Precise cosmological parameter estimation using COSMOREC

We use the new cosmological recombination code, COSMOREC, for parameter estimation in the context of (future) precise measurements of the cosmic microwave background (CMB) temperature and polarization anisotropies. We address the question of how previously neglected physical processes in the recombination model of RECFAST affect the determination of key cosmological parameters, for the first time performing a model-by-model computation of the recombination problem. In particular, we ask how the biases depend on different combinations of parameters, e. g. when varying the helium abundance or the effective number of neutrino species in addition to the standard six parameters. We forecast how important the recombination corrections are for Planck, and a survey cosmic-variance limited up to l = 2000. For example, we find that for a A cold dark matter (ACDM) analysis, n(s) may be biased by over 3 sigma and 7 sigma in each case, whereas when Y-p is allowed to vary its value can be biased by over 2 sigma and 7 sigma. We find similar shifts for other parameter combinations. Furthermore, we ask which recombination corrections are really crucial for CMB parameter estimation, and ask whether an approach based on a redshift-dependent correction function to RECFAST is sufficient.