Extended Mosaic Observations with the Cosmic Background Imager

TL;DR

This work combines two years of CBI observations to produce a high-resolution CMB power spectrum over $400 < \ell < 3500$ and calibrates it to the WMAP scale. Using MCMC analyses of the primary anisotropy ($\ell<2000$) with flat+weak-$h$ priors, it yields precise cosmological parameters and shows that high-$\ell$ data help break degeneracies present in WMAP-only analyses; it also explores a potential high-$\ell$ excess power attributed to the Sunyaev-Zeldovich effect to constrain $\sigma_8^{\rm SZ}$, finding a peak around $\sim 0.96$ with notable uncertainties. The results favor a flat, near scale-invariant universe and provide evidence for a small negative running of the scalar spectral index, while confirming damping at small scales and supporting a consistent picture when combined with ACBAR and BIMA data. Overall, the study demonstrates the power of combining mosaic and deep-field CBI data with cross-calibration to tighten cosmological constraints and to probe secondary anisotropies on arcminute scales.

Abstract

Two years of microwave background observations with the Cosmic Background Imager (CBI) have been combined to give a sensitive, high resolution angular power spectrum over the range 400 < l < 3500. This power spectrum has been referenced to a more accurate overall calibration derived from WMAP. The data cover 90 deg^2 including three pointings targeted for deep observations. The uncertainty on the l > 2000 power previously seen with the CBI is reduced. Under the assumption that any signal in excess of the primary anisotropy is due to a secondary Sunyaev-Zeldovich anisotropy in distant galaxy clusters we use CBI, ACBAR, and BIMA data to place a constraint on the present-day rms mass fluctuation sigma_8. We present the results of a cosmological parameter analysis on the l < 2000 primary anisotropy data which show significant improvements in the parameters as compared to WMAP alone, and we explore the role of the small-scale cosmic microwave background data in breaking parameter degeneracies.

Figures (12)

• Figure 1: Comparison of Jupiter temperatures measured with the CBI and with WMAP. The spectrum of Jupiter in this frequency range is not thermal owing to an absorption feature. The individual channel CBI temperatures of Jupiter in the 26--36 GHz range are shown by the filled blue circles, and those of WMAP in the range 22--94 GHz are shown by the open pink squares. The dotted and solid lines, respectively, show the best fit slopes to WMAP and CBI data over the 20 to 40 GHz range. The CBI temperatures shown here were determined assuming $T_{J}(32 {\rm GHz})=152 \pm 5$K Mason99; the systematic error bar on this calibration is shown as a dotted red line. The WMAP error bars include both random and systematic errors.
• Figure 2: The 2000+2001 CBI Spectrum. The "even" binning is shown in red and the "odd" binning in light blue. Orange stars indicate the thermal noise variance; green triangles indicate the statistical source correction which has been subtracted from the power spectrum. The solid black line is the WMAP$\Lambda$CDM model with a pure power-law primordial spectrum (model spectrum is file wmap_lcdm_pl_model_yr1_v1.txt, available on the WMAP website http://lambda.gsfc.nasa.gov).
• Figure 3: The 2000+2001 CBI window functions ("even" binning).
• Figure 4: The 2000+2001 CBI window functions ("odd" binning).
• Figure 5: Marginalized likelihood curves for a range of individual cosmological parameters, each shown for three CMB datasets: " WMAP only" (blue/dotted); "CBI + WMAP" (red/dashed); and "CBI + ALL" (green/solid). In all cases a flat plus weak-$h$ prior is used.