Robert Kerr (Mathematical Institute, University of Warwick, UK). Tools for identifying resolved-scale turbulence in convection simulations.

Now that mesoscale models of atmospheric convection are regularly being run with horizontal resolutions of less than a kilometre, we need to start thinking about these calculations as large-eddy simulations where resolved-scale turbulence, not parameterised turbulence, should do most of the mixing at those scales.  Unfortunately, many heavily parameterised atmospheric models seem unable to be run in an LES mode.  The first goal of this presentation is to give examples of atmospheric convection calculations in this regime that do, and do not, generate resolved-scale turbulence and the consequences of when they do not. The second goal is to identify analysis tools developed for understanding turbulence in simulations of Rayleigh-Bénard convection between heated and cooled plates and show how they could be used in the atmospheric context to identify when there is resolved scale turbulence and what type of turbulence it is. Three possible types of turbulence are considered, all of which generate -5/3 regimes in the energy spectra.