Lj. Han et M. Gurnis, How valid are dynamic models of subduction and convection when plate motions are prescribed?, PHYS E PLAN, 110(3-4), 1999, pp. 235-246
A detailed comparison between fully dynamic and kinematic plate formulation
s has been made in models of mantle convection. plate velocity is computed
self-consistently from fully dynamic plate models with temperature- and str
ess-dependent viscosity and preexisting mobile faults. In fully dynamic mod
els, the now is driven solely by internal buoyancy, while in kinematic mode
ls the now is driven by a combination of the prescribed surface velocity an
d internal buoyancy. Only a temperature-dependent viscosity. close to the e
ffective viscosity determined from the fully dynamic models, is used in the
kinematic models. The two types of models give very similar temperature st
ructures and slab evolutionary histories when the effective viscosity and s
urface velocity are nearly identical. In kinematic plate models, the additi
onal work introduced by the prescribed velocity boundary condition is appar
ently dissipated within the lithosphere: and has little influence on the co
nvection under the lithosphere. In models with periodic lateral boundary co
nditions, slabs sink into the lower mantle at an oblique angle and this con
trasts with the vertical sinking which occurs with reflecting boundary cond
itions. Models show that we can simulate Fully dynamic models with kinemati
c models under either periodic boundary conditions or reflecting boundary c
onditions. (C) 1999 Elsevier Science B.V. All rights reserved.