Variance similarity in shallow cumulus topped mixed layers
| Neggers, Roel | ECMWF |
| Stevens, Bjorn | Department of Atmospheric and Oceanic Sciences, UCLA |
| Neelin, David | Department of Atmospheric and Oceanic Sciences, UCLA |
Thermodynamic variance similarity in shallow cumulus topped mixed layers is studied using large-eddy simulation (LES) results. The simulations are based on a range of different shallow cumulus cases, including marine steady state cumulus as well as continental diurnal cycles. The latter are based on observations from the Southern Great Plains (SGP) site of the Atmospheric Radiation Measurement (ARM) program, and from the Small Cumulus Microphysics Study (SCMS).
First the performance of standard similarity scaling based on the surface latent and sensible heat fluxes is evaluated. The results confirm the observation that surface-based bottom-up scaling fails in the top half of the subcloud mixed layer. A local variance budget analysis is then perfomed to derive new scales for the thermodynamic variances at mixed layer top. Its essential new features are that i) the local vertical gradient is retained and ii) the mixed layer turnover timescale is used as the relevant adjustment timescale for the dissipation of variance. These new scales successfully capture the time-development of the variance at mixed layer top during the diurnal cycle cases. Furthermore, the data-collapse over all cases of the mixed layer variance profiles in the top half of the mixed layer allows for the definition a new similarity scaling law.
Finally, a further detailed analysis of these new scales is performed which leads to some new conclusions on the spatial structure of shallow cumulus convection. The physics introduced in the variance by the local vertical gradient and the turnover timescale relate the area fraction of the active cumulus clouds to the depth structure of the mixed layer. This results in a new parameterization for the variance at shallow cumulus cloud base, which successfully captures the phase and magnitude as diagnosed in observations and LES.
This poster will be displayed at the ARM Science Team Meeting.