Influences of surface energy partitioning on convective initiation and organization: Observations and cloud-permitting model experiments

Interactions between the land-surface, boundary-layer, and convective clouds remain challenging to model and observe. In this study, cloud-permitting hindcast experiments (perturbed soil moisture) and observations (e.g., boundary-layer profiles in the U.S. Southern Great Plains) were used to identify systematic responses of convective clouds to surface energy partitioning, across multiple states and spatial scales. The results reveal a set of tropospheric states in which a drier but warmer surface is more likely to initiate deep convection, by deepening the boundary layer and reducing the convective inhibition energy. This 'dynamically-limited' regime can be contrasted with a thermodynamic regime in which deep convection is favored over a cooler but wetter surface, by lowering the boundary layer and accumulating convective available potential energy in a shallower layer. A few cases of simulated surface influences on convective initiation exhibit upscale growth of convective clouds into larger mesoscale systems, suggesting soil moisture feedback on heavy precipitation events. The extent to which convective initiation is dynamically limited varies greatly among boundary-layer parameterizations, and is an important target for improving drought-onset prediction.