Ma. Ravine et Jp. Morgan, GEOID EFFECTS OF LATERAL VISCOSITY VARIATION NEAR THE TOP OF THE MANTLE - A 2D MODEL, Earth and planetary science letters, 119(4), 1993, pp. 617-625
Whereas the oceanic lithosphere is underlain by a low-viscosity layer,
the asthenosphere, it appears unlikely that this layer is well develo
ped beneath continental cratons. This difference could represent a lat
eral viscosity variation of two to three orders of magnitude near the
top of the mantle. Previous mantle flow/geoid models have typically ne
glected this variation. To investigate the impact of such a lateral st
rength variation on mantle flow and the resulting geoid, we solve for
density-driven flow in a 2D box of uniform-viscosity fluid. A dichotom
ous lateral viscosity variation near the top of the mantle is modeled
by having a no-slip boundary condition over a section of the top of th
e box, while allowing the remainder of the top to be free-slip. For va
rying fractions of no-slip on the upper boundary, and various values o
f the wavenumber, depth and phase of the driving density anomaly, mode
l geoids are found. In contrast to viscous flow with only a depth-depe
ndent viscosity, mode coupling in the geoid is observed. This occurs p
rincipally at a wavelength equal to the lengthscale of the lateral vis
cosity structure, with amplitudes as large as 20-40% in some cases. Th
us, coupling can produce a significant signal in the geoid at a wavele
ngth longer than that of the driving load. This result suggests that t
he accuracy of models of the geoid that assume a solely radial viscosi
ty structure may be limited to the 60-80% level found in recent studie
s [1,2]. We speculate that some of the Earth's low-order geoid is due
to flow coupling resulting from the degree two to five components of l
ateral ocean-continent structure.