ATCA imaging of clusters of galaxies found in the SUMSS

Introduction

Theoretical predictions suggest that there should be many clusters of galaxies at distances out to a redshift of z=1 (see Jenkins et al., 2001). However, comparatively few clusters have been found at this redshift. It is important to develop new methods for detecting clusters at high redshift. Traditional methods employed to find clusters involve searching for optical or IR overdensities of galaxies in a survey. However, when going to higher redshifts these methods suffer contamination from faint sources at lower redshift, as well as contrast problems with bright sources at lower redshift. X-ray and radio surveys will therefore be very important at finding high redshift clusters.

The new generations of X-ray satellites (XMM, Newton and Chandra) are expected to find many high-redshift clusters. The X-ray surveys detect the bremsstrahlung radiation from the cluster plasma. This emission is dependent on the density squared of the electrons in the plasma and is therefore biased to detecting high-density regions. The X-ray surveys are therefore likely to be biased towards finding relaxed clusters with a well-established cooling flow. For this reason it is important to find other selection techniques to complement the X-ray work.

Radio sources are widely regarded as tracers of high-density regions at high redshifts (e.g. Venemans et al. 2002). In the literature there are examples of clusters being found using deep imaging around a single radio source (e.g. Best 2000, Nakata et al., 2001). Also groupings of radio sources in the NRAO VLA Sky Survey (NVSS) have been used as tracers for clusters of galaxies (e.g. Croft et al., 2001, Cotter et al., 2002). These searches will only find a subset of the cluster population and as in the case of the X-ray surveys they will involve biases. Interactions between galaxies are thought to trigger radio sources (e.g. Best et al., 2002). It is therefore likely that the radio searches will be biased towards merging systems. This is complementary to the X-ray searches that may be biased towards relaxed systems.

Using the SUMSS catalogue to identify potential clusters

The Sydney University Molonglo Sky Survey (SUMSS) is a radio survey carried out using the Molonglo Observatory Synthesis Telescope (MOST). It is carried out at 843 MHz, with a resolution of 43 arcseconds, complete to about 6 mJy (Bock et al., 1999). It is very similar in frequency, resolution and completeness to the NVSS. Building on the success of the clusters found using the NVSS we have been using SUMSS to search for clusters in the Southern Hemisphere. We searched for groupings of 5 radio sources in a 7-arcminute-diameter circle in the SUMSS catalogue (Mauch et al., 2003).

This produced a list of 120 potential clusters. After examining SuperCOSMOS (Hambly et al., 2001) R-band images of these fields we were able to remove obvious chance alignments where several of the sources had clearly unrelated identifications, and some low-redshift clusters. This procedure left 60 candidates for clusters expected to lie at a redshift greater than z ~0.3 (beyond the SuperCOSMOS R-band plate limit for radio source host galaxies).

ATCA observations

All 60 of these candidates have been observed at 20 cm and 13 cm at ATCA using the 6-km array (Buttery et al., 2001). These observations were carried out with the goal of providing more accurate source positions to facilitate optical identifications. The observations have provided some very interesting results. About thirty percent of the cluster candidates showed some sources that were resolved out by the 6-km observations. We expected that this was diffuse emission, which is often associated with clusters (Buttery et al., 2002). Very little is known about this diffuse emission and only a few cases of it have been found to date (see e.g. Clarke and Enblin 2000, Govoni et al., 2001, Thierbach et al., 2003). To determine whether this really was diffuse emission further observations were made with ATCA in November 2002 using the 1.5-km array (see
Figure 1).

Figure 1: The left hand figure shows the ATCA 6-km array map at 20 cm in greyscale overlaid with the SUMSS contours at 5, 6, 8, 12, 24 and 48 mJy. The top left source is completely resolved out. The right hand figure shows the ATCA 1.5-km array in greyscale overlaid with the same SUMSS contours. The top left source is now apparent - confirming that it represents extended emission.

Optical and IR follow-up

In addition to the radio observations, 40 of the potential clusters have been imaged to V~23, R~23 and I~22 with the ANU 2.3-m telescope. These observations are designed to identify clusters out to a redshift of z ~0.7 and to make an estimate of the redshift for clusters below that redshift by locating the position of the 4000-Å break.

Clusters with no identifications out to a redshift of z ~0.7 are high-z candidates. A sample of these have been observed in service observations on IRIS2 on the Anglo-Australian Telescope as well as with SofI in December 2002 on ESO's New Technology Telescope at La Silla. This allows redshift estimation for clusters where the 4000-Å break lies between R and J.

The optical and IR follow-up of our cluster candidates is ongoing, but we have now found clear evidence that there are clusters associated with at least some of the radio sources in several of our candidate fields (Figure 2).

Figure 2: The figure shows a J-band image taken with the NTT showing numerous IR sources surrounding one of the radio sources, as well as a radio source identification. The sources labelled with crosses are not present (or only very faintly present) in the R-band image. This indicates that the 4000-Å break lies between R and J implying a redshift of z ~1. The figure has ATCA 20-cm contours at 0.5, 0.6, 0.8, 1.6, 3.2, 6.4 and 12 mJy overlaid.

Our programme will now advance to multi-object spectroscopy with Gemini-S and VLT, to identify the cluster members without doubt and to investigate their stellar populations and the dynamics of the clusters.

Conclusions

Our results give further evidence to indicate that radio sources trace high-density regions of the universe at high redshifts. We have also found a large number of examples of diffuse emission in clusters. This will contribute to the debate about the nature of these diffuse sources.

References

Best, P.N. (2000), MNRAS 317, 720
Best, P.N. et al., (2002), MNRAS 330, 17
Bock, D. et al., (1999), AJ 117, 1578
Buttery, H. et al., (2001), ASP Conference Proceedings "A New Era in Cosmology"
Buttery, H. et al., (2002), RevMexAA "Galaxy Evolution:Theory and Observations"
Clarke, T. E. and Enblin, T. A. (2000), AAS 197
Cotter, G. et al., (2002), MNRAS 331, 1
Croft, S. et al., (2001), AAS 199 Govoni, F. et al., (2001), A&A 376, 803
Hambly, N.C. et al., MNRAS 326, (2001), 1279
Jenkins, A. et al., (2001), MNRAS 321, 372
Mauch, T. et al., (2003), MNRAS submitted
Nakata, F. et al., (2001), PASJ 53, 1139
Thierbach, M. et al., (2003), A&A 397, 53
Venemans, B.P. et al., (2002), ApJL 569, L11

Helen Buttery, Garret Cotter, Dick Hunstead and Elaine Sadler
(h.buttery@mrao.cam.ac.uk)