A computationally efficient parameterization for 3D radiative effects at the surface in large-eddy simulations

The amount of solar radiation that reaches the Earth’s surface is strongly influenced by clouds and aerosols. This results in a complex pattern of radiation at the surface with cloud shadows and regions with enhanced solar radiation. Cloud enhancements can cause surface irradiances up to 25% higher than under clear sky conditions. Insight into these peaks is valuable to operators of the electricity grid, as these peaks can cause grid voltages to exceed the safety limits. Furthermore, the radiation at the surface impacts the surface fluxes, boundary layer turbulence and clouds. Therefore, proper modelling of surface radiation is important for predictions of solar energy production and for detailed weather forecasts. Current 1D calculations of radiative transfer cannot capture the complex pattern of surface radiation, whereas 3D calculations are very costly. In this study, we developed a simple method to account for the 3D radiative effects at the surface in LES. The horizontal spreading of the diffuse radiation is accounted for by applying a spatial filter to the surface diffuse radiation field. With this filter, we add only little extra complexity to the existing 1D calculation. Therefore, our method is computationally efficient. The filtering of the diffuse radiation is applied to the results of a Large Eddy Simulation for a summer day in Cabauw, the Netherlands, on which shallow cumulus clouds formed during the day. The results of the LES simulation are compared to detailed high-quality observations (1Hz). Without the filtering, the cloud enhancements are not captured at all, and the probability distribution of global radiation is unimodal, whereas the observed distribution is bimodal. After filtering, our results closely match with the observations. The probability distribution of global radiation is now bimodal and cloud enhancements are simulated. We found that small changes in the filter width do not strongly influence the results. Furthermore, we showed that the width of the filter can be parameterized as a linear function of one variable related to the cloud field, e.g. the maximum cloud size or the cloud base height. Hence, this work presents a proof-of-concept for our method to come to more realistic surface irradiances by filtering diffuse radiation at the surface.