Nonnucleosynthetic Constraints on the Baryon Density and Other Cosmological Parameters

Because the baryon-to-photon ratio η_10 is in some doubt, we drop nucleosynthetic constraints on η_10 and fit the three cosmological parameters (h, Ω_M, η_10) to four observational constraints: Hubble parameter h_o = 0.70 ± 0.15, age of the universe t_0 =14^+7_-2 Gyr, cluster gas fraction f_o ≡ f_G h^3/2 = 0.060 ± 0.006, and effective shape parameter Γ_o = 0.255 ± 0.017. Errors quoted are 1 σ, and we assume Gaussian statistics. We experiment with a fifth constraint Ω_o = 0.2 ± 0.1 from clusters. We set the tilt parameter n = 1 and the gas enhancement factor Upsilon = 0.9. We consider cold dark matter models (open and Ω_M = 1) and flat ΛCDM models. We omit HCDM models (to which the Γ_o constraint does not apply). We test goodness of fit and draw confidence regions by the Δχ^2 method. CDM models with Ω_M = 1 (SCDM models) are accepted only because the large error on h_o allows h 0.4. The combination of the four other constraints with Ωo approx 0.2 is rejected in CDM models with 98% confidence, suggesting that light may not trace mass. ΛCDM models give similar results. In all of these models, η_10 ~> 6 is favored strongly over η_10 ~< 2. This suggests that reports of low deuterium abundances on QSO lines of sight may be correct and that observational determinations of primordial ^4He may have systematic errors. Plausible variations on n and Upsilon in our models do not change the results much. If we drop or change the crucial Γ_o constraint, lower values of Ω_M and η_10 are permitted. The constraint Γ_o = 0.15 ± 0.04, derived recently from the IRAS redshift survey, favors Ω_M approx 0.3 and η_10 approx 5 but does not exclude η_10 approx 2