2005ApJ...635...22S

Near-future Sunyaev-Zel'dovich (SZ) surveys such as ACT, SPT, APEX, and Planck will soon find thousands of galaxy clusters. Multifrequency arcminute-resolution SZ observations can, in principle, determine each cluster's gas temperature (T_e), bulk velocity (v_pec), and optical depth (τ). However, the frequency bands and detector sensitivity employed by upcoming surveys will generally not be sufficient to disentangle the degeneracy between these three cluster parameters, even in the absence of SZ signal contamination from point sources and imperfect primary microwave background subtraction. Assuming contaminants can be removed, we find that near-future SZ surveys will be able to constrain well two cluster gas parameters that are linear combinations of τT_e, τv_z, and τT²_e. Because the SZ intensity shift is nearly a linear function of τT_e, τv_z, and τT²_e, a correspondence exists between the two effective gas parameters that SZ surveys can constrain and simple line-of-sight integrals through the three-dimensional cluster. We illustrate the parameter constraints and correspondence to line-of-sight integrals using three-dimensional N-body+hydro cluster simulations and a Markov-chain Monte Carlo method. We show that adding an independent temperature measurement to upcoming SZ data breaks the parameter degeneracy and that the cluster effective velocity thus constrained is approximately the optical-depth-weighted velocity integrated along the cluster line of sight. A temperature prior with an error as large as 2 keV still gives bulk velocity errors of 100 km/s or less, even for a more typical cluster with an electron temperature of 3 keV, for ACT-like SZ observations in the absence of signal contamination. The Markov-chain constraints on v_pecand τ that we obtain are more encouraging and most likely more accurate than those obtained from Fisher matrices.