On the effects of the damping mechanisms in an atmospheric general circulation model

Abstract

A version of ECMWF's first spectral model with a comprehensive physical parameterization package has been used for long-term integrations in perpetual January mode. The model has 9 sigma-levels in the vertical and the expansion of the horizontal fields are truncated at total wavenumber n = 31. Inorder to simulate the observed January climate both in terms of meanfields as well as the spectral distribution of the kinetic energy as accurately aspossible, a series of experiments has been performed with varying parameterizations of some physical processes, especially the damping mechanisms. Theinclusion of a simple gravity wave drag is found to have a substantial impact on the simulations. The position and depth of the mean sea level quasi-stationary low pressures in the northern hemisphere is more in accordance with their observed counterparts with the gravity wave drag included, and the excessive stratospheric polar night jet, which is developed in the model without the drag, is effectively controlled by this parameterization. The sensitivity due to changes in the vertical diffusion of momentum, heat and moisture has been considered by crude variations in the mixing length. It turns out that even large changes in the value of the mixing length have a quite modest influence on the mean fields simulated by the model. The horizontal diffusion is necessary in order to represent the effect of the unresolved scales on the explicitly predicted scales. The parameterization of this effect is incorporated in the model as a linear term utilizing that in a spectral formulation, a linear diffusion can easily be formulated with any scale dependency. It is found that an increase of the diffusion in the smallest scales may result in an increase of eddy kinetic energy of the larger scales and even an increase in the total eddy kinetic energy. Moreover it appears that it is quite possible to formulate the linear diffusion of the model to simulate the observed spectral variations of kinetic energy for the medium and smaller scales.