The estimation of the SZ effects with unbiased multifilters

In this work we study the performance of linear multifilters for the estimation of the amplitudes of the thermal and kinematic Sunyaev-Zel'dovich effects. We show that when both effects are present, estimation of these effects with standard matched multifilters is intrinsically biased. This bias is due to the fact that both signals have basically the same spatial profile. We find a new family of multifilters related to the matched multifilters that cancel this systematic bias, hence we call them Unbiased Matched Multifilters. We test the unbiased matched multifilters and compare them with the standard matched multifilters using simulations that reproduce the future Planck mission's observations. We find that in the case of the standard matched multifilters the systematic bias in the estimation of the kinematic Sunyaev-Zel'dovich effect can be very large, even greater than the statistical error bars. Unbiased matched multifilters cancel effectively this kind of bias. In concordance with other works in the literature, our results indicate that the sensitivity and resolution of Planck will not be enough to give reliable estimations of the kinematic Sunyaev-Zel'dovich of individual clusters. However, since the estimation with the unbiased matched multifilters is not intrinsically biased, it can be possible to use them to statistically study peculiar velocities in large scales using large sets of clusters.