Schematic diagram of Non-local PBL / Vertical diffusion

       (Click to see how the equations are calcuated in codes!)


 Introduction of Non-local PBL / Vertical diffusion

 Turbulence in the atmospheric boundarylayer (ABL) causes mixing of heat, moisture, momentum, and passive scalars. Global weather forecasting and climate models typically describe the turbulent mixing with an eddy diffusivity based on local gradients of wind and potential temperature. Such a so-called local K approach is treated, for instance, by Louis (1979). Local K theory may fail in the unstable boundary layer because the influence of large eddy transports is not accounted for (e.g., Wyngaard and Brost 1984; Holtslag and Moeng 1991), and entrainment effects are not treated in such an approach. This may affect the profiles of mean quantities, especially at locations where dry convection is of importance in ABL.

 A nonlocal ABL scheme is used in the NCAR Community Climate Model, Version 2 (CCM2). The non-local ABL scheme is based on the work by Troen and Mahrt (1986) and Holtslag et al. (1990). It utilizes an eddy diffusivity profile and incorporates the nonlocal effects of transport by large eddies in a simplified manner (Holtslag and Moeng 1991). The latter represents the effects of dry convective plumes whose vertical scale is the depth of the boundary layer. Within this scheme, the boundary-layer depth is calculated explicitly. It also appears that the non-local diffusion scheme is more robust from a numerical point of view, because it is less sensitive to stability oscillations (Beljaars 1991)

The vertical diffusion parameterization in SNUGCM provides the interface to the turbulence parameterization, computes the molecular diffusivities, and finally computes the tendencies of the input variables. The diffusion equations are actually solved implicitly, so the tendencies are computed from the difference between the final and initial profiles.