Nonlinear equilibration of baroclinic eddies: the role of boundary layer processes and seasonal forcing

In this thesis, the influence of boundary layer processes and seasonal forcing on baroclinic eddy equilibration is studied to understand how the baroclinic adjustment is modified when taking into account these two factors. A modified β plane multilevel quasi-geostrophic (QG) model with an interactive stratification and a simplified parameterization of atmospheric boundary layer physics is used as the atmospheric model in this study. Comparisons between experiments with the modified QG model and the traditional QG model with fixed stratification show that it is necessary to include the interaction between vertical eddy heat flux and stratification to obtain a realistic equilibrated state, i.e. robust isentropic slope. A slab surface model is also coupled with the atmospheric model to provide an interactive surface temperature distribution in some experiments in this study. The effect of boundary layer processes is first studied under the situation with fixed underlying surface temperature. The boundary layer vertical thermal diffusion, along with the surface heat exchange, is found primarily responsible for the limitation of the PV homogenization in the boundary layer. The boundary layer processes can influence the eddy activity in at least two competing ways. First, in the eddy energy budget, all the boundary layer processes act to damp the eddy energy directly. On the other hand, the boundary layer processes also influence the mean flow, which can further influence the eddy behavior. For the boundary layer thermal damping, the indirect effect on the eddy activity becomes dominant. Stronger boundary layer thermal forcing results in stronger meridional temperature gradients and eddy heat fluxes. For the frictional dissipation, the resulting changes in the zonal wind and in the location of the critical latitude lead to a meridional variation of the eddy forcing, which can further result in a non-monotonic response of the mean flow. The role of the boundary layer processes in the baroclinic eddy equilibration is further studied using a simple air-sea thermally coupled model. Although in the coupled system, each boundary layer process has more and different ways to influence the equilibrium state, their effect on the lower level PV homogenization is very robust. Surface friction and surface 3 heat flux all act to damp the lower level PV mixing and stronger surface damping prevents the PV homogenization more strongly, but the way in which each boundary layer process affects the PV homogenization is very different. Baroclinic eddy equilibration under seasonal forcing is studied in both the atmospheric model and the coupled model. In the situation with specified seasonal variation of the underlying surface temperature, a Northern-Hemisphere like seasonal variation of the surface temperature is used to act on the atmospheric flow through the boundary layer processes and the radiative-convective heating. Under slowly varying seasonal forcing, the eddy and the mean flow behavior is characterized by four clearly divided time intervals: an eddy inactive time interval in the summer, a mainly dynamically determined eddy spinup time interval starting from mid-fall and lasting less than one month, a quasi-equilibrium time interval for the zonal mean flow available potential energy from late fall to late spring and a mainly external forcing determined eddy spindown time interval from late winter to late spring. In spite of the strong seasonality of the eddy activity, a robust PV structure is still observed through all the seasons. It is found that besides baroclinic eddies, the boundary layer thermal forcing as well as the moist convection all can help maintain the lower level PV structure. The sensitivity study of the eddy equilibration to the time scale of the external forcing also indicates that the time scale separation between the baroclinic adjustment and the external forcing in midlatitudes is only visible for external forcing cycle one year and longer. The seasonality study with the coupled model confirms the conclusions obtained in the uncoupled model. Thesis Supervisor: Peter H. Stone Title: Professor of Climate Dynamics