Lambda-CDM and the WMAP power spectrum beam profile sensitivity

Abstract: We first discuss the sensitivity of the WMAP CMB power spectrum to systematic errors by calculating the raw CMB power spectrum from WMAP data. We find that the power spectrum is surprisingly sensitive to the WMAP radiometer beam profile even at the position of the first acoustic peak on ~1 degree scales. Although the WMAP beam profile core is only 12.6arcmin FWHM at W, there is a long power-law tail to the beam due to side-lobes and this causes significant effects even at the first peak position. We then test the form of the beam-profile used by the WMAP team which is based on observations of Jupiter. We stacked radio source beam profiles as observed in each WMAP band and found that they showed a wider profile in Q, V, W than the Jupiter profile. We have now checked that this is not due to any Eddington or other bias in our sample by showing that the same results are obtained when radio sources are selected at 1.4GHz and that our methods retrieve the Jupiter beam when it is employed in simulations. Finally, we show that the uncertainty in the WMAP beam profile allows the position as well as the amplitude of the first peak to be changed and how this could allow simpler cosmologies than standard Lambda-CDM to fit the CMB data.