Abstract
The origin of orogenic andesitic magmas is tested by calculations of REE fractionation in hydrous melts derived from partial melting of subducted ocean basalt in eclogite facies. New data on the subsolidus phase proportions of basaltic eclogite, the enrichment of LREE in altered ocean basalts, and experimentally determined REE partition coefficients ( K D's ) between garnet and melt have been included in trace element fractionation equations. Non-modal melting of phases combined with variation in K D's during melting is a unique feature of these calculations. Variation of K D , melting proportions, initial proportion of subsolidus phases, degree of melting, and initial REE concentrations yield a wide range of input parameters that produce REE profiles in partial melts of basaltic eclogite matching REE profiles of some orogenic andesites. The positive correlation of REE concentration with silica content for many andesitic suites can be accounted for by non-modal melting if quartz (or a similar phase with low REE K D values) melts at a high melting proportion and garnet melts at a low melting proportion during the first stages of fusion. However, no mineralogic fractionation scheme can account for REE/silica systematics if REE K D values are linearly decreasing with increasing melting. Earlier workers who have used similar calculations to discredit the eclogite fractionation model have set overly strict, and sometimes incorrect, constraints concerning the range in REE K D values for garnet, the subsolidus proportions of phases in basaltic eclogite, and the relative concentrations of REE in subducted ocean crust undergoing partial melting.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have