Modeling major element variation in primary melts from a two-component mantle: application to the E-MORB and N-MORB relationship at the EPR
Abstract
A wealth of published (and unpublished) geochemical data exists for the 9°-10°N segment of the East-Pacific Rise (EPR) providing an unprecedented opportunity to investigate which processes control geochemical variation in erupted lava flows. Recent models that attempt to explain the range of trace element and isotopic characteristics in N- to E-MORB propose they result from mixing of end-member melts from a two-lithology mantle (e.g. Stracke and Bourdon, 2008; Waters et al., 2011), followed by fractional crystallization processes. Smith et al. (2000) defined E-MORB as K/Ti>11; with N-MORB K/Ti<11. Thorough sampling of the EPR and analysis of major element variation shows a distinctive increasing trend in K/Ti ratios and other incompatible element ratios with decreasing MgO in E-MORB that cannot be accounted for by fractional crystallization. Clearly mixing is involved in the production of these melts, however it would be expected that the production of the enriched member would result in the same fractionation trend shown by the N-MORB compositions from initial primitive MgO compositions (~9.5 wt %) to evolved MgO compositions (~6.5wt%). At any one enriched composition the range of MgO is limited to less than 1wt%. We propose that the lack of primitive MgO compositions for the most enriched melts (end-member mixing) is due to the influence of low MgO pyroxenite melts on the deepest melts in the mantle. Some of these enriched, evolved melts are allowed to escape mixing with shallower, more voluminous and relatively depleted peridotite melts via channelized melt transport. We present new results from a parameterization of major element melt compositions for two-lithology pyroxenite and peridotite melting. The model is based on simple 2D corner flow of the mantle beneath a spreading ridge with thermal advection-diffusion and the latent heat budget for the two-component melting. We use this model to predict the range of possible primitive melt compositions available beneath the ridge including spatial constraints on the origin of E-MORB end-member melts.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFM.V23B2568W
- Keywords:
-
- 1037 GEOCHEMISTRY / Magma genesis and partial melting;
- 8410 VOLCANOLOGY / Geochemical modeling;
- 8416 VOLCANOLOGY / Mid-oceanic ridge processes