Critical evaluation and proposed refinement of the Troen and Mahrt (1986) boundary layer model

A widely used bulk model of the atmospheric boundary layer (BL) was proposed by Troen and Mahrt (1986) (hereinafter TM). The TM parameterization scheme was conceived for use in models where only a coarse resolution in the BL can be achieved (e.g. climate models and weather prediction models up to the meso-beta scale). Nevertheless, this parameterization scheme is still widely adopted in high resolution local area models, even in complex terrain areas. In various LES-based tests of BL parameterizations, the TM model is typically found to fail in predicting the entrainment rate, often overestimating it in a BL with strong shear, and underestimating it in conditions of free convection. As K-closures can still be of interest in NWP, possible refinements to the TM bulk model, correcting its shortcomings, are discussed, with the further aim of reducing as much as possible the number of empirical constants in favour of conceptually based parameterisations.