Abstract:
We investigate the effects of thermal diffusion on the evolution of mantle plumes by means of three-dimensional numerical modeling forward and backward in time. Mantle plumes are fed by a hot, low-viscous material from the thermal boundary layer. The material of the plumes is mainly advected toward the Earth's surface with some effects of thermal diffusion. However, the feeding can become weaker with time, and then thermal diffusion can take over and control the evolution of the plumes. Numerical experiments forward in time show that a week feeding of mantle plumes by the hot material from the boundary layer results in the diffusive disappearance of plume tails first and plume heads later. This is the most likely explanation for the seismically detected low-velocity mantle structures (mantle plumes) with prominent heads and almost invisible tails at midmantle depths. We develop restoration models (backward in time) to recover strong features of mantle plumes in the geological past after they have dissipated due to thermal diffusion and analyze effects of thermal diffusion and temperature-dependent viscosity on the reconstruction of the mantle plumes. We investigate the impact of thermal diffusion on the performance of our restoration (variational data assimilation) algorithm. For a given range of Rayleigh number Ra and two values of the viscosity ratio r (between the upper and lower boundaries of the model domain) we show that (1) the residuals between the temperature predicted by the forward model and that reconstructed by the backward modeling become larger and (2) the restoration process becomes poorer as Ra decreases and r increases. We assimilate temperature obtained from high-resolution seismic tomography data for the southeastern Carpathians and show that present diffused mantle structures can be restored to their prominent state in the Miocene times. We discuss the problems of smoothness of model input and output data, errors associated with the modeling, and some other challenges in the data assimilation for thermoconvective flow in the mantle. Copyright 2006 by the American Geophysical Union.
