A novel approach to introduce 3D cloud structure into 1D radiative transfer

Detailed geometry and structure of cloud fields cannot be resolved by large scale models despite of the important roles they play in precipitation and radiative transfer processes. We have developed a novel approach to represent statistical characteristics of unresolved three-dimensional (3D) cloud structure in radiative transfer. Subgrid clouds are described by Probability Density Functions (PDFs) and two-point spatial correlation functions of cloud condensate. The new radiative transfer approach has the following features: (1) It is a one-dimensional approach designed to capture most of the 3D effects; (2) Cloud overlap in any direction are treated consistently through spatial correlation functions (thus no overlap assumptions); (3) no need for the cloud fraction concept. The calculated fluxes using the stochastic approach in general agrees well with those from 3D Monte Carlo simulations.