3.3. Interferometric Observations
The stability and spatial filtering inherent to interferometry has been exploited to make high-quality images of the SZE. The stability of an interferometer is due to its ability to perform simultaneous differential sky measurements over well-defined spatial frequencies. The spatial filtering of an interferometer also allows the emission from radio point sources to be separated from the SZE emission.
There are several other features that allow an interferometer to achieve extremely low systematics. For example, only signals that correlate between array elements will lead to detected signal. For most interferometers, this means that the bulk of the sky noise for each element will not lead to signal. Amplifier gain instabilities for an interferometer will not lead to large offsets or false detections, although, if severe, they may lead to somewhat noisy signal amplitude. To remove the effects of offsets or drifts in the electronics, as well as the correlation of spurious (noncelestial) sources of noise, the phase of the signal received at each telescope is modulated, and then the proper demodulation is applied to the output of the correlator.
The spatial filtering of an interferometer also allows the emission from radio point sources to be separated from the SZE emission. This is possible because at high angular resolution ( 10") the SZE contributes very little flux. This allows one to use long baselines, which give high angular resolution, to detect and monitor the flux of radio point sources, while using short baselines to measure the SZE. Nearly simultaneous monitoring of the point sources is important, as they are often time variable. The signal from the point sources is then easily removed, to the limit of the dynamic range of the instrument, from the short-baseline data, which are sensitive also to the SZE.
For the reasons given above, interferometers offer an ideal way to achieve high brightness sensitivity for extended low-surface brightness emission, at least at radio wavelengths. Most interferometers, however, were not designed for imaging low-surface brightness sources. Interferometers have been built traditionally to obtain high angular resolution and thus have employed large individual elements for maximum sensitivity to small-scale emission. As a result, special-purpose interferometric systems have been built for imaging the SZE (Carlstrom et al., 1996; Jones et al., 1993; Padin et al., 2001). All of them have taken advantage of low-noise HEMT amplifiers (Pospieszalski et al., 1995) to achieve high sensitivity.
The first interferometric detection (Jones et al., 1993) of the SZE was obtained with the Ryle Telescope (RT). The RT was built from the 5 Kilometer Array, consisting of eight 13 m telescopes located in Cambridge, England, operating at 15 GHz with East-West configurations. Five of the telescopes can be used in a compact E-W configuration for imaging of the SZE (Grainger et al., 2002; Grainge et al., 1993; Saunders et al., 2003; Jones et al., 2003; Grainge et al., 1996; Jones et al., 1993; Grainge et al., 2002b; Grainge et al., 2002a).
The OVRO and BIMA SZE imaging project uses 30 GHz (1 cm) low-noise receivers mounted on the OVRO (1) and BIMA (2) mm-wave arrays in California. They have produced SZE images toward 60 clusters to date (Carlstrom et al., 1996; Carlstrom et al., 2000; LaRoque et al., 2003; Grego et al., 2000; Grego et al., 2001; Reese et al., 2000; Joy et al., 2001; Reese et al., 2002; Patel et al., 2000). A sample of their SZE images is shown in Figure 5. All contours are multiples of 2 of each image, and the full-width at half maximum (FWHM) of the synthesized beam (PSF for this deconvolution) is shown in the lower left-hand corner of each image. Figure 5 also clearly demonstrates the independence of the SZE on redshift. All of the clusters shown have similarly high X-ray luminosities, and, as can be seen, the strength of the SZE signals are similar despite the factor of 5 in redshift. The OVRO and BIMA arrays support two-dimensional configurations of the telescopes, including extremely short baselines, allowing good synthesized beams for imaging the SZE of clusters at declinations greater than ~ - 15 degrees.
Figure 5. Deconvolved interferometric SZE images for four galaxy clusters over a large redshift range (0.17 z 0.89). The contours are multiples of 2, and negative contours are shown as solid lines. The FWHM ellipse of the synthesized beam (PSF) is shown in the lower-left corner of each panel. The rms, , appears in the top of each panel. Radio point sources were removed from three of the images shown. The interferometer was able to separate the point source emission from the SZE by using the high-resolution data obtained with long baselines. All of the clusters shown have similarly high X-ray luminosities, and, as can be seen, the strength of the SZE signals are similar despite the factor of 5 in redshift, illustrating the independence of the SZE on redshift.
The RT, OVRO, and BIMA SZE observations are insensitive to the angular scales required to image low-redshift (z << 0.1) clusters. Recently, however, the Cosmic Background Imager (CBI; Padin et al. 2001) has been used to image the SZE in a few nearby clusters (Udomprasert et al., 2000). The CBI is composed of 13 0.9 m telescopes mounted on a common platform, with baselines spanning 1 m to 6 m. Operating in 10 1 GHz channels spanning 26 - 36 GHz, it is sensitive to angular scales spanning 3' to 20'. The large field of view of the CBI, 0.75 degrees FWHM, makes it susceptible to correlated contamination from terrestrial sources (i.e., ground emission). To compensate, they have adopted the same observing strategy as for single-dish observations (Section 3.2), by subtracting from the cluster data, data from leading and trailings fields offset by ± 12.5 minutes in Right Ascension from the cluster.
Interferometric observations of the SZE, as for single-dish observations, are just beginning to demonstrate their potential. Upcoming instruments will be over an order of magnitude more sensitive. This next generation of interferometric SZE instruments will conduct deep SZE surveys covering tens, and possibly hundreds, of square degrees. While not as fast as planned large-format bolometric arrays, the interferometers will be able to survey deeper and provide more detailed imaging. In particular, the high resolution and deep imaging provided by future heterogeneous arrays will provide a valuable tool for investigating cluster structure and its evolution. Such studies are necessary before the full potential of large SZE surveys for cosmology can be realized.
1 An array of six 10.4 m telescopes located in the Owens Valley, CA, operated by Caltech. Back.
2 An array of 10 6.1 m mm-wave telescopes located at Hat Creek, CA, operated by the Berkeley-Illinois-Maryland-Association. Back.