Cosmological Parameters from Cosmic Background Imager Observations and Comparisons with BOOMERANG, DASI, and MAXIMA

TL;DR

This work analyzes Cosmic Background Imager (CBI) data, extending CMB power-spectrum measurements to high multipoles ($300<\ell<3500$) to constrain a minimal inflationary parameter set $\{\Omega_{\rm tot}, \Omega_{\Lambda}, \Omega_b h^2, \Omega_{\rm cdm} h^2, n_s, \tau_C, \ln {\cal C}_{10}\}$. It introduces radical data compression into bandpowers, applies a comprehensive set of priors (weak-$h$, flat, LSS, HST-$h$, SN), and uses offset-lognormal likelihoods to robustly infer cosmological parameters, including tight constraints on geometry, matter densities, and the Hubble constant. The results show a flat universe with a nearly scale-invariant spectrum, low matter density, a cosmological constant, and baryon density in line with BBN; they remain robust when subsets of data are removed or when different binning schemes are used, and they are consistent with other CMB experiments (BOOMERANG, DASI, MAXIMA, VSA). Overall, the study strengthens the ΛCDM inflationary paradigm and demonstrates the power of high-$\ell$ CMB observations in linking early-universe physics to present-day cluster scales.

Abstract

We report on the cosmological parameters derived from observations with the Cosmic Background Imager (CBI), covering 40 square degrees and the multipole range 300 < l < 3500. The angular scales probed by the CBI correspond to structures which cover the mass range from 10^14 to 10^17 M_sun, and the observations reveal, for the first time, the seeds that gave rise to clusters of galaxies. These unique, high-resolution observations also show damping in the power spectrum to l ~ 2000, which we interpret as due to the finite width of the photon-baryon decoupling region and the viscosity operating at decoupling. Because the observations extend to much higher l the CBI results provide information complementary to that probed by the Boomerang, DASI, Maxima, and VSA experiments. As the observations are pushed to higher multipoles no anomalies relative to standard models appear, and extremely good consistency is found between the cosmological parameters derived for the CBI observations over the range 610 < l < 2000 and observations at lower l [abridged].

Figures (14)

• Figure 1: Features in the anisotropy spectrum (from Paper III). The first acoustic peak is seen at high sensitivity in the BOOMERANG Netterfield02, DASI Halverson02, and MAXIMA Lee01 observations, while the second and third acoustic peaks are seen at lower sensitivity (the rectangles indicate the 68% confidence intervals on band-power). The circles (dark blue) and squares (green) show the odd and even binnings of the CBI results from the joint spectrum of the three mosaic fields (see Paper III. Note that the two binnings are highly correlated with each other and are not independent measurements of the power spectrum). The damping tail is clearly seen in the CBI spectrum, and, in the region of overlap, all four experiments are in excellent agreement, as is discussed in § \ref{['sec:optimal2']}. The black curve is the joint model also discussed in § \ref{['sec:optimal2']}.
• Figure 2: 1D projected likelihood functions calculated for the CBIo140+DMR data. All panels include the weak-$h$ (solid dark blue) and LSS+weak-$h$(short-dash-dotted red) priors. (LSS is the large-scale structure prior.) The $\Omega_k$ panel also shows what the whole ${\cal C}_\ell$-database gives before the weak-$h$ prior is imposed (black dotted). We note that even in the absence of CMB data there is a bias towards the closed models lange01. In the other panels, flat+weak-$h$ (long-dashed-dotted light blue) and LSS+flat+weak-$h$ (dashed green) are plotted. Notice how stable the $n_s$ determination is, independent of priors. We see here that, under priors ranging from the weak-$h$ prior to the weak-$h$+LSS+flat priors, the CBI provides a useful measure of four out of the six fundamental parameters shown. This is independent of the first acoustic peak, where the CBI has low sensitivity, and is also largely independent of the spectrum below $\ell \sim 610$ for all but $\Omega_b h^2$ (see text).
• Figure 3: Results obtained using DMR alone. This gives an idea of the role of the LSS prior in sharpening up detections for DMR. Note that DMR did reasonably well by itself in first indicating for this class of models that $n_s \sim 1$bh95. Of course it could not determine $\omega_b$ and the structure in $\Omega_k$ and $\Omega_\Lambda$ can be traced to ${\cal C}_\ell$-database constraints lange01. Comparison with Fig. \ref{['fig:likecbimO']} shows the greatly improved constraints when the CBI data are added.
• Figure 4: Same as Fig. \ref{['fig:likecbimO']}, except that CBI data at $\ell<610$ have been discarded. We see that, under the weak-$h$ and LSS+weak-$h$ priors, $\Omega_{\rm k}$ is peaked near zero, as is the case for the whole data set, showing that in the CBI data the evidence for a flat universe is coming from the $\ell$-range above the second acoustic peak where damping dominates, as well as from the lower-$\ell$ range. We see here that under increasingly restrictive priors the CBI data at $\ell>610$ provide useful constraints on $n_s$, $\Omega_m h^2$ and $\Omega_\Lambda$, showing that these results are driven by the shape (and level) of the spectrum at high-$\ell$ independent of the results at low-$\ell$. Note, however, that when the CBI data are restricted to $\ell>610$ they do not provide a useful measure of $\Omega_b h^2$ (compare with Fig. \ref{['fig:likecbimO']}).
• Figure 5: Same as Fig. \ref{['fig:likecbimO']}, except for CBIe140+DMR. By comparing this figure with Fig. \ref{['fig:likecbimO']}, we see that the particular choice of bin boundaries does not make a significant difference to the parameter estimation. This can also be seen from the comparison of the even and odd binning cases in Table \ref{['tab:paramscbitests']}.