Australia Telescope National Facility

Radio Images of the Globular Cluster 47 Tucanae
D. McConnell ,
R. Deacon ,
J.G. Ables
, PASA, 18 (2), in press.

Next Section: Acknowledgements
Title/Abstract Page: Radio Images of the
Previous Section: Image Analysis and Results
Contents Page: Volume 18, Number 2

Discussion

Of the 15 pulsars with positions well determined by pulsar timing (Freire et al. 2000), 11 correspond to nine sources in the ATCA image at 1.4 GHz. Pulsars G and I are separated by less than 0

$.\!\!^{\prime\prime}$ 2 and appear as a single point source. Pulsar T lies 2

$.\!\!^{\prime\prime}$ 8 from the G/I pair. Pulsars L and O are separated by $\sim$ 5

$^{\prime \prime }$ and appear as a single extended source. Thus nine discrete sources in the image account for 11 of the pulsars. Pulsars M, H and N correspond to no emission above the 3 $\sigma$ level of 54 $\mu\mbox{Jy}$ and are reported by Camilo et al. (2000) to be faint -- 70 $\pm$ 20 $\mu\mbox{Jy}$ , 90 $\pm$ 20 $\mu\mbox{Jy}$ and 30 $\pm$ 10 $\mu\mbox{Jy}$ respectively. The core region of the cluster (Figure 1) shows five sources in addition to the nine identified with pulsars. Based on the source density discussed in section 3 above we expect 3.1 $\pm$ 1.8 sources above the 5 $\sigma$ detection threshhold, consistent with all five being unrelated to the globular cluster. As noted in section 3, the spectral index measurements in many cases have large uncertainties. However, they are consistent with all pulsar related sources having spectral index

$-3 < \alpha < -2$ , typical of pulsar emission. Three of the numbered sources (142, 176 and 178) have spectral index range

$-0.8 < \alpha < 1.2$ , characteristic of extragalactic radio sources. The remaining two (183, 186) have steeper spectra and may correspond to any of the seven pulsars yet to be located with pulse timing. Two sources, numbered 2 and 5 of Table 2, (MA2000) were identified as possible pulsars based on their spectral index ( $-2.4\pm2.0$ and $-2.4\pm0.7$ respectively) and their apparent variability. Source 2 lies 248

$^{\prime \prime }$ ( $\sim10$ core radii) from the cluster centre. The improved sensitivity of the new image gives a more precisely determined spectral index of $-0.1\pm0.9$ , weakening the case for it being a pulsar. Source 5 can now be identified with the pulsar 47 Tuc J. A remarkable feature of the 47 Tuc pulsars is their strong scintillation. Camilo et al. (2000) quote estimates of the flux density of 14 pulsars; only two have mean flux greater than the sensitivity (nominally $\sim$ 300 $\mu\mbox{Jy}$ at 1.4 GHz) of the observing system used for their discovery. Intensity enhancements have allowed pulsars with mean flux

$\raisebox{-0.3ex}{\mbox{$\stackrel{<}{_\sim} \,$}}$ 30 $\mu\mbox{Jy}$ to be detected - a factor of 10 below the nominal threshhold. The Compact Array image reported here has a 5 $\sigma$ detection limit of 90 $\mu\mbox{Jy}$ , and so only half of the pulsars detectable with the Parkes telescope are visible in the image. Future improvements to the image sensitivity may come in two ways. Additional integration time is expected to continue reducing the fluctuation level in the image. Secondly, as noted above, the sensitivity of the image reported here is considerably poorer than expected from formal noise estimates. Some improvement may be gained from understanding the origin of this degradation. Finally we note that the pulsars already detected in 47 Tuc probably represent the brightest set of a larger population. Camilo et al. (2000) have estimated that as many as 200 potentially observable pulsars may exist in the cluster. Such a population of pulsars may be visible with the Compact Array as an extended radio source in the cluster core. The luminosity function of pulsars (Lyne, Manchester and Taylor, 1985) is well described by

$dN = -L^{-1} d \log L$ down to a minimum luminosity at 400 MHz of $\sim$ 0.3 mJy kpc², equivalent to 0.024 mJy kpc² at 1.4 GHz, assuming a spectral index of -2. This relation was used by Camilo et al. (2000) to estimate the size of the pulsar population. We use the same relation to estimate the total flux density from the whole population to be $\sim$ 4 mJy, with about half of this from pulsars below the ATCA detection limit and appearing in the extended component. The detection of an extended radio source in the cluster core, and a measurement of its angular extent would have give an interesting alternative estimate of the pulsar populsation and its distribution within the cluster core.

Next Section: Acknowledgements
Title/Abstract Page: Radio Images of the
Previous Section: Image Analysis and Results
Contents Page: Volume 18, Number 2

ASKAP

Public