Abstract
Abstract
Accurate collision-induced absorption profiles for H2–He pairs, in the rototranslational band of H2, are computed accounting for the full anisotropic interaction potential. The calculations are time consuming and complicated compared to those pursued in the isotropic potential approximation. A machine learning approach is implemented in order to produce highly accurate data on a dense frequency grid, by combining data computed in the full calculation with those from the isotropic approximation. Thus an extensive, highly accurate, data base can be obtained for a set of frequencies, temperatures, and ortho-H2/para-H2 fractions, appropriate for use in modeling of planetary atmospheres, in particular of the gas giants.
Subject
Computer Science Applications,History,Education