Colloquium: Dr. Jake Gristey, Univ. of Colorado Boulder

Energy originating from the sun and reaching Earth’s surface, known as the surface solar irradiance (SSI), drives virtually all Earth-atmosphere processes. The SSI is strongly modulated by clouds, including shallow cumulus clouds that are encountered frequently across the globe. The detailed 3D spatial structure of shallow cumulus clouds, their rapid evolution, and aerosol embedded within the cloud field, all lead to complex variability in SSI that has proved challenging to understand and predict. In this talk, I will show how this SSI variability is captured concisely by the bi-modal shape of the SSI probability density function (PDF), which is robustly observed but only reproduced by 3D radiative transfer. I will demonstrate that, despite common beliefs, 3D radiative effects do not disappear with space-time averaging. To address these issues, machine-learning algorithms are explored to construct a mapping between observationally constrained cloud and aerosol properties, derived from large eddy simulation, and the corresponding SSI PDF, derived from 3D radiative transfer, for a variety of shallow cumulus cases at the Southern Great Plains Atmospheric Observatory. I will reveal the drastic improvements in prediction of SSI relative to traditional 1D radiative transfer, while bypassing the computational expense of 3D radiative transfer. I will also share an approach that quantifies the relative importance of each predictor to understand the physical controls on SSI variability. The new findings have important implications for solar renewable energy assessment, highlight the significance of the absence of 3D radiative effects in weather and climate modeling, and provide a route forward for efficient parameterization of 3D radiative effects at the surface.