The Development of a Cyclone–Anticyclone Asymmetry within a Growing Baroclinic Wave

This study examines the cyclone–anticyclone asymmetry within idealized baroclinic waves. For this purpose a primitive equation (PE) model and a quasigeostrophic (QG) model are used. The Rossby number appears to be an essential external parameter for the cyclone–anticyclone asymmetry. Consequently, the Rossby number effect on nonlinear baroclinic waves is a central point of this study. First, the baroclinic stability of a vertically antisymmetric dipole jet flow is analyzed. Increasing the Rossby number in the PE model has the effect that the eddies of the fastest growing normal mode are tilted toward the northwest–southeast direction. The increase of this tilt is almost proportional to the Rossby number and arises due to leading-order corrections to the QG model. The nonlinear development of the fastest growing normal mode is simulated for various Rossby numbers. The PE model simulations exhibit the development of strong cyclones and weak anticyclones even when the cyclonic and anticyclonic shears in the basic-state flow are symmetrical. This asymmetry increases with increasing Rossby number, while in the QG model such an asymmetry does not appear at all due to a flow symmetry conservation. Perturbing this flow symmetry in a linearized QG model reveals that the QG solution is unstable with respect to symmetry-breaking disturbances. These disturbances grow due to baroclinic, as well as barotropic, energy conversions. Corresponding simulations with symmetry-breaking disturbances are carried out with the nonlinear QG model. The simulated developments strongly resemble those obtained from the PE model without an additional perturbation. A conceptual model for the symmetry-breaking in the PE model is proposed. In this model, the simultaneous action of the symmetry-breaking instability and a non-QG forcing causes the cyclone–anticyclone asymmetry. The latter mechanism appears as an inhomogeneous forcing that is proportional to the Rossby number and can, therefore, explain the influence of this parameter on the PE model results.