Abstract
Abstract. The treatment of unresolved cloud–radiation interactions in weather and climate models has considerably improved over the recent years, compared to conventional plane-parallel radiation schemes, which previously persisted in these models for multiple decades. One such improvement is the state-of-the-art Tripleclouds radiative solver, which has one cloud-free and two cloudy regions in each vertical model layer and is thereby capable of representing cloud horizontal inhomogeneity. Inspired by the Tripleclouds concept, primarily introduced by Shonk and Hogan (2008), we incorporated a second cloudy region into the widely employed δ-Eddington two-stream method with the maximum-random overlap assumption for partial cloudiness. The inclusion of another cloudy region in the two-stream framework required an extension of vertical overlap rules. While retaining the maximum-random overlap for the entire layer cloudiness, we additionally assumed the maximum overlap of optically thicker cloudy regions in pairs of adjacent layers. This extended overlap formulation implicitly places the optically thicker region towards the interior of the cloud, which is in agreement with the core–shell model for convective clouds. The method was initially applied on a shallow cumulus cloud field, evaluated against a three-dimensional benchmark radiation computation. Different approaches were used to generate a pair of cloud condensates characterizing the two cloudy regions, testing various condensate distribution assumptions along with global cloud variability estimate. Regardless of the exact condensate setup, the radiative bias in the vast majority of Tripleclouds configurations was considerably reduced compared to the conventional plane-parallel calculation. Whereas previous studies employing the Tripleclouds concept focused on researching the top-of-the-atmosphere radiation budget, the present work applies Tripleclouds to atmospheric heating rate and net surface flux. The Tripleclouds scheme was implemented in the comprehensive libRadtran radiative transfer package and can be utilized to further address key scientific issues related to unresolved cloud–radiation interplay in coarse-resolution atmospheric models.
Highlights
Radiation schemes in coarse-resolution numerical weather prediction and climate models, commonly referred to as general circulation models (GCMs), have traditionally been claimed to be impaired by the poor representation of clouds (Randall et al, 1984, 2003, 2007)
Three-dimensional (3-D) radiative effects related to subgrid horizontal photon transport, which in reality manifests itself most pronouncedly in regions characterized by strong horizontal gradients of optical properties, such as cloud side boundaries (Jakub and Mayer, 2015, 2016; Klinger and Mayer, 2014, 2016), are currently still neglected in the majority of GCMs
We assume the maximum overlap of optically thicker cloudy regions in pairs of adjacent layers and abbreviate this extended overlap rule to the “maximum2random overlap”. This assumption implicitly places the optically thicker cloudy region towards the interior of the cloud in the horizontal plane, which is in line with the core–shell model
Summary
Radiation schemes in coarse-resolution numerical weather prediction and climate models, commonly referred to as general circulation models (GCMs), have traditionally been claimed to be impaired by the poor representation of clouds (Randall et al, 1984, 2003, 2007). In order to make the method applicable to GCMs, Shonk et al (2010) derived a global estimate of cloud horizontal variability in terms of fractional standard deviation (FSD), which can be used to split the mean LWC into two components along with the LWC distribution assumption They incorporated a generalized vertical overlap parameterization, called the exponential-random overlap, accounting for the aforementioned problematics in strongly sheared conditions. Per is to document the present Tripleclouds implementation in the comprehensive radiative transfer package libRadtran (Mayer and Kylling, 2005; Emde et al, 2016) Another aim of this study is to explore the TC potential for shallow cumulus clouds, applying various solver configurations diagnosing atmospheric heating rate and net surface flux.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
