Statistical characterization of cosmic microwave background temperature patterns in anisotropic cosmologies

We consider the issue of characterizing the coherent large-scale patterns from cosmic microwave background (CMB) temperature maps in globally anisotropic cosmologies. The methods we investigate are reasonably general; the particular models we test them on are the homogeneous but anisotropic relativistic cosmologies described by the Bianchi classification. Although the temperature variations produced in these models are not stochastic, they give rise to a 'non-Gaussian' distribution of temperature fluctuations over the sky that is a partial diagnostic of the model. We explore two methods for quantifying non-Gaussian and/or non-stationary fluctuation fields in order to see how they respond to the Bianchi models. We first investigate the behaviour of phase correlations between the spherical harmonic modes of the maps. Then we examine the behaviour of the multipole vectors of the temperature distribution which, though defined in harmonic space, can indicate the presence of a preferred direction in real space, i.e. on the 2-sphere. These methods give extremely clear signals of the presence of anisotropy when applied to the models we discuss, suggesting that they have some promise as diagnostics of the presence of global asymmetry in the Universe.