A new high- and low-frequency scattering parameterization for cirrus and its impact on a high-resolution numerical weather prediction model

Abstract

A new high- and low-frequency scattering parameterization for cirrus is presented. The parameterization is based on an ensemble model of cirrus ice crystals, which has been previously demonstrated to follow observed mass-dimensional power laws and predict the volume extinction coefficient of cirrus to within current experimental uncertainties. The scalar optical properties of the ice crystal ensemble are calculated using the methods of T-matrix and geometric optics, and the bulk scalar optical properties are calculated by integrating the single-scattering properties over a moment estimation parameterization of the particle size distribution (PSD). The moment estimation parameterization estimates the PSDs from the second moment and in-cloud temperature. Therefore, the bulk scalar optical properties are directly related to a prognostic variable of a general circulation model (GCM), i.e., the ice water content (IWC). The radiative impact of the new parameterization is assessed by directly comparing model predictions against CERES reflected short-wave (SW) flux measurements. It is demonstrated that for one case, the new optical parameterization reduces the short-wave bias by about 60% relative to the former effective dimension-based parameterization. However, both parameterizations predict long-wave (LW) fluxes that are generally within the CERES measurement uncertainty of ±10 Wm-2. There is, however, a systematic negative bias in the reflected SW flux at TOA (i.e., clouds are generally too bright). © 2013 AIP Publishing LLC.