The maintenance of northern summer stationary waves is studied with da
ta from a 15-year integration of the general circulation model (GCM) e
xperiment performed at the Geophysical Fluid Dynamics Laboratory. The
model has relatively high resolution (rhomboidal 30 wavenumbers, 9 ver
tical levels) and simulates the summertime stationary waves reasonably
well. A steady, linear, baroclinic model is used to understand the va
rious forcing mechanisms for the northern summer stationary waves. The
linear model response to global diabatic heating is found to play a d
ominant role in maintaining the summertime stationary waves in the GCM
, especially in the subtropics. This response to diabatic heating show
s a baroclinic structure in the vertical with a node at about sigma =
0.5. On the other hand, stationary nonlinear interaction terms are fou
nd to be largely responsible for the extratropical, equivalent barotro
pic stationary wave features. It is hypothesized that this nonlinear i
nteraction is a result of the thermally induced stationary waves inter
acting with the local orography. The direct linear response to orograp
hy is found to be rather insignificant, however. Transient vorticity a
nd heat fluxes also tend to play a negligible role in explaining the s
ummer stationary wave patterns. Further decomposition of the linear mo
del response to global diabatic heating indicates that the response to
the Indian monsoon and the western Pacific heat source is of primary
importance in determining the global stationary wave pattern. This lar
ge heat source not only determines the stationary flow features locall
y, but also remotely controls the flow structure over the whole Pacifi
c, North America, and the Atlantic region. Thus, variabilities in the
Indian monsoon and the western Pacific heating may exert a strong infl
uence on the global climate variability.