Precise Estimation of Cosmological Parameters Using a More Accurate Likelihood Function

Abstract

The estimation of cosmological parameters from a given data set requires a construction of a likelihood function which, in general, has a complicated functional form. We adopt a Gaussian copula and constructed a copula likelihood function for the convergence power spectrum from a weak lensing survey. We show that the parameter estimation based on the Gaussian likelihood erroneously introduces a systematic shift in the confidence region, in particular for a parameter of the dark energy equation of state w. Thus, the copula likelihood should be used in future cosmological observations.