Biases on the cosmological parameters and thermal Sunyaev-Zel'dovich residuals

We examine the biases induced on cosmological parameters when the presence of secondary anisotropies is not taken into account in cosmic microwave background analyses. We first develop an exact analytical expression for computing the biases on parameters when any additive signal is neglected in the analysis. We then apply it in the context of the forthcoming Planck experiment. For illustration, we investigate the effect of the sole residual thermal Sunyaev-Zel'dovich signal that remains after cluster extraction. We find, in particular, that analyses neglecting the presence of this contribution introduce on the cosmological parameters n s and biases, at least 6.5 and 2.9 times their 1 confidence intervals. The b parameter is also biased to a lesser extent.