资源环境科技发展态势分析平台

Effects of cloud condensate vertical alignment on radiative transfer process were investigated using cloud resolving model explicit simulations, which provide a surrogate for subgrid cloud geometry. Diagnostic results showed that the decorrelation length L-cw varies in the vertical dimension, with larger L-cw occurring in convective clouds and smaller L-cw in cirrus clouds. A new parameterization of L-cw is proposed that takes into account such varying features and gives rise to improvements in simulations of cloud radiative forcing (CRF) and radiative heating, i.e., the peak of bias is respectively reduced by 8 W m(-2) for SWCF and 2 W m(-2) for LWCF in comparison with L-cw = 1 km.

The role of L-cw in modulating CRFs is twofold. On the one hand, larger L-cw tends to increase the standard deviation of optical depth sigma(tau), as dense and tenuous parts of the clouds would be increasingly aligned in the vertical dimension, thereby broadening the probability distribution. On the other hand, larger sigma(tau) causes a decrease in the solar albedo and thermal emissivity, as implied in their convex functions on tau. As a result, increasing (decreasing) L(cw)leads to decreased (increased) CRFs, as revealed by comparisons among L-cw = 0, L-cw = 1 km and L-cw = infinity. It also affects the vertical structure of radiative flux and thus influences the radiative heating. A better representation of up in the vertical, dimension yields an improved simulation of radiative heating. Although the importance of vertical alignment of cloud condensate is found to be less than that of cloud cover in regards to their impacts on CRFs, it still has enough of an effect on modulating the cloud radiative transfer process. (C) 2016 The Authors. Published by Elsevier B.V.