Beecroft Institute for Particle Astrophysics and Cosmology

Abstract:

Although the broad outlines of the appropriate pipeline for cosmological likelihood analysis with CMB data has been known for several years, only recently have we had to contend with the full, large-scale, computationally challenging problem involving both highly-correlated noise and extremely large datasets ($N > 1000$). In this talk we concentrate on the beginning and end of this process. First, we discuss estimating the noise covariance from the data itself in a rigorous and unbiased way; this is essentially an iterated minimum-variance mapmaking approach. We also discuss the unbiased determination of cosmological parameters from estimates of the power spectrum or experimental bandpowers.

Abstract:

We revisit the possibility that we inhabit a compact multi-connected flat, or nearly flat, Universe. Analysis of COBE data has shown that, for such a case, the size of the fundamental domain must be a substantial fraction of the horizon size. Nevertheless, there could be several copies of the Universe within the horizon. If the Milky Way was once a quasar we might detect its 'ghost' images. Using new large quasar catalogues we repeat the search by Fagundes & Wichoski for antipodal quasar pairs. By applying linear theory to account for the peculiar velocity of the Local Group, we are able to narrow the search radius to 134 arcsec. We find seven candidate antipodal quasar pairs within this search radius. However, a similar number would be expected by chance. We argue that, even with larger quasar catalogues, and more accurate values of the cosmological parameters, it is unlikely to be possible to identify putative ghost pairs unambiguously, because of the uncertainty of the correction for peculiar motion of the Milky Way.