Aspects of Eddy Momentum Fluxes in the General Circulation of the Troposphere

Abstract

This thesis investigates the roles of eddy momentum fluxes (EMFs) in several aspects

of the tropospheric circulation. The thesis begins by examining how large-scale oro-

graphic forcing and ENSO-like equatorial heating affect horizontal EMFs in a multi-

layer primitive equation model (the GCM). For the orographic forcing a transition is

seen from a linear regime, in which the response is dominated by zonal propagation,

to a nonlinear regime, in which the response mostly propagates meridionally. The

dynamics of this transition are investigated by comparing the behavior of the GCM

with simpler models and it is also shown that the transition leads to relatively larger

stationary EMFs in the nonlinear regime. In the equatorial heating experiments the

winds in the deep tropics accelerate as the strength of the heating is applied and even-

tually superrotate. The superrotation is driven by the EMFs due to the stationary

waves excited by the heating, and is closely related to the breakdown of the GCM’s

linear response.

Complementing these, the second half of the thesis consists of two projects which

study the relationship between EMFs and mid-latitude jets. First, the EMFs in

the lower troposphere of two idealized models as well as in a reanalysis dataset are

investigated. The pattern of lower tropospheric EMFs is remarkably similar across

the datasets, with even a two-layer quasi-geostrophic (QG) model capturing the basic

structure seen in the reanalysis data of EMF divergence in the center of the jet at

slow phase speeds and EMF convergence at fast phase speeds. A physical picture of

the underlying dynamics is presented based on the dynamics of the QG model.

Finally, the Fluctuation-Dissipation Theorem (FDT) is used to predict how the

two-layer QG model responds to small, zonally-symmetric perturbations. It is found

that the FDT can accurately predict the models response to barotropic perturbations

(perturbations with the same magnitude and sign in each layer) but is less successful

at predicting the response to baroclinic perturbations (perturbations with the same

magnitude but opposite sign in each layer). Some alternative strategies for producing

better FDT estimates are investigated, and these may be relevant for studies with

more comprehensive atmospheric models.