9219434 Sacks The role of subduction and magmatism in the growth of the continental crust and the evolution of the crust/mantle system has long been recognized, but a detailed, process-oriented under- standing of the subduction cycle has been elusive. In this project the principal investigators will combine experimental petrology and geochemistry, observational seismology, isotope and trace element geochemistry and laboratory convection experiments with numerical approaches to the thermal structure and large-scale convection of the slab and mantle wedge, and porous flow models for melt and fluid migration in the arc environment. Heat flow and various analyses of seismic data, from primarily the Japan arc, will constrain the thermal structure and rheological properties used in the initial calculations (2-D and some 3-D) of flow in the slab and mantle, using a temperature-dependent viscosity. Models developed for Japan, where seismic constraints are strongest, will be used as a base for Central America and for the Cascades, where a young hot plate is subducting and conditions are far from steady state. Models for transport of hydrous fluids in the vicinity of the slab, and of melts through the mantle will be calculated using the stress and flow fields from the thermal and convection modeling. Plate scale mass transport delivers chemical elements derived from the continental crust, hydrosphere and lithosphere to the deep source of arc magmas; fluid and melt transport processes translate some fraction of these elements back to the surface in crust building magmas. Documentably slab-derived elements such as 10Be (1.5 Ma half-life), B and U (75 ka half-life in disequilibria series) will be incorporated as chemical tracers in fluid/melt migration models, with fluxes, transport paths and characteristic transport times constrained from chemistry of subconducting sediments and erupting lavas. Fluid partitioning studies on key elements will further const rain input to migration models. Predictions from coupled thermal, convection and transport models regarding spatial distribution, character and chemistry of lavas along the length and across the width of arcs will be compared with observed values, and evaluated against observed variations in subduction parameters from arc to arc to arrive at an integrated model of the subduction cycle and its role in crustal growth and evolution of the mantle. ***
