­What is the Global Impact of 3D Radiative Transfer?

In nature, clouds are complex 3D structures that interact with radiation in all directions. Many models now account for sub-grid cloud structure, but due to computational constraints, they all neglect the effects of horizontal transport in global simulations of weather and climate. Concerned for the crudeness of the approximation and discrepancies between observations and modeled radiation fields, several studies have quantified the ‘3D radiative effect’ of clouds, finding that it can be large, particularly for broken cloud fields. Nevertheless, and perhaps out of convenience, a common assumption has been that 3D effects average out over the diurnal cycle in the shortwave and can be neglected completely in the longwave. Thanks to the new SPeedy Algorithm for Radiative Transfer through Cloud Sides (SPARTACUS), which is sufficiently fast for use in a general circulation model, this hypothesis can now be tested.

SPARTACUS parameterizes 3D effects in a 1D atmospheric column using ‘cloud effective size', a parameter related to the edge length of the clouds within a model grid box. In the first part of this talk, we will use satellite and ground-based radar measurements, and cloud-resolving model simulations, to derive a best estimate for this parameter and its uncertainty. The dependence of the cloud effective size on cloud regime and other indicators will be discussed. We then use these estimates in multi-year free-running coupled simulations of the ECMWF model. Crucially these simulations allow the 3D effects to feedback on the climate system and allow the global impact of 3D radiative transfer to be quantified for the first time. The inclusion of 3D radiative effects is found to change annual-mean surface fluxes by up to 10 W m^-2 locally. In a global-mean sense, surface downwelling shortwave and longwave fluxes increase by around 1 W m^-2, which leads to a significant warming of the land surface.