The Hot and Cold Spots in the Wilkinson Microwave Anisotropy Probe Data Are Not Hot and Cold Enough

This paper presents a frequentist analysis of the hot and cold spots of the cosmic microwave background data collected by the Wilkinson Microwave Anisotropy Probe (WMAP). We compare the WMAP temperature statistics of extrema (number of extrema, mean excursion, variance, skewness and kurtosis of the excursion) to Monte-Carlo simulations. We find that, on average, the local maxima (high temperatures in the anisotropy) are too cold and the local minima are too warm. In order to quantify this claim we describe a two-sided statistical hypothesis test which we advocate for other investigations of the Gaussianity hypothesis. Using this test we reject the isotropic Gaussian hypothesis at more than 99% confidence in a well-defined way. Our claims are based only on regions that are outside the most conservative WMAP foreground mask. We perform our test separately on maxima and minima, and on the north and south ecliptic and Galactic hemispheres and reject Gaussianity at above 95% confidence for almost all tests of the mean excursions. The same test also shows the variance of the maxima and minima to be low in the ecliptic north (99% confidence), but consistent in the south; this effect is not as pronounced in the Galactic north and south hemispheres.