Is cosmology consistent?

TL;DR

This work tests end-to-end consistency of the cosmological framework by combining high-precision CMB measurements with large-scale structure and Ly$\alpha$ forest data. It introduces a lossless compression of 105 CMB bandpowers into 24 bandpowers with full window and covariance information, and fits an 11-parameter adiabatic inflationary model, examining degeneracies and cross-checks with non-CMB priors. The results favor a nearly flat, $\Lambda$-dominated universe with $\omega_b\approx0.02$, $\omega_{\rm dm}$ and $\Omega_\Lambda$ constrained by PSCz and Hubble priors, and place a robust upper limit on tensor modes ($r<0.5$) and a neutrino mass sum of $<4.2$ eV, while indicating a slight red tilt ($n_s\approx0.9$). Overall, the analysis demonstrates strong internal consistency among CMB datasets and external consistency with other cosmological measurements, supporting a coherent standard cosmology and providing constraints that favor small-field inflation models.

Abstract

We perform a detailed analysis of the latest CMB measurements (including BOOMERaNG, DASI, Maxima and CBI), both alone and jointly with other cosmological data sets involving, e.g., galaxy clustering and the Lyman Alpha Forest. We first address the question of whether the CMB data are internally consistent once calibration and beam uncertainties are taken into account, performing a series of statistical tests. With a few minor caveats, our answer is yes, and we compress all data into a single set of 24 bandpowers with associated covariance matrix and window functions. We then compute joint constraints on the 11 parameters of the ``standard'' adiabatic inflationary cosmological model. Out best fit model passes a series of physical consistency checks and agrees with essentially all currently available cosmological data. In addition to sharp constraints on the cosmic matter budget in good agreement with those of the BOOMERaNG, DASI and Maxima teams, we obtain a heaviest neutrino mass range 0.04-4.2 eV and the sharpest constraints to date on gravity waves which (together with preference for a slight red-tilt) favors ``small-field'' inflation models.

Figures (12)

• Figure 1: CMB data used in our analysis. Error bars do not include calibration or beam errors which allow substantial vertical shifting and tilting for some experiments.
• Figure 2: Combination of all data from Figure \ref{['cmbdataFig']}. These error bars include the effects of beam and calibration uncertainties, which cause long-range correlations of order 5%-10% over the peaks. In addition, points tend to be anti-correlated with their nearest neighbors, typically at the level of a few percent. The horizontal bars give the characteristic widths of the window functions (see text).
• Figure 3: Each curve shows the number of standard deviations ("sigmas") at which a given experiment is inconsistent with all others when its power spectrum $\delta T$ is multiplied by a constant $r$.
• Figure 4: Constraints on individual parameters using only CMB and LSS information. The quoted 95% confidence limits are where each curve drops below the dashed line.
• Figure 5: Like the previous figure, but adding the prior $h=0.72\pm 0.08$.