Department of Astrophysical Sciences, Princeton University,
Princeton, NJ 08544, USA
The Wilkinson Microwave Anisotropy Probe (WMAP) has made it possible to test Gaussianity of fluctuations in the cosmic microwave background with unprecedented sensitivity, thanks to its careful instrumental design minimizing systematic errors. Constraints on the CMB non-Gaussianity offer a fundamental test of inflation and general relativistic perturbation theory. On the other hand, extragalactic radio sources produce significant non-Gaussian signals, making physical interpretations of the measured non-Gaussianity complex. Applying to the WMAP 1-year data a statistical tool we have developed for separately measuring the primordial and the source non-Gaussianities, we find: (1) CMB is consistent with Gaussian fluctuations, as predicted by inflation. The r.m.s. contribution from a second order (non-linear) term is smaller than 2x10-5 of the leading term. Note that the second-order perturbation theory predicts 10-100 times smaller value. (2) We have detected the source contribution at 41GHz (Q band); thus, the WMAP map in Q band is weakly non-Gaussian at small angular scales because of the sources! This measurement has been used to determine the source contribution to the power spectrum in Q band. Our method can be applied to the Planck experiment, for which there is a chance to detect non-Gaussianity predicted by the second-order perturbation theory.A step-by-step instruction to clean and combine the WMAP sky maps for the CMB analysis is given in Section 2.