The role of lateral entrainment in the shallow convective cloud response to boundary layer perturbations

Description

Shallow convective clouds cover large areas of the oceans in the trade wind regions and have important thermodynamic and radiative impacts on the climate and weather. A previous study using observations from the Barbados observatory has shown that the cloud fraction near cloud base of these cumuli remains remarkably invariant over longer time scales, and primarily responds on short time scales to perturbations in the boundary layer to restore equilibrium (the cumulus valve mechanism). Meanwhile, cloudiness from convective outflow aloft underneath the trade inversion better correlates with changes in the large-scale forcing on longer time scales. Global models could not reproduce this observed variability of cloudiness near cloud base or aloft, indicating that some important processes relating forcing on a variety of timescales to the cloud response are not accurately captured in current shallow convection parameterizations. We use the IFS global model to help to understand the reasons why the parametrizations might not capture this variability with the aim of improving the model representation of trade cumulus cloud. In the IFS, shallow convection is parameterized using a mass flux scheme with a closure aiming to restore dry static energy equilibrium in the boundary layer on short time scales. While this closure is consistent with the hypothesised cumulus valve mechanism, the variability in cloudiness produced by the scheme is not consistent with that observed. Single column model testing of idealised equilibrium cases as well as observed non-equilibrium cases (MAGIC transect and ARM SGP daytime convection) reveal that the treatment of organised lateral entrainment is a crucial determinant for the cloud response to forcing in the current parameterized framework. Since the convective updraft fraction does not explicitly translate into a cloud fraction, cloud mass detrained from the rising plume largely determines the cloud fraction profile. We explore parameter sensitivities in an ensemble SCM to achieve a more realistic cloud fraction profile in response to the forcing.