We describe a search for clusters of galaxies via their Sunyaev-Zel'dovich (SZ) effect at 31 GHz with the Sunyaev-Zel'dovich Array (SZA). The SZA is a purpose-built 8-telescope interferometer located at the Owens Valley Radio Observatory in California, equipped with a dual-band (31 GHz and 90 GHz) receivers and a wide-band 8 GHz digital correlator. The receivers employ HEMT technology for amplification, and the data presented in this thesis were obtained with the 31 GHz system.
To demonstrate the workings of the instrument, we present the analysis of commissioning observations on three high-redshift, X-ray selected clusters of galaxies whereby we obtain estimates of electron temperatures, gas masses, and total cluster masses. Data were subsequently taken on a 5.2 square degree region of the sky to a sensitivity of ~ 0.25 mJy/beam. This survey was analyzed for the properties of foreground sources of emission and to place a constraint on σ-s, the amplitude of mass density fluctuations on a scale of 8Mpc/h.
The SZA survey generated a catalog of 110 compact sources with fluxes greater than 1mJy at 31 GHz. The median spectral index for these sources when compared to counterparts found at 5 GHz with the Very Large Array (VLA) is found to be 0.48 +/- 0.45, consistent with a distribution of steep-spectrum synchrotron sources. The differential source counts as a function of flux for this sample is dN/dS = (0.0055 +/- 0.0006) (S)^(-1.969 +/- 0.008), where the flux, S is expressed in mJy and the area is in square arcminutes. This value is consistent with previous measurements and theoretical predictions. We further demonstrate the presence of a dust-associated diffuse foreground in fields near the plane of the Galaxy, which could be due to dipole emission from dust grains spinning in a magnetic field.
We perfrom a Bayesian statistical analysis on the source-extracted survey data to place a constraint on σ-s, accounting for systematic uncertainties in the SZ flux-Mass relation and in the mass function. At the time of this writing we have not yet accounted for other sources of systematic errors, some of which will significantly impact our result. These include intrinsic assumptions of clsuter physics contained within the simulations used to relate cluster flux to mass; the correlation between sources of emission and clusters of galaxies; and sample variance, due to the relatively small size of our survey. Overlooking these additional sources of errors, we place an upper limit on σ-s of 0.68 +/- 0.04 at 98% confidence, qualifying that this result should not be considered final until the previously-mentioned sources of systematic uncertainties are properly incorporated in our analysis.