Thermal evolution of the subduction interface: Coupled petrologic and geodynamic study of high-pressure rocks of the Rio San Juan Complex, Dominican Republic

Pressure-temperature (P-T) estimates from subduction-related metamorphic rocks such as eclogites and blueschists are often used to constrain the thermal conditions of fossil subduction zone plate interfaces. However, the metamorphic rock record indicates peak temperatures 100-300°C warmer on average than those predicted by geodynamic models for modern subduction zones.. To shed light on the difference in the fossil and modern subduction zone thermal structures, we compare newly acquired P-T estimates using quartz-in-garnet and zircon-in-garnet elastic thermobarometry combined with Zr-in-rutile thermometry to newly constructed geodynamic models for the thermal evolution of the Rio San Juan Complex, Dominican Republic. The geodynamic models are 2-D coupled kinematic-dynamic models that use the fossil subduction parameters, such as time-dependent convergence velocity and plate age. Global plate reconstruction models provide constraints on these parameters. Data from regional geological and petrological studies constrain the duration of subduction and the thermal history of the subducting plate. New analyses on an eclogite sample reveal a counter-clockwise pressure-temperature (P-T) path, with peak conditions at ~1.7 GPa and ~650ºC. Previous thermobarometric studies on the same sample indicate similar temperatures but significantly higher peak pressures, up to ~2.3 GPa, during the early stage of subduction around 110–104 Ma. In contrast, lower-grade blocks indicate isobaric cooling to ~400ºC and clockwise P-T paths during a later stage (80–62 Ma). Our thermal modeling results indicate that subduction initiation between two relatively young tectonic plates (< 30 Ma) can explain the relatively low-P, high-T data from both new and previous analyses. However, the highest pressures derived by the previous study require a rapid deepening of the maximum depth of slab-mantle decoupling during the early stages of subduction. Such deepening can be explained by a plate velocity increase around 110 Ma. Our modeling results further indicate the subduction rate increase results in significant cooling of the interface at a given depth, consistent with the isobaric cooling observed in the rock record. Migration of the spreading center across the study area may explain the change from counter-clockwise to clockwise paths, and higher temperature gradients of some of the blocks.