Scale-adaptive and flow-adaptive sub-grid methods

"Dynamic" methods for sub-grid turbulence modelling have a long history in the engineering community but have been relatively little explored within meteorology. However, recent explorations have suggested that they might be a valuable tool within the turbulent grey zone. The basic idea is to compare the small-scale flow within a simulation to the filtered form of the flow. We would like to assume that the sub-grid parameterization should be applicable at both scales. By minimizing any mismatch in this assumption, we can set the values of parameter(s) within the parameterization in a way that adapts to both the scale difference of interest and the local character of the flow.

In this project we study the parameter estimations arising from a dynamic Smagorinsky model for large-eddy simulations of shallow convection as a function of the filter scales, and location within the flow (e.g. within cloud or within the mixed layer). We then feed back the dependencies found into grey zone simulations of the same cases. Can such information improve the simulations, and, if so, what aspects of the parameter depndencies are most important to be able to capture?

Some links for this work:

Talks:

1. A talk summarizing the project presented by Bob Plant at a turbulent grey zone workshop.