A closer look at CMB constraints on WIMP dark matter

We use Cosmic Microwave Background data from the WMAP, SPT, BICEP, and QUaD experiments to obtain constraints on the dark matter particle mass $m_\chi$, and show that the combined data requires $m_\chi > 7.6$ GeV at the 95% confidence level for the $\chi \chi \rightarrow b \bar b$ channel assuming $s-$wave annihilation and a thermal cross section $<\sigma_{\rm a} v> = 3 \times 10^{-26}$ cm$^3/$s. We examine whether the bound on $m_\chi$ is sensitive to $\sigma_8$ measurements made by galaxy cluster observations. The large uncertainty in $\sigma_8$ and the degeneracy with $\Omega_{\rm m}$ allow only small improvements in the dark matter mass bound. Increasing the number of effective neutrino-like degrees of freedom to $N_{\rm eff} = 3.85$ improves the mass bound to $m_\chi > 8.6$ GeV at 95% confidence, for the $\chi \chi \rightarrow b \bar b$ channel. We also study models in which dark matter halos at $z<60$ reionize the Universe. We compute the Ostriker-Vishniac power resulting from partial reionization at intermediate redshifts $10<z<60$, but find the effect to be small. We discuss the importance of the large angle polarization as a complementary probe of dark matter annihilation. By performing Monte Carlo simulations, we show that future experiments that measure the $EE$ power spectrum from $20 < l < 50$ can exclude $m_\chi \sim$ 10 GeV at the 2 (3) $\sigma$ level provided the error bars are smaller than 4 (3) $\times$ cosmic variance. We show that the Planck experiment will significantly improve our knowledge of dark matter properties.