Ultra high pressure (UHP) terrains: Lessons from thermal modeling

Abstract

The thermal history of subducted/exhumed crustal fragments is explored to depths of 125 km using a forward modeling approach. The finite element code uses an adaptive gridding technique that allows deformation and differential movement of subgrids to simulate tectonic mass flow. The geometry and velocity of plate convergence and the large-scale strain partitioning (for example, subduction channel, mobile tectonic fragments) are described explicitly as functions of time. The evolution of metamorphic conditions during descent and subsequent exhumation of a terrane within the subduction channel is computed for different kinematic scenarios. These include a range of velocities, relative to a stable continental block, for the subducting slab, the exhuming (UHP) fragment, and a subduction channel of given width. Though essentially generic in emphasis, our models are designed to help understand real UHP occurrences, most especially their exhumation history. Boundary conditions and the geometric set-up were chosen according to three case studies of known UHP terranes (eastern China, Western Norway, Western Alps). Simulations indicate that the temperature reached at a given depth during the subduction stage is crucially dependent on the convergence rate, width of the subduction channel, location of the fragment within the subduction channel, and on rates of shear heating and convection of the subcontinental mantle. The P-T-t path experienced by a tectonic fragment during its exhumation from Pmax is determined chiefly by its size (thickness) and exhumation velocity; the rate of ongoing subduction is less relevant given the rapid exhumation required to preserve UHP assemblages. The depth at which Tmax is reached by an exhuming UHP fragment and the dP/dT slope of its exhumation path may critically affect the extent of irreversible phase transitions during decompression. Comparing data for the three UHP case studies with corresponding simulation results indicates general agreement of the P-T paths. Predicted T-t evolution paths are also compatible with well-constrained mineral chronometry from the study areas. This agreement between field data and simulation results leads to a number of observables that may be useful in relating UHP occurrences to their evolution. For example, no metamorphic field gradients are expected for fragments of less than 5 km thickness, and the spread in mineral ages reflecting the high-pressure stage is predicted to be less than 1 Ma. Therefore, observed variations in metamorphic conditions for the stage near Tmax and P(Tmax) may reflect real differences of evolution in very large fragments only; and the same is valid for a substantial spread in metamorphic age data pertaining to stages near Pmax or Tmax. Tectonic extrusion of fragments along a subduction channel is not the only mechanism proposed to explain UHP occurrences. However, the broad agreement between several test cases and the present simulation results lend support to the viability of the kinematic scenario adopted. Where field evidence indicates a similar emplacement mechanism and allows the (initial) size of the UHP fragment to be inferred, the MELONPIT model provides a powerful integrative tool to produce a quantitative picture of the dynamics of UHP terranes.