EC-01. Systematic Changes in Shallow Cumulus Cloud Field Evolution due to Shortwave 3D Radiative Responses

Abstract
Shallow cumulus cloud fields, found widely across the globe, exhibit highly three-dimensional (3D) spatial structure. It follows that simulating realistic interaction of solar radiation with these cloud fields presents both an intellectual and computational challenge. In atmospheric models, the radiation is almost always simplified to one-dimensional, where each column of the model is treated separately. This simplification breaks down dramatically for fine scale models (e.g., large eddy simulation; LES), where horizontal transport of solar radiation between columns is frequent and has important radiative effects. While these 3D radiative effects have been well documented, the response of the cloud field development, evolution, and persistence is relatively less well known. Existing work is limited to case studies that suggest these responses could be substantial with direct relevance for applications from renewable energy to climate science. In recognition of the importance of shortwave 3D radiative responses in shallow cumulus cloud fields, and aided by recent advances in computational resources, coupling of 3D radiation with LES has attracted notable attention. We will demonstrate the performance of one such approach that applies an emulator of 3D radiative transfer for coupling with the System for Atmospheric Modeling (SAM). We propose to simulate a large number of continental shallow cumulus cases using both the new coupled 3D radiation SAM setup, and the traditional 1D radiation SAM setup. Systematic comparison of the two sets of simulations permits a statistical approach to quantifying 3D radiative responses on shallow cumulus cloud field evolution, moving beyond a unique or small number of case studies, and enabling assessment of the robustness and variability of the responses. Given that robust radiative responses are identified, it also remains to demonstrate that the results from the coupled 3D radiation SAM setup are more realistic by comparing with observations. We argue that, given the differences that can occur between LES and observations for any given case, a reliable comparison with observations is only possible via a statistical approach across a large number of cases as enabled by the new emulator. Opportunities and challenges associated with this approach will be discussed.