Recent work has shown that the zonal, equatorially symmetric, time-varying
part of a model of the flow at the surface of the Earth's core can be well
explained by only two standing waves, and that by making certain assumption
s these waves may be inverted for rms B-s (the component of field pointing
away from the rotation axis) and a quantity parametrizing friction or excit
ation (F) of the waves. Here, we discuss the two-wave fit, and describe the
implications of models of rms B-s and friction/excitation for the dynamic
state of the core, for the torque balance on axial cylinders, and for recen
t numerical simulations of the geodynamo. We find several possible explanat
ions for why only two standing waves are needed to fit the data, including
the possibility that it is due to the resolution of the core flow model rat
her than conditions within the core itself. We find that the fits of rms B-
s and F suggest that the role of inertia should not be discounted in the co
re, and that care should be taken in constructing geodynamo simulations so
that the effective friction at the core-mantle boundary does not swamp the
inertial term. A ratio of the magnitude of the two appears to be O(1) in th
e Earth's core: we believe that, ideally, a numerical model of the Earth's
core should reproduce this result.