Journal of the Meteorological Society of Japan, Vol. 81, No. 3,

To broaden the range of known circulations and to test existing theory, a variety of issues are examined concerning the dynamics of flows in thick, thin, and transitional atmospheric layers. The circulations are produced numerically using a primitive equation model subject to simple heating functions. To confine the motions to a thin upper layer, the heating is chosen to produce a flow with either an exponential (EXP) vertical structure, or one that is linear (LIN) aloft while vanishing below. Five sets of solutions are created to define the terrestrial and jovian axisymmetric states, some basic terrestrial states, and the transitional jovian states for the two structures. The axisymmetric cases examine how the surface drag, static stability, rotation rate, and layer thickness influence the flow character. The standard theory is extended to allow for a weaker drag and the solutions confirm that at lower rates the Hadley cells become wider, and the thermal fronts sharper and double. In the absence of any drag, the cells disappear and a thermal wind prevails globally. But in the absence of a background static stability, the cells become more intense and create their own stable temperature field. For normal parameter values, the Hadley cells adhere to the theoretical form as the rotation rate increases, except when their width falls below 3 of latitude. Furthermore, when the heated layer is thin and the jets are confined aloft, the cells develop vertically bimodal amplitudes, while remaining deep and exhibiting the usual widths. The basic 3-D terrestrial cases examine the role of the heating rate, static stability, surface drag, and rotation rate on the flow character. The mean jets exist within a limited latitudinal range, with their location being as much dependent on the heating amplitude as on the heating distribution. When the background static stability is absent, the standard circulation theory becomes less valid as the cells and baroclinic instability become more intense and act together to stabilize low and middle latitudes. However, when the drag is reduced, the baroclinic instability becomes much weaker and confined to lower levels because of suppression by the jet’s stronger barotropic component. Other forms of baroclinic instability can be produced by creating double-jet flows, either by increasing the rotation rate or by adding an extra source of baroclinicity in low latitudes. The transitional jovian cases examine how the multiple jets behave as the active layer is varied between thick and thin for the LIN and EXP structures. In all cases, the jet widths remain constant with latitude, but their amplitudes vary, peaking either in low or middle latitudes depending on how the baroclinicity is distributed. An extra baroclinicity in low latitudes produces a jet whose barotropic instability can drive an equatorial superrotation, regardless of layer thickness. The eddy-driven jets have a similar dynamics for all layer thicknesses but, unlike the steady LIN jets, the EXP jets also migrate equatorward and, on rare occasions, poleward.