Effects of Reduction in Parameterized Heat Transport to Convective Boundary Layer Simulations at Gray-Zone Resolutions

This study examines the effects of reduction in parameterized vertical heat transport to convective boundary layer (CBL) simulations at gray-zone resolutions. The subgrid-scale (SGS) transport is reduced according to a decrease in horizontal grid size in a computationally simple but physically based way, through curve fitting to the large-eddy simulation (LES) output.

The reduction algorithm possesses three features. First, nonlocal transport by the strong updrafts of organized structures and local transport by remaining small-scale eddies are separately parameterized. Second, SGS nonlocal transport is formulated by multiplying a grid-size dependency function to total nonlocal transport profile fitted to the LES results. Finally, SGS local transport is reduced according to the grid size decrease by multiplying a grid-size dependency to total local transport profile that is calculated by an eddy-diffusivity formula. No explicit grid-size dependency is considered for momentum transport, as the temperature fields modify the momentum fields.

An idealized CBL is simulated at gray-zone resolutions using the new algorithm. Results are evaluated against the LES outcome, and compared with a convectional nonlocal planetary boundary layer (PBL) parameterization. By reducing the parameterized vertical heat transport, improvements over the conventional scheme appear in mean profiles, resolved and SGS vertical transport profiles and their grid-size dependency, as well as energy spectrum. Real-case simulations for convective rolls show that the simulated roll structures are more robust with stronger intensity regardless of resolution when the new algorithm is used.