Covariance matrix calculated from nuclear models

A production method of a covariance matrix from a nuclear model calculation is described. An optical model is used to estimate the covariance matrix of the 54Fe total cross section, and a Hauser-Feshbach and a precompound models are used for the covariance matrices of 54,56Fe(n,p) reaction cross sections. These matrices are calculated from the uncertainties of level density parameters, a precompound parameter, and the optical potential parameters for neutron, proton, and α-particle. The neutron optical potential parameters and their uncertainties are evaluated from the experimental total and elastic scattering data, while the other parameters and their uncertainties are evaluated from the experimental 54,56Fe(n,p), (n,α), (n,2n), (n,Xp) and (n,Xα) reaction cross sections. (author)