Handout (175.1 kB)
The presentation will focus on the examination of several popular parameterizations of mixing lengths to simulate the GABLS case in a single column model. These formulations include: 1) a local Richardson number scheme used in the Canadian GEM operational model, 2) the Bougeault-Lacarrere (1989) scheme, 3) a length scale proposed by Therry and Lacarrere (1983), and 4) a modified version of the Lenderink and Holtslag (2004) formulation. It will be shown that the first two length scales tend to give a too deep stable boundary layer, while the two other schemes provide a much improved simulation when compared to the reference LES results.
Examination of the impact of those formulations on mesoscale forecasts with the Canadian GEM model will also be presented at the Symposium.