Quantifying baroclinic adjustment in the atmosphere from data

The baroclinic adjustment hypothesis as proposed by Smagorinsky (1963) and developed by Green (1970) and Stone (1978) assumes that the climatological mean state of the atmosphere stays close to neutral with respect to baroclinic instability. Since there are more than one way of achieving neutrality, this assumption alone does not uniquely constrain the climate state. Furthermore, while the observed zonal mean state may be the result of baroclinic adjustment, the lack of “initial condition” to compare with makes it difficult to quantify the efficacy of the adjustment.

We attempt to quantify the extent of baroclinic adjustment in the atmosphere by comparing the observed zonal-mean state with that of the eddy-free reference state. The eddy-free reference state is constructed from the rearrangement of the observed 3D state into a zonally symmetric state enforcing the material conservation of potential vorticity and potential temperature, as well as Kelvin's circulation and mass. (This amounts to a 2D geostrophic adjustment problem.) The reference state obtained from the NCEP reanalysis data shares many salient features with the straight zonal mean of the same data, but the difference between the two states brings out the role of baroclinic eddies in the latter: accelerating the flow near the surface and within the storm track regions, while decelerating the upper-tropospheric flow in the subtropics. We will then compare the stability of the two states and assess the effect of baroclinic adjustment.