Physical Sciences

The kinematic Sunyaev-Zel'dovich (kSZ) signal is a powerful probe of the cosmic baryon distribution. The kSZ signal is proportional to the integrated free electron momentum rather than the electron pressure (which sources the thermal SZ signal). Since velocities should be unbiased on large scales, the kSZ signal is an unbiased tracer of the large-scale electron distribution, and thus can be used to detect the "missing baryons" that evade most observational techniques. While most current methods for kSZ extraction rely on the availability of very accurate redshifts, we revisit a method that allows measurements even in the absence of redshift information for individual objects. It involves cross-correlating the square of an appropriately filtered cosmic microwave background (CMB) temperature map with a projected density map constructed from a sample of large-scale structure tracers. We show that this method will achieve high signal-to-noise when applied to the next generation of high-resolution CMB experiments, provided that component separation is sufficiently effective at removing foreground contamination. Considering statistical errors only, we forecast that this estimator can yield S/N≈3, 120 and over 150 for Planck, Advanced ACTPol, and a hypothetical Stage IV CMB experiment, respectively, in combination with a galaxy catalog from WISE, and about 20% larger S/N for a galaxy catalog from the proposed SPHEREx experiment. We show that the basic estimator receives a contribution due to leakage from CMB lensing, but that this term can be effectively removed by either direct measurement or marginalization, with little effect on the kSZ significance. We discuss possible sources of systematic contamination and propose mitigation strategies for future surveys. We compare the theoretical predictions to numerical simulations and validate the approximations in our analytic approach. This work serves as a companion paper to the first kSZ measurement with this method, where we used CMB temperature maps constructed from Planck and WMAP data, together with galaxies from the WISE survey, to obtain a 3.8-4.5σ detection of the kSZ2 amplitude.