The thermal Sunyaev-Zel'dovich effect power spectrum in light of Planck

The amplitude of the thermal Sunyaev-Zel'dovich effect (tSZ) power spectrum is extremely sensitive to the abundance of the most massive dark matter haloes (galaxy clusters) and therefore to fundamental cosmological parameters that control their growth, such as sigma(8) and Omega(m). Here we explore the sensitivity of the tSZ power spectrum to important non-gravitational ('subgrid') physics by employing the cosmo-OWLS suite of large-volume cosmological hydrodynamical simulations, run in both the Planck and 7-year Wilkinson Microwave Anisotropy Probe (WMAP7) best-fitting cosmologies. On intermediate and small angular scales (l greater than or similar to 1000, or theta less than or similar to 10 arcmin), accessible with the South Pole Telescope (SPT) and the Atacama Cosmology Telescope (ACT), the predicted tSZ power spectrum is highly model dependent, with gas ejection due to active galactic nuclei (AGN) feedback having a particularly large effect. However, at large scales, observable with the Planck telescope, the effects of subgrid physics are minor. Comparing the simulated tSZ power spectra with observations, we find a significant amplitude offset on all measured angular scales (including large scales), if the Planck best-fitting cosmology is assumed by the simulations. This is shown to be a generic result for all current models of the tSZ power spectrum. By contrast, if the WMAP7 cosmology is adopted, there is full consistency with the Planck tSZ power spectrum measurements on large scales and agreement at the 2 sigma level with the SPT and ACT measurements at intermediate scales for our fiducial AGN model, which Le Brun et al. have shown reproduces the 'resolved' properties of the Local Group and cluster population remarkably well. These findings strongly suggest that there are significantly fewer massive galaxy clusters than expected for the Planck best-fitting cosmology, which is consistent with recent measurements of the tSZ number counts. Our findings therefore pose a significant challenge to the cosmological parameter values preferred (and/or the model adopted) by the Planck primary cosmic microwave background analyses.