HR: 14:20h
AN: V42F-03    [PDF]
TI: Modeling Metamorphism Using Pseudosections
AU: * Brown, M
EM: mbrown@geol.umd.edu
AF: University of Maryland, Lab for Crustal Petrology, Dept of Geology, College Park, MD 20742
AU: Johnson, T
EM: timj@geol.umd.edu
AF: University of Maryland, Lab for Crustal Petrology, Dept of Geology, College Park, MD 20742
AB: To maximize our potential to recover full information from mineral assemblages and microstructures in metamorphic rocks we must be able to model in a reliable manner mineral equilibria in a majority of systems of geological interest. As experimental and thermodynamic data for a wider range of compositional end-members have become available, and models for activity-composition relations have improved, so chemical systems for which was are able to produce equilibrium phase diagrams have become increasingly more complex and now more closely approximate natural rocks. Pseudosections are equilibrium phase diagrams constructed for fixed bulk compositions or bulk compositional ranges against varying intensive parameters (P, T, X, aH2O, etc.) using internally consistent sets of thermodynamic data (e.g. Holland and Powell, 1998, JMG). Pseudosections allow precise quantitative predictions to be made on multivariant stability fields of paragenetic associations; however, we must ask how close are these predictions to real values for intensive parameters of interest. Previous studies of subsolidus to suprasolidus metapelitic rocks from southern Brittany have used reaction microstructures and conventional thermobarometry to suggest the Variscan lower crust followed a clockwise P-T evolution (Jones and Brown, 1990, JMG; Brown and Dallmeyer, 1996, JMG). We construct pseudosections for the MnNCKFMASH system and use average P-T calculations to investigate in more detail metamorphic evolution of these rocks. For migmatites, phase relations predicted based on the superimposition of the P-T path inferred from microstructural relations among mineral phases onto pseudosections calculated for an average metapelite composition, contoured for proportions of L and Grt, are broadly consistent with those inferred from petrography. The sequential occurrence of Ky, Ky + St and Sil suggests prograde evolution to P $>$ 8 kbar at 625§C decreasing to P around 6 kbar at 650§C, and followed by increasing P and T to the metamorphic peak. A major melting step occurred at 750§C and 9 kbar by incongruent breakdown of Ms. At the metamorphic peak of 8 kbar and 800§C, ~25 mol % L and $>$20 mol % Grt are predicted from volatile phase-absent melting that also consumed Bt. Retrograde evolution began with decompression and cooling, allowing crystallization of melt and replacement of Grt by Bt + Sil; this was followed by near-isothermal decompression from around 6 kbar to 4 kbar. The decompression segment was associated with widespread development of Crd, which commonly occurs as large peritectic porphyroblasts with leucosome in interboudin partitions indicating a second episode of melt generation. Crd growth that accompanied decompression melting is only predicted in melt-depleted low-Mn rocks. Preserved bell-shaped spessartine profiles in Grt suggest that Mn was effectively removed from the reacting rock volume. Microstructural relations in upper amphibolite facies metapelites from the unit structurally overlying the migmatites suggest retrograde development of St, And and white mica in these rocks. Based on a pseudosection contoured for the quantity of H2O required to saturate the assemblage at P-T, we suggest that an influx of an H2O-rich volatile phase is required, which we infer to have been derived from crystallising melt in underlying migmatites and granites. The correspondence of the predictions from the pseudosections with classical thermobarometric estimates that account for all dilutents within participating phases is encouraging, suggesting that quantitative inferences are valid using either method. When used in conjunction with geochronology, accurate constraints may be placed on the P-T-t evolution of rocks and sophisticated tectonothermal models advanced.
DE: 3660 Metamorphic petrology
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting