A Refined Method of Parameterizing Absorption Coefficients among Multiple Gases Simultaneously from Line-by-Line Data

Abstract

An extension of the correlated-k distribution method that uses spectral-mapping techniques was derived to parameterize line-by-line absorption coefficients for multiple gases simultaneously for use in three-dimensional atmospheric models. In a variation from previous correlation techniques, this technique ensures exact correlation of absorption frequencies within a probability interval for all gases through all layers of the atmosphere when multiple gases are considered. The technique is physical since, in reality, gases are correlated in wavelength space. The technique, referred to as the ?multiple-absorber correlated-k distribution spectral-mapping method,? was found to be accurate to <0.7% of incident radiation for 16 probability intervals per wavelength interval, integrated over 0.4?1000-?m wavelengths and accounting for 11 absorbing gases simultaneously and multiple layers, compared with an exact line-by-line solution. A method was also derived to reduce the number of probability intervals required for a radiative transfer solution without suffering the same inaccuracy as merely reducing the number of probability intervals when parameterizing the absorption coefficient. The new coefficients were tested in a global model, and results were compared with mean thermal-IR irradiance data.