Insight into the sensitivity of the global moist isentropic streamfunction can be gained by using the statistical transformed Eulerian mean (STEM) circulation to parameterize the streamfunction in terms of zonally averaged second order statistics. Since the STEM has a relatively simple functional form, we can calculate the sensitivity of the STEM by calculating functional derivatives that give the first order changes to the global moist isentropic streamfunction to perturbations in the circulation statistics.
These functional derivatives are interpreted in the relatively simple context of midlatitude surface perturbations in an idealised moist model which incorporates the latent heating of water vapor but not its radiative effects. These experiments show that the zonal mean thermodynamic response to surface heating raises the streamfunction onto higher isentropes near the surface anomaly, but that changes in the eddy covariance are responsible for shifting the circulation maximum poleward. These changes are interpreted in terms of the functional derivatives, demonstrating how the functional derivatives of the STEM conveniently reveal the background sensitivity of the moist isentropic circulation to perturbations.