Ph. Stone et B. Nemet, BAROCLINIC ADJUSTMENT - A COMPARISON BETWEEN THEORY, OBSERVATIONS, AND MODELS, Journal of the atmospheric sciences, 53(12), 1996, pp. 1663-1674
Isentropic slopes calculated from Northern Hemisphere analyses of the
zonal mean state of the atmosphere are compared with isentropic slopes
calculated from baroclinic adjustment theory. In midlatitudes, the is
entropic slopes are relatively close to the adjusted values in the lay
er from 2 to 7 km (800 to 400 mb). They also have very little variatio
n in latitude and season, which implies that baroclinic eddy fluxes su
pply a strong negative feedback to changes in isentropic dopes. The ve
rtical structure of the temperature field in the ''adjusted'' layers a
nd the location of the layers suggest that vertical eddy heat fluxes p
lay a significant role in this feedback process. Isentropic slopes are
also calculated from simulations of the current climate by the GISS M
odel II GCM and the NCAR CCM2. Both models have regions of apparent ba
roclinic adjustment similar to that in the Northern Hemisphere atmosph
ere. However, compared to the observations, the agreement of the simul
ations with baroclinic adjustment theory is not as good, and the isent
ropic slopes show stronger latitudinal and seasonal variations. The di
screpancies are associated primarily with errors in the models' meridi
onal temperature gradients in the lower troposphere in midlatitudes. T
he seasonal changes in these gradients are much larger than in the obs
ervations, particularly in the CCM2, even though the model simulations
were constrained by climatological sea surface temperatures.