MODELING HIGH TEMPERATURE AND PRESSURE PHASE EQUILIBRIA, AND PARTIAL MELTING OF GABBROIC GNEISS, MILFORD SOUND, NEW ZEALAND

Granulite facies hornblende gabbro and diorite gneisses, of Pembroke Valley New Zealand, were metamorphosed at lower crustal pressures of up to 14 kbar and experienced partial melting. Partial melting produced garnet-bearing trondhjemite patches and veins that have been interpreted as products of water-saturated hornblende breakdown. This origin was tested by modeling water-saturated metamorphism and partial melting of a hornblende gabbro (observed assemblage=Amp + Pl + Czo + Grt + Bt + Chl + Qtz) in the system MnNaCaKFeMgAlSiH. Three P-T pseudosections (MnNaCaKFeMgAlSiH with zoisite; MnNaCaKFeMgAlSiH with clinozoisite; and MnNaCaKFeFe³⁺MgAlSiH with clinozoisite solid solution) illustrate the effects of ferric iron and epidote group minerals on phase relations. Without ferric iron, only zoisite stability can be predicted and the peak metamorphic minerals (hornblende + garnet + ‘epidote’ + plagioclase + melt) are stable between 10.8 and 14.5 kbar at 625 to ca. 725^oC. Without ferric iron and forced clinozoisite stability, the peak metamorphic minerals are only stable between 13.4 and 13.7 kbar at 640 to ca. 660^oC. With ferric iron, clinozoisite-epidote solid solution can be modeled more realistically and the peak minerals are stable between 8.7 and 12.6 kbar at 650 to >850^oC. Addition of ferric iron also reduces the range of predicted muscovite stability, lowers the minimum pressure for paragonite stability, and decreases the minimum pressure and increases maximum temperature of epidote group mineral stability. Models predict that garnet mode increases, and plagioclase and biotite modes decrease dramatically just above the solidus, compatible with peritectic garnet growth. Amphibole remains stable and it’s mode increases across the solidus for this bulk composition, unlike prior studies that predicted consumption of amphibole. Preliminary results from pseudosections suggest that ferric iron-bearing chemical system can predict peak metamorphic conditions, partial melting, and growth of peritectic garnet; and that phase equilibria models should include ferric iron in order to reasonably predict mineral assemblages.