FIVE-YEAR<i>WILKINSON MICROWAVE ANISOTROPY PROBE</i>OBSERVATIONS: LIKELIHOODS AND PARAMETERS FROM THE<i>WMAP</i>DATA

This paper focuses on cosmological constraints derived from analysis of WMAP data alone. A simple LCDM cosmological model fits the five-year WMAP temperature and polarization data. The basic parameters of the model are consistent with the three-year data and now better constrained: Omega_b h^2 = 0.02273+-0.00062, Omega_c h^2 = 0.1099+-0.0062, Omega_L = 0.742+-0.030, n_s = 0.963+0.014- 0.015, tau = 0.087+-0.017, sigma_8 = 0.796+-0.036. With five years of polarization data, we have measured the optical depth to reionization, tau>0, at 5 sigma significance. The redshift of an instantaneous reionization is constrained to be z_reion = 11.0+-1.4 with 68% confidence. This excludes a sudden reionization of the universe at z=6 at more than 3.5 sigma significance, suggesting that reionization was an extended process. Using two methods for polarized foreground cleaning we get consistent estimates for the optical depth, indicating an error due to foreground treatment of tau~0.01. This cosmological model also fits small-scale CMB data, and a range of astronomical data measuring the expansion rate and clustering of matter in the universe. We find evidence for the first time in the CMB power spectrum for a non-zero cosmic neutrino background, or a background of relativistic species, with the standard three light neutrino species preferred over the best-fit LCDM model with N_eff=0 at >99.5% confidence, and N_eff > 2.3 (95% CL) when varied. The five-year WMAP data improve the upper limit on the tensor-to-scalar ratio to r < 0.43 (95% CL), for power-law models. With longer integration we find no evidence for a running spectral index, with dn_s/dlnk = -0.037+-0.028.