Fitting and smoothing of opacity data

Abstract

Techniques are described for the fitting and smoothing of Rosseland mean opacities, |$\kappa_{R}(T,\rho ),$| where T is the temperature and ρ the mass density, and for calculating the first two derivatives, |$\partial \kappa_\text {R}/ \partial T \vert{\rho}\enspace \text {and}\enspace \partial \kappa_\text {R}/ \partial \rho \vert_{T}$|⁠. The codes OPALFIT and OPFIT are available for the processing of opacities calculated by the projects OPAL and OP. Both projects use techniques of opacity sampling (that is to say, a number of frequency points insufficient to resolve all detailed structures in monochromatic opacities), which lead to some lack of smoothness in the tabulated values of |$\kappa_\text {R}(T,\rho )$|⁠. It is shown that improved accuracy for the derivatives is obtained by using a technique of two-dimensional smoothing.