| Abstrakt: |
A numerical code has been developed to track the distribution of trace elements in crustal rocks undergoing melting. The model handles diffusion with moving boundaries and accounts for the processes of diffusion, dissolution and precipitation in a partially molten system. Among the various input parameters for modelling, source composition (i.e. modal abundance) variations, diffusion coefficients and partition coefficients are found to exert a significant control on the melt chemistry. The other inputs such as melt reaction stoichiometry, kinetics of melting and grain size of protolith have lesser influence. Exploration of the general behaviour indicates that for systems in which disequilibrium melting of the kind considered in this paper occurs, trace element concentrations may be used to constrain the composition of the protolith or the timescales of melting, depending on the specific circumstances. After exploring some general features of melting in a pelitic system, the model is applied to calculate trace element distributions in migmatites from the Lesser Himalayan rocks in Sikkim, India. We focus on the distribution of trace elements during the initial stages of melt formation. These partially molten rocks show disequilibrium distribution of trace elements, and the numerical code is capable of quantitatively reproducing many of the observed patterns. The results of the modelling indicate that the observed melts in this zone were formed within 50,000 years and that segregation of melts (into leucosome and restite) was complete between 50,000 and 250,000 years. These short timescales may point to deformation-enhanced melt segregation at least on a hand specimen scale. It is important to distinguish between timescales of segregation over these scales and timescales of removal of melt on an outcrop scale to form plutons-the latter, requiring higher degrees of melting and larger distances of migration, take longer. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink