Heterogeneity in mid-ocean ridge sources

Abstract

This talk will review some theoretical considerations concerning the melting of heterogeneous sources and the implications for interpretation of melt inclusion data, particularly olivine-hosted inclusions from mid-ocean ridge environments. We have apparently adequate models that assume a locally or globally homogenous source and can explain major and trace element chemistry of basalts and correlations with physical variables like melt flux and axial depth. Such models are the basis of the consensus view of the average thermal and compositional state of the upper mantle and the range of their variability. However, a variety of data (most notably from melt inclusion diversity within samples and suites) and lines of reasoning (such as the need for recycling to maintain fertility of the source) require that heterogeneity exists within particular melting regimes as well as among them. We can classify heterogeneity by scale. There is some small scale (perhaps ~10 cm) below which the source melts as if it were effectively homogeneous and some large scale (perhaps ~ 1 km) above which the source components can be treated as chemically and thermally independent and their melts simply mixed; these scales may be different for different elements and lithologies. The challenge is to develop tools that predict behavior at intermediate scales where incomplete thermal and/or chemical interaction occur and hence standard mixing theory is confounded by complex feedbacks between composition and melting behavior and between melting and equilibration. Additional challenges center on the kinetics of interactions during melt migration through heterogeneous media.

Additional details