THERMODYNAMIC MODELING OF EPIDOTE-AMPHIBOLITE ASSEMBLAGES IN METAMORPHOSED CUMULATE GABBROS FROM THE UPPER ALLOCHTHON, MID-NORWAY: STRATEGIES TO EXTRACT A P-T PATH FROM STRONGLY ZONED PHASES

The Upper Allochthon in the Scandinavian Caledonides contains Early Ordovician ophiolite fragments consisting of basalts and cumulate gabbros. The gabbros have the metamorphic assemblage: hornblende-plagioclase-diopside-epidote. The hornblende has actinolite cores, epidote has Al-rich cores and Fe-rich rims, commonly with three distinct compositional layers, and plagioclase zoning seems chaotic.

Modeling was done using Perple_X to try to understand the prograde metamorphic P-T path. First, green hornblende rims on actinolite cores, and little-zoned diopsides, were assumed to be prograde, but different parts of epidote and plagioclase were tested in models because it was unclear which were in equilibrium during peak prograde conditions. We also tested average plagioclase with the idea that deformation-induced recrystallization would have exposed all plagioclase compositions to the grain boundary metamorphic fluid. This required a large array of models, first applied in an exploratory way and then comprehensively. Mineral compositions were adjusted to stoichiometric solution model constraints, Fe³⁺ was estimated for diopside and amphiboles, and the mineral compositions combined to make model rocks.

Most models yielded inconsistent, geologically nonsensical results. Only those using average plagioclase and epidote mantles yielded consistent results for peak prograde conditions in 8 of 9 samples (~525°C, 6 kbars). These and models using modified bulk rock compositions indicated low-Fe epidote and albite should have been stable during early prograde conditions. New models tested this using diopside, the most actinolitic amphibole, the most Fe-poor epidote, and the most albitic plagioclase found in mineral cores and inclusions. These yielded amazingly consistent results (considering the poor constraints) among all 9 samples (~350°C, 6 kbars). Lastly, models indicated that the thin Fe-rich epidote rims may have resulted from isothermal decompression from peak conditions, dissolving unstable Fe-poor and precipitating Fe-rich epidote rims. Careful petrography, coupled with a range of hypothetical model trials, can yield consistent and geologically plausible P-T estimates—results unlikely to be obtained by merely calculating a P-T estimate from, say, two core compositions.