EC-02. Understanding and predicting the distribution of surface solar irradiance beneath shallow cumulus clouds

Abstract
Ubiquitous shallow cumulus clouds strongly modulate surface solar irradiance (SSI). Their three-dimensional (3D) spatial structure leads to complex variability in SSI distribution. This variability is captured by the SSI probability density function (PDF), which is typically bi-modal representing separately the cloud shadows and the gaps between. Our interest is to understand the relationship between cloud field properties, which we derive from large eddy simulation, and the SSI PDF shape, which we calculate using 3D radiative transfer. Applying machine learning algorithms, we find that it is possible to predict the SSI PDF shape using just a handful of key cloud field properties. This approach permits quantification of the relative importance of each cloud field property to understand the controls on SSI variability, and also provides an opportunity to bypass the computational expense of 3D radiative transfer. An additional complexity arises from embedded aerosol between the clouds that also influences SSI; we will show preliminary results demonstrating how aerosol perturbs the SSI PDF shape and plans to incorporate aerosol properties into our SSI prediction. These findings have important implications for solar renewable energy assessment and highlight the significance of the absence of 3D radiative effects in weather and climate forecasting.