The Anisotropy of the Microwave Background to l = 3500: Deep Field Observations with the Cosmic Background Imager

TL;DR

The paper reports deep-field measurements of the cosmic microwave background anisotropy with the Cosmic Background Imager, extending the measured spectrum to ell up to 3500. It confirms the damping of primary anisotropies and maps the tail of the spectrum out to ell ~ 2000, while finding a significant excess power at ell > 2000 that exceeds standard intrinsic models. The analysis employs a robust maximum-likelihood framework with careful subtraction of discrete radio sources, complemented by OVRO 31 GHz source measurements and NVSS-based statistics to control foregrounds. The authors discuss possible origins for the high-ℓ excess, including secondary Sunyaev-Zeldovich anisotropy, and emphasize the need for further observations to pinpoint its origin and implications for sigma_8 and structure formation.

Abstract

We report measurements of anisotropy in the cosmic microwave background radiation over the multipole range l ~ 200 - 3500 with the Cosmic Background Imager based on deep observations of three fields. These results confirm the drop in power with increasing l first reported in earlier measurements with this instrument, and extend the observations of this decline in power out to l \~ 2000. The decline in power is consistent with the predicted damping of primary anisotropies. At larger multipoles, l = 2000 - 3500, the power is 3.1 sigma greater than standard models for intrinsic microwave background anisotropy in this multipole range, and 3.5 sigma greater than zero. This excess power is not consistent with expected levels of residual radio source contamination but, for sigma_8 >~ 1, is consistent with predicted levels due to a secondary Sunyaev-Zeldovich anisotropy. Further observations are necessary to confirm the level of this excess and, if confirmed, determine its origin.

Figures (7)

• Figure 1: The distributions of the weights of the visibilities as a function of $\ell$ for the three deep fields. Since the CBI sensitivity does not vary much with baseline or frequency, these weights are roughly proportional to the on sky integration time as a function of $\ell$.
• Figure 2: CBI joint Deep Field window functions. The alternate binning window functions are displayed in the upper panel; the primary binning window functions are displayed in the lower panel. Vertical lines denote the bandpower bin divisions discussed in the text. The window function of the high-$\ell$ rebinning is shown as a dashed line. The expected value of band power for a given spectrum, $C_l$, in one bin is $\Sigma_l\; W(l)/l \times l(l+1)C_l/2\pi$
• Figure 3: Power spectra from the individual CBI deep fields shown for the primary binning only. The 08h, 14h, and 20h data are denoted by squares (blue), circles (green), and crosses (red), respectively. The horizontal location of each point on the power spectrum is $\ell_{\rm eff}$ and the horizontal error bars indicate the full extent of each band.
• Figure 4: Joint CBI deep field power spectra for the primary (blue open circles) and alternate (green open squares) binnings. The placement of the points in $\ell$ is determined as in figure \ref{['fig:deepspec']}. Also shown are the thermal noise power spectrum (open stars) and the residual source power spectrum (solid triangles).
• Figure 5: Comparison of the CBI deep field power spectrum with BOOMERANG newboom, DASI dasispectrum, and MAXIMA new_maxima results. The rectangles indicate the 68% confidence intervals on band-power; for BOOMERANG, the solid rectangles indicate the 68% confidence interval for the statistical and sample variance errors, while the hatched rectangles shows the amount by which a $\pm1\sigma$ error in the beamwidth ($12\hbox{$.\mkern-4mu^\prime$}9 \pm 1\hbox{$.\mkern-4mu^\prime$}4$) would shift the estimates (all up or all down together). Results from the CBI primary binning are shown as blue open circles, and the alternate binning results are shown as green open squares. The high value of the power seen in the first bin of the primary and alternate binnings relative to other observations has rather low significance ($\sim 1.7 \sigma$) due to the large sample variance in the deep field spectra at low $\ell$; this is discussed further in the text.