, FX = (1.5 ± 0.2) × 10-13 erg cm-2 s-1, our extinction-corrected B-band upper limit would correspond to an X-ray-to-optical flux ratio FX/FB ≳ 150, which is too low to independently suggest that the X-ray source detected by Swift and Suzaku is indeed the PSR J1028 − 5819 counterpart. We computed the upper limit to the pulsar optical luminosity using as a reference its DM-based distance of  ~ 2.3 kpc (Keith et al. 2008). We note that while the uncertainty in the measured DM is negligible (DM = 96.525 ± 0.002 cm-3 pc), the Galactic electron density model of Cordes & Lazio (2002) is known to have an average uncertainty of  ~ 30%, which, however, can be much larger for individual objects. As in the first Fermi-LAT catalogue of γ-ray pulsars (Abdo et al. 2010a), we assumed the typical 30% uncertainty on the nominal DM-based distance of PSR J1028 − 5819. This corresponds to an uncertainty of 0.7 kpc, which we assume in the following discussion. The B-band flux upper limit corresponds to an optical luminosity LB ≲ 9.2    ×    1029 erg s-1 for PSR J1028 − 5819, after accounting for the assumed distance uncertainty. Given the pulsar spin-down age of  ~ 90 kyr, we can assume that its, yet undetected, optical emission be purely non-thermal, as normally observed in the youngest pulsars (see, e.g. Mignani 2011). Thus, for the pulsar’s spin-down power, Ė ~ 8.43 × 1035 erg s-1, our luminosity upper limit implies an optical emission efficiency ηB ≡ LB/Ė ≲ 1.09    ×    10-6. Both limits are consistent with the optical luminosity and efficiency computed for most pulsars older than  ~ 0.1 Myr (Zharikov et al. 2006) and confirm an apparent turnover in the pulsar optical-emission properties. Similarly, the pulsar X-ray flux derived from the best-fit PL mode would correspond to an X-ray luminosity LX = (9.5 ± 5.9) × 1031 erg s-1, after accounting for the X-ray flux and pulsar distance uncertainties. This value would imply an X-ray emission efficiency ηX ≡ LX/Ė = (1.13 ± 0.70) × 10-4, for an assumed beaming factor f = 1. The derived X-ray efficiency would be consistent, within the large scatter, with the trend observed for the Fermi pulsars (Marelli et al. 2011).

4.2. The pulsar multi-wavelength spectrum

We compared the extinction-corrected B and V-band spectral flux upper limits with the extrapolations in the optical domain of the X-ray and γ-ray spectra of PSR J1028 − 5819. For the X-ray spectrum, we used as a reference our best-fit PL model with spectral index ΓX = 1.7 ± 0.2, while for the γ-ray spectrum we assumed a PL with exponential cutoff, spectral index Γγ = 1.25 ± 0.17, and cutoff energy Ec = 1.9 ± 0.5 GeV, obtained from a reanalysis of the Fermi data (see also Marelli et al. 2011). The multi-wavelength spectral energy distribution (SED) of PSR J1028 − 5819 is shown in Fig. 7, where we accounted for the 1σ uncertainty in the extrapolations of the X and γ-ray PL spectra. The optical spectral flux upper limits are clearly above the extrapolation of the γ-ray PL, which does not rule out the optical emission of PSR J1028 − 5819 being associated with the low-energy tail of the γ-ray spectrum, as suggested for most γ-ray pulsars identified in the optical (Durant et al. 2011). This would imply that there is a break in the optical-to-X-ray spectrum, as observed in most pulsars (see, e.g. Mignani et al. 2010). In this case, the observed optical flux would be B = 28 − 34, a range only marginally reachable by current 8 m-class telescopes. On the other hand, the optical upper limits are also compatible with the extrapolation of the X-ray PL, which does not allow one yet to confirm the presence of a possible spectral break in the optical-to-X-ray spectrum. In general, the SED suggests that no single PL can fit the optical-to-γ-ray spectrum of the pulsar, with at least one spectral break required between the X and γ-ray energy ranges. This trend has also been observed in other Fermi pulsars (e.g., Mignani et al. 2011), and probably related to complex particle energy and density distributions in the neutron star magnetosphere.

thumbnail Fig. 7

Upper limits to the extinction-corrected optical fluxes of PSR J1028 − 5819 compared with the low-energy extrapolations of the X-ray and γ-ray PLs that most closely fits the Suzaku and Fermi (Abdo et al. 2009a) data (black solid lines). The yellow and blue-shaded areas indicate the 1σ uncertainty in the extrapolations of the X and γ-ray PLs, respectively. The plotted optical flux upper limits are corrected for interstellar extinction based upon an NH = 1021 cm-2.

4.3. The candidate PWN

Finally, we verified whether the diffuse emission marginally detected in the Suzaku data (Sect. 3.5) would be compatible, if it were indeed a PWN, with the PSR J1028 − 5819 energetics and distance. PWNe have been detected in the X-rays around a number of young and energetic pulsars (see, e.g. Kargaltsev & Pavlov 2010, for a recent compilation), with luminosities equal to  ~ 10-4–10-2 of the pulsar spin-down power Ė. In the case of PSR J1028 − 5819, one would then expect an X-ray luminosity LX,PWN ~ 8.5 × 1031 erg s-1 for its PWN, assuming LX,PWN/Ė = 10-4. At the PSR J1028 − 5819 distance of  ~ 2.3 ± 0.7 kpc, this would correspond to an integrated flux of FX,PWN ~ (1.3 ± 0.8) × 10-13 erg cm-2 s-1 for the PWN. By using the WebPIMMS tool12 assuming a PWN spectral index ΓX,PWN = 2 and a hydrogen column density NH ~ 1021 cm-2, this would correspond to a count-rate of (8 ± 5) × 10-3 counts s-1 in the 0.2–10 keV energy range. Following, for instance Bamba et al. (2010), a PWN around a  ~ 100 kyr old pulsar would have an angular extent of  ~ 11.′5\hbox{$11\farcm5$}–21.′5\hbox{$21\farcm5$} at a distance of 2.3 ± 0.7 kpc, i.e. comparable to that of the diffuse emission detected in the Suzaku data. For the expected PWN extension, we would then derive an expected count rate of  ~ (0.2 − 3) × 10-5 counts s-1 arcmin-2. This value is up to a factor of 30 lower than the actual count-rate measured for the diffuse emission around PSR J1028 − 5819. However, it is not incompatible with it if one assumes, e.g. either a correspondingly larger LX,PWN/Ė, which would still be within the observed range for the known PWNe, or a non-uniform PWN surface brightness. Thus, it is indeed plausible that the diffuse emission detected by Suzaku is associated with the PSR J1028 − 5819 PWN.

5. Summary and conclusions

We have performed the first deep observations of the Fermi pulsar PSR J1028 − 5819 with the VLT. Unfortunately, owing to a bright star F-type star (star A; V = 13.12)  ~ 4″ away from the pulsar radio position (Keith et al. 2008) we could only set limits of V ~ 25.3 and B ~ 25.4 on its optical flux, which are admittedly affected by both the residuals after the subtraction of star A’s PSF and the coarse accuracy of the radio coordinates. These limits correspond to an optical luminosity LB ≲ 9.2    ×    1029 erg s-1 and emission efficiency ηB ≲ 1.09    ×    10-6, which are consistent with the pulsar spin-down age of  ~ 90 kyr. We have also analysed archival Suzaku observations of PSR J1028 − 5819, which provide a more secure detection of the pulsar’s X-ray candidate counterpart and support its identification as a pulsar based on a more accurate characterisation of the X-ray spectrum, which is best-fit by a PL with spectral index ΓX = 1.7 ± 0.2. The inferred pulsar’s X-ray efficiency (ηX ~ 1.13 × 10-4) would be consistent with the trend observed for the Fermi pulsars (Marelli et al. 2011). The pulsar multi-wavelength spectrum cannot be fit by a single PL, implying that there is at least one spectral break in the optical-to-γ-ray energy range, as observed in other Fermi pulsars. We have also found tentative evidence of X-ray diffuse emission around PSR J1028 − 5819, with a surface brightness and angular extent consistent with the observed X-ray luminosity and size of known PWNe. More observations are required to complete the multi-wavelength study of PSR J1028 − 5819 at low energies. On the optical side, high-spatial resolution observations with the HST, possibly in the near-UV to exploit the low interstellar extinction along the line of sight, that would be coupled with a much more accurate radio position, are needed to resolve the pulsar emission against the PSF wings of star A. Moreover, optical spectroscopy of star A is needed to provide a more precise classification of its spectral type, allowing one to more accurately predict its X-ray flux. On the X-ray side, follow-up observations with XMM-Newton and Chandra are needed to obtain a more accurate characterisation of the pulsar’s X-ray spectrum, accounting for any possible contribution from star A, and detect X-ray pulsations, which will confirm the proposed identification. Chandra observations will also be crucial in confirming the existence of the diffuse emission marginally detected by Suzaku and study its spectrum and morphology.

10

We note that the frame times of the Swift/XRT in PC mode and the Suzaku/XIS in full-frame mode are 2.507 s and 8 s, respectively, which makes it impossible to search for pulsations at the PSR J1028 − 5819 91.4 ms period.

11

For comparison with the luminosities quoted in Zharikov et al. (2006), which are computed in the B band.

Acknowledgments

A key contributor to the FORS2 pipeline, Carlo Izzo (ESO), prematurely passed away on June 23, 2011 after fighting courageously a short illness. Thanks to Carlo’s work, the results from many FORS2 observations were published, making his legacy forever lasting. R.P.M. thanks Paul Ray

for checking the PSR J1028 − 5819 coordinates and Mario van den Ancker and Sabine Moheler (ESO) for investigating the FORS2 geometric distortions. The authors thank Fabio Mattana for useful comments. This research made use of data obtained from Data ARchives and Transmission System (DARTS), provided by Center for Science-satellite Operation and Data Archives (C-SODA) at ISAS/JAXA. P.E. acknowledges financial support from the Autonomous Region of Sardinia through a research grant under the programme PO Sardegna FSE 2007 − 2013, L.R. 7/2007 “Promoting scientific research and innovation technology in Sardinia”. This research has made use of data obtained from the High Energy Astrophysics Science Archive Research Center (HEASARC) and the Leicester Database and Archive Service (LEDAS), provided by NASA’s Goddard Space Flight Center and the Department of Physics and Astronomy, Leicester University, UK, respectively. S.R.O. acknowledges the support of the UK Space Agency.

References

  1. Abdo, A. A., Ackermann, M., Atwood, W. B., et al. 2009a, ApJ, 695, L72 [NASA ADS] [CrossRef] [Google Scholar]
  2. Abdo, A. A., Ackermann, M., Ajello, M., et al. 2009b, ApJS, 183, 46 [NASA ADS] [CrossRef] [Google Scholar]
  3. Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010a, ApJ, 720, 272 [NASA ADS] [CrossRef] [Google Scholar]
  4. Abdo, A. A., Ackermann, M., Ajello, M., et al. 2010b, ApJS, 187, 460 [NASA ADS] [CrossRef] [Google Scholar]
  5. Abramowski, A., Acero, F., Aharonian, F., et al. 2011, A&A, 525, A46 [Google Scholar]
  6. Ackermann, M., Ajello, M., Baldini, L., et al. 2011, ApJ, 726, 35 [NASA ADS] [CrossRef] [Google Scholar]
  7. Appenzeller, I., Fricke, K., Fürtig, W., et al. 1998, The Messenger, 94, 1 [NASA ADS] [Google Scholar]
  8. Arnaud, K. A. 1996, Astronomical Data Analysis Software and Systems V, eds. G. H. Jacoby, & J. Barnes (San Francisco: ASP), 101, 17 [Google Scholar]
  9. Atwood, W. B., Abdo, A. A., Ackermann, M., et al. 2009, ApJ, 697, 1071 [NASA ADS] [CrossRef] [Google Scholar]
  10. Bamba, A., Anada, T., Dotani, T., et al. 2010, ApJ, 719, L116 [NASA ADS] [CrossRef] [Google Scholar]
  11. Bertin, E., Mellier, Y., Radovich, M., et al. 2002, in ADASS XI, ASP Conf. Ser., 281, 228 [Google Scholar]
  12. Bignami, G. F., & Caraveo, P. A. 1996, ARA&A, 34, 331 [NASA ADS] [CrossRef] [Google Scholar]
  13. Breeveld, A. A., Curran, P. A., Hoversten, E. A., et al. 2010, MNRAS, 406, 1687 [NASA ADS] [Google Scholar]
  14. Breeveld, A. A., Landsman, W., Holland, S. T. et al., 2011, in GRB 2010 Symp., AIP Conf. Ser., in press [arXiv:1102.4717] [Google Scholar]
  15. Burrows, D. N., Hill, J. E., Nousek, J. A., et al. 2005, SSRv, 120, 165 [Google Scholar]
  16. Cordes, J. M., & Lazio, T. J. W. 2002 [arXiv:astro-ph/0207156] [Google Scholar]
  17. Durant, M., Kargaltsev, O., & Pavlov, G. G. 2011, ApJ, 743, 38 [NASA ADS] [CrossRef] [Google Scholar]
  18. Fitzpatrick, E. L. 1999, PASP, 111, 63 [NASA ADS] [CrossRef] [Google Scholar]
  19. Hobbs, G., Lorimer, D. R., Lyne, A. G., & Kramer, M. 2005, MNRAS, 360, 974 [NASA ADS] [CrossRef] [Google Scholar]
  20. Kargaltsev, O., & Pavlov, G. G. 2010, in X-ray Astronomy 2009, AIP Conf. Proc., 1248, 25 [Google Scholar]
  21. Keith, M. J., Johnston, S., Kramer, M., et al. 2008, MNRAS, 389, 1881 [NASA ADS] [CrossRef] [Google Scholar]
  22. Koyama, K., Tsunemi, H., Dotani, T., et al. 2007, PASJ, 59, 23 [Google Scholar]
  23. Krautter, J., Zickgraf, F.-J., Appenzeller, I., et al. 1999, A&A, 350, 743 [NASA ADS] [Google Scholar]
  24. Lasker, B. M., Lattanzi, M. G., McLean, B. J., et al. 2008, AJ, 136, 735 [NASA ADS] [CrossRef] [PubMed] [Google Scholar]
  25. Lattanzi, M. G., Capetti, A., & Macchetto, F. D. 1997, A&A, 318, 997 [NASA ADS] [Google Scholar]
  26. Marelli, M., De Luca, A., & Caraveo, P. A. 2011, ApJ, 733, 82 [NASA ADS] [CrossRef] [Google Scholar]
  27. Mignani, R. P. 2011, AdSpR, 47, 1281 [NASA ADS] [CrossRef] [Google Scholar]
  28. Mignani, R. P., Sartori, A., De Luca, A., et al. 2010, A&A, 515, A110 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  29. Mignani, R. P., Shearer, A., De Luca, A., et al. 2011, A&A, 533, A101 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  30. Mitsuda, K., Bautz, M., Inoue, H., et al. 2007, PASJ, 59, 1 [Google Scholar]
  31. Newberry, M. V. 1991, PASP, 103, 122 [NASA ADS] [CrossRef] [Google Scholar]
  32. Poole, T. S., Breeveld, A. A., Page, M. J., et al. 2008, MNRAS, 383, 627 [NASA ADS] [CrossRef] [Google Scholar]
  33. Predehl, P., & Schmitt, J. H. M. M. 1995, A&A, 293, 889 [NASA ADS] [Google Scholar]
  34. Raymond, J. C., & Smith, B. W. 1977, ApJS, 35, 419 [NASA ADS] [CrossRef] [Google Scholar]
  35. Roming, P. W. A., Kennedy, T. E., Mason, K. O., et al. 2005, Space Sci. Rev., 120, 95 [NASA ADS] [CrossRef] [Google Scholar]
  36. Serlemitsos, P. J., Soong, Y., Chan, K.-W., et al. 2007, PASJ, 59, 9 [Google Scholar]
  37. Skrutskie, M. F., Cutri, R. M., Stiening, R., et al. 2006, AJ, 131, 1163 [NASA ADS] [CrossRef] [Google Scholar]
  38. Smith, D. A., Guillemot, L., Camilo, F., et al. 2008, A&A, 492, 923 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  39. Sugizaki, M., Kamae, T., & Maeda, T. 2009, PASJ, 61, S55 [NASA ADS] [Google Scholar]
  40. Thompson, D. J. 2008, Rep. Prog. Phys., 71, 116901 [NASA ADS] [CrossRef] [Google Scholar]
  41. Uchiyama, Y., Maeda, Y., Ebara, M., et al. 2008, PASJ, 60, 35 [NASA ADS] [Google Scholar]
  42. Weltevrede, P., Johnston, S., Manchester, R. N., et al. 2010, PASP, 27, 64 [Google Scholar]
  43. Zacharias, N., Finch, C., Girard, T., et al. 2010, AJ, 139, 2184 [NASA ADS] [CrossRef] [Google Scholar]
  44. Zharikov, S. V., Shibanov, Yu. A., & Komarova, V. N. 2006, AdSpR, 37, 1979 [NASA ADS] [CrossRef] [Google Scholar]

All Tables

Table 1

Coordinates and fluxes of the stars detected close to the PSR J1028 − 5819 position (see Fig. 1).

In the text
Table 2

Summary of the Suzaku/XIS spectral analysis.

In the text

All Figures

thumbnail Fig. 1

The 40″ × 40″ VLT/FORS2 image of PSR J1028 − 5819 in the vHIGH filter (4800 s). North to the top, east to the left. The cross marks the pulsar radio-interferometry position (Keith et al. 2008), while the circle (~5″ radius) marks the position of the candidate X-ray counterpart detected by Swift/XRT (Abdo et al. 2009a; Marelli et al. 2011). Stars close to the radio position are labelled (A–D). The bright star A (V = 13.12) is saturated in the FORS2 image. The intensity scale in the image has been stretched to help visualise the fainter stars (B–D).
In the text
thumbnail Fig. 2

Left panel: Swift/XRT image (20′ × 20′) of the field of PSR J1028–5819 in the 0.3–10 keV energy range. The source highlighted by the green circle is the pulsar’s X-ray counterpart (Abdo et al. 2009a; Marelli et al. 2011). Right panel: zoom-in of the Swift/XRT image (2.′5×2.′5)\hbox{$2\farcm5 \times 2\farcm5)$} on the source position. The red cross marks the ATCA position of PSR J1028–5819 (Keith et al. 2008).
In the text
thumbnail Fig. 3

Same field as in Fig. 1 (40″ × 40″) but obtained by the Swift/UVOT through the UVM2 filter (4115 s). North to the top, east to the left. The group of stars around star A (see Fig. 1) is unresolved at the UVOT spatial resolution.
In the text
thumbnail Fig. 4

Left panel: Suzaku/XIS image (25′ × 25′) of the field of PSR J1028–5819 in the 0.2–10 keV energy range. The pulsar’s X-ray counterpart, also detected by the Swift/XRT (Abdo et al. 2009a; Marelli et al. 2011; Fig. 2), is at the centre of the field-of-view. Right panel: zoom-in of the Suzaku/XIS image (6.′5×6.′5)\hbox{$6\farcm5 \times 6\farcm5)$} at the source position. As in Fig. 2, the red cross marks the ATCA position of PSR J1028–5819 (Keith et al. 2008).
In the text
thumbnail Fig. 5

Upper panel: Suzaku/XIS spectrum of the candidate PSR J1028 − 5819 X-ray counterpart obtained using data from the three XIS CCDs. The background level is marked in blue and the best-fit PL model in red. Lower panel: fit residuals. The residuals show some (rather small) structures. In the 1.5–2 keV interval, they are probably associated with the Si K-edge remain, while around 5 keV they likely arise from the different responses of the BI and FI detectors around this energy. However, we note that most residuals are within 1σ from the best-fit PL model.

In the text
thumbnail Fig. 6

Surface brightness profiles of the PSR J1028 − 5819 X-ray counterpart (black) detected in the Suzaku/XIS data and of a reference X-ray source (red). The excess with respect to the reference source differs from zero at the  ~ 5σ level (accounting for statistical errors only). This suggests that diffuse X-ray emission surrounds PSR J1028 − 5819 on angular scales of a few arcminutes.
In the text
thumbnail Fig. 7

Upper limits to the extinction-corrected optical fluxes of PSR J1028 − 5819 compared with the low-energy extrapolations of the X-ray and γ-ray PLs that most closely fits the Suzaku and Fermi (Abdo et al. 2009a) data (black solid lines). The yellow and blue-shaded areas indicate the 1σ uncertainty in the extrapolations of the X and γ-ray PLs, respectively. The plotted optical flux upper limits are corrected for interstellar extinction based upon an NH = 1021 cm-2.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.