Thermochemical models of oceanic upper mantle

The thermal state of the oceanic upper mantle is reassessed by developing new progressively more sophisticated models of the thermal evolution of oceanic lithosphere. Model Properties are constrained by experimental mineral physics data for the thermal properties of the mantle, statistical analyses of the major geophysical observables (topography and surface heat flow), and sensitivity analyses of physical parameters. Predictions of geothermal structure, plate thickness, subsidence rates, and net sea floor heat flow are discussed. Sensitivity analyses demonstrate that physically simple plate cooling models cannot fit geophysical observations unless thermal expansivity is substantially lower than experimental estimates for forsterite. More physically complete models resolve improved fits to geophysical observations with less artificial adjustments in physical properties. The inclusion of an insulating oceanic crust is found to significantly impact the cooling behavior of oceanic lithosphere. To estimate the total amount of heat vented to the oceans by ventilated hydrothermal circulation, we integrate the deficit between modeled and measured heat ow over young sea floor. A model with insulating oceanic crust predicts a loss of about 6.6 TW and the fraction of heat extracted on ridge axes is about 50% of the total. Lastly, the thermal structure of oceanic lithosphere is linked to the chemistry of Ocean Island Basalts (OIB) via the 'LID effect'. We developed a two-dimensional numerical model of multi-phase coarsening, dffusive trace element partitioning, and near-fractional melting. The model is applied to the generation of melts in the OIB source, and results are compared to observed correlations in La/Sm and Sm/Yb with sea floor age. Using the model we explain Rare-Earth element (REE) systematics in Global OIB, constrain the LAB temperature to be approximately 1225⁰ C, suggest that experimental partition coeffcients may be a few tens of percent away from appropriate effective values in OIB melting, suggest that the mean grain radius in the OIB source is on the order of 2.5-5 mm, demonstrate that the OIB source is in a state of extreme disequilibrium, show that the REE composition of the bulk OIB source is both heavily enriched and similar to the MORB source, and we show that the potential temperature of the OIB sourceis likely < 1400⁰C.