Infrared Cooling in Cloudy Atmospheres: Precision of Grid Point Selection for Numerical Models

Abstract

The infrared layer temperature change in a cloudy atmosphere normally shows warming at the base of the cloud and intense cooling at the top of the cloud. In a model that uses broad-band radiative transfer to calculate atmospheric temperature changes, errors on the order of 6.0 K day?1 can occur at the top of a cloud if only selected grid points are used. Calculations using grid points at discrete intervals produce a smoothing effect over the entire cooling profile and increase errors. Two case studies were used to demonstrate these problems.