Redshift estimation of clusters by wavelet decomposition of their Sunyaev-Zel'dovich morphology

A method for estimating redshifts of galaxy clusters based solely on resolved Sunyaev-Zel'dovich (SZ) images is proposed. Given a high-resolution SZ cluster image (with a FWHM of similar to 1 arcmin), the method indirectly measures its structure-related parameters ( amplitude, size, etc.) by fitting a model function to the higher-order wavelet moments of the cluster's SZ morphology. The applicability and accuracy of the wavelet method are assessed by applying the method to maps of a set of clusters extracted from hydrodynamical simulations of cosmic structure formation. The parameters, derived by a fit to the spectrum of wavelet moments as a function of scale, are found to show a dependence on redshift z that is of the type x(z) = x(1) exp(-z/x(2)) + x(3), where the monotony of this functional behaviour and the nondegeneracy of those parameters allow inversion and estimation of the redshift z. The average attainable accuracy in the z estimation relative to 1+z is similar to 4-5 per cent out to z similar or equal to 1.2, which is comparable with photometric redshifts. For single-frequency SZ interferometers, in which the ambient fluctuating CMB is the main noise source, the accuracy of the method drops slightly to similar to 6-7 per cent. Other complications addressed include instrumental noise, cold cores and systematic trends in baryon fraction with cluster mass