Modified correlated k-distribution methods for remote sensing applications

Abstract

In this paper we discuss the application of correlated k-distribution methods for the computation of radiative transfer through the atmosphere in the visible, near IR, and IR spectral regions. Whereas the correlated k-distribution is a good approximation overall to line-by-line calculations, it is inaccurate in regions of strong absorber overlap. Several approaches (band splitting, strength-based splitting, double and triple correlated k) are proposed to modify the conventional correlated k-distribution and improve its predictions in these regions of absorber line overlap. A comparison with line-by-line computations in the infrared spectral region demonstrates that a combination of these methods produces brightness temperatures within less than 0.25 K of those obtained from line-by-line computations at significantly less computation costs. In the visible and near IR, when atmospheric scattering is included, the new treatment of overlapping mixed gases also produces excellent agreement with the line-by-line model.