No Higher Criticism of the Bianchi-corrected Wilkinson Microwave Anisotropy Probe data

Abstract

Higher Criticism (HC) has been proposed by Donoho & Jin as an effective statistic to detect deviations from Gaussianity. Motivated by the success of the Bianchi VIIh model in addressing many of the anomalies observed in the Wilkinson Microwave Anisotropy Probe (WMAP) data (Jaffe et al.), we present calculations in real and in wavelet space of the HC statistic of the Bianchi-corrected WMAP first-year data. At the wavelet scale of 5°, the HC of the WMAP map drops from a value above the 99 per cent confidence level (c.l.) to a value below the 68 per cent CL when corrected by the Bianchi template. An important property of the HC statistic is its ability to locate the pixels that account for the deviation from Gaussianity. The analysis of the uncorrected WMAP data pointed to a cold spot in the Southern hemisphere, centred at l ~ 209°, b ~ −57°. The HC of the Bianchi-corrected map indicates that this spot remains prominent, albeit at a level completely consistent with Gaussian statistics. Consequently, it is debatable how much emphasis should be placed on this residual feature, but we consider the effect of modestly increasing the scaling of the template. A factor of only 1.2 renders the spot indistinguishable from the background level, with no noticeable impact on the results published in Jaffe et al. for the low-l anomalies, large-scale power asymmetry or wavelet kurtosis. A trivial interpretation would be that the Bianchi template may require a small enhancement of power on scales corresponding to the wavelet scale of 5°.