Determination of cosmological parameters from cosmic shear data

We study how parameter error forecasts for tomographic cosmic shear observations are affected by sky coverage, density of source galaxies, inclusion of cosmic microwave background experiments, simultaneous fitting of nondark energy parameters, and the parametrization of the history of the dark energy equation-of-state parameter $w(z)$. We find tomographic shear-shear power spectra on large angular scales ($l<1000$) inferred from all-sky observations, in combination with Planck, can achieve σ(w₀)=0.06 and σ(wₐ)=0.09 assuming the equation-of-state parameter is given by w(z)=w₀+wₐ[1-a(z)] and that nine other matter content and primordial power spectrum parameters are simultaneously fit. Taking parameters other than w₀, wₐ, and Ωₘ to be completely fixed by the cosmic microwave background (CMB), we find errors on w₀ and wₐ that are only 10% and 30% better, respectively, justifying this common simplifying assumption. We also study ``dark energy tomography'' : reconstruction of $w(z)$ assumed to be constant within each of five independent w bins. With smaller-scale information included by use of the Jain and Taylor ratio statistic, we find σ(wᵢ)<0.1 for all five w bins and σ(wᵢ)<0.02 for both w bins at $z<0.8$. Finally, addition of cosmic shear can also reduce errors on quantities already determined well by the CMB. We find the sum of neutrino masses can be determined to ±0.013 eV and that the primordial power spectrum power-law index, nS, as well as dnₛ/dlnk, can be determined more than a factor of 2 better than by Planck alone. These improvements may be highly valuable since the lower bound on the sum of neutrino masses is 0.06 eV as inferred from atmospheric neutrino oscillations, and slow-roll models of inflation predict nonzero dnS/dlnk at the forecasted error levels when |nS-1|>0.04.