Potential vorticity sources and sinks in ocean and atmosphere (Invited Presentation)

Potential vorticity has a 'geography' that expresses the rotational constraint on waves,instability, eddies and mean circulation. The spherical Earth and large-scale circulation (through its shear and density structure) provide the basic geography of pv, along with the singular contribution due to potential density gradients at boundaries. The ocean and atmosphere have many analogous dynamical structures, yet differ greatly in their pv sources and sinks. Here we describe boundary and convective sources of pv.

Experiments (numerical and laboratory-) aimed at the ocean involve injection of pv from the boundaries of the fluid as a dominant effect. These range from gcm investigation of the effect of continental slopes on the wind-driven circulation to lab- and numerical simulation of stratified, rotating flow over topographic ridges and bumps. The familiar flow past a tall, isolated mountain causes a shedding of predominantly cyclonic vorticity produced at the intersection of the sloping boundary and interior stratification. At finer scale, internal hydraulics and internal wave dynamics also act as pv sources.

In the atmosphere, at much larger scale, we describe a life cycle of flow past an isolated mountain, which (in a weakly damped world) fills the globe with annular banded jets and a pv staircase; this could plausibly be a model for the annular modes prominent in extratropical variability.

PV mixing, toward homogenization, is a key property that distributes the 'sharp' pv signals from boundaries, often to fill a much greater volume of fluid (for example, an entire ocean basin). A second key property is the radiation of 'b plumes' (pseudo-) westward from an intense disturbance; these are essentially low-frequency Rossby waves with appropriate vertical mode structure. The b plumes, including their topographic cousins, fill out the general circulation. A third key property is the barrier effect of intense jets, which develop strong pv gradients at some levels and homogenization at others. The jets can be energetically dominant features, but they interact through radiation and advection with a much larger region.

Convection penetrates deep into the ocean at a few sites, and dense, downslope flows occur in response. By contrast, the troposphere is widely populated with cumulus towers that traverse the entire layer. We can model the forcing of larger scale circulation by such invasions by noting the dipole or tripole pv structure induced by buoyancy forcing. Radiation of b plumes to a distance depends on these distributions.