V44A-03:
A record of lower crustal chamber sizes from Icelandic olivine-hosted melt inclusions

Thursday, 18 December 2014: 5:30 PM
John Maclennan, University of Cambridge, Cambridge, United Kingdom
Abstract:
Melt inclusions hosted in olivines from single Icelandic eruptions of basalt have highly variable compositions, particularly in terms of their incompatible element concentrations, concentration ratios and Pb-isotopes. This compositional diversity reflects variation in the composition of mantle melts supplied to individual volcanic plumbing systems. The variance of inclusion compositions decreases as the forsterite content of the host olivine decreases: a relationship that is readily understood in terms of concurrent mixing and crystallisation of primitive melts. The availability of a large number of melt inclusion analyses from Icelandic olivines (>500) allows for robust quantification of the relative rates of melt cooling and passive tracer mixing in the lower crustal chambers where inclusion entrapment occurred. This coupling of cooling and mixing is likely to reflect convection in the melt layer of the chambers. The fact that the Icelandic observations preserve a record of both mixing and cooling indicates that the characteristic timescales of these processes are similar. A simple 1D model of convection in a basaltic sill was developed, with the cooling either controlled by temperature dependence of viscosity of the melt or by the rate of conduction of heat through the rocks of the chamber roof. The timescales of cooling of melt and mixing of passive tracers are proportional to different powers of the Rayleigh number of the sill. Therefore, models that match the Icelandic observations involve sills with a surprisingly small range of thicknesses: the model timescales of mixing and cooling are similar only if the thickness of the sill is <10 m. These sill thicknesses are similar to those proposed for the lower oceanic crust based on observations from ophiolites, and may indicate that small sills play a dominant role in the evolution of basaltic melts.