Foreground contributions to the cosmic microwave background

Abstract

A detailed search has been made for evidence of residual foreground contributions to the cosmic microwave background (CMB) in the map generated by the Wilkinson Microwave Anisotropy Probe (WMAP), a map that has been (nominally) cleaned for the foreground already. We find positive results in that various features relate to Galactic properties. For example, on the largest angular scales we find significant differences between the power in the fluctuations for positive and negative Galactic latitudes and between the four Galactic quadrants. There are also differences between the power spectrum at latitudes within 10 degrees of the plane and at higher latitudes. The mean temperature shows similar variations. An explanation in terms of Galactic effects seems inescapable. In an effort to find the origin of these Galactic-style effects we have examined evidence from Galactic gamma-rays, specifically from the EGRET instrument. We are mindful that the CMB maps examined have already been 'cleaned' (for CR and other effects) in a rather complex way, but, in our view, the cleaning has left some potentially serious 'contaminations'. A correlation is found between gamma-ray intensities and the CMB and other CR indicators. For example, regions of the Galaxy having (line-of-sight) steep CR energy spectra have low mean CMB temperatures, and the important Loop I edge region, where the CR intensity is high, has a high mean temperature. Most of the large-scale Galactic asymmetries (e.g. north-south difference and quadrant variations) have analogues in CR asymmetries and also in some other Galactic properties, such as the column density of gas. Thus, it is possible to hypothesize about direct CR-induced contributions, although it may be that CRs are simply the indicators of Galactic 'conditions' which are influencing the residual CMB fluctuations. Irrespective of the actual cause of the correlations we have endeavoured to extrapolate to the situation where the residual foreground is minimized. The effect on the usually derived cosmological properties is briefly examined. The least that can be said is that the 'error' in some of these properties has been underestimated.