Petrology of a non-classical Barrovian inverted metamorphic sequence from the western Arunachal Himalaya, India

In this study, we reconstruct the inverted metamorphic sequence in the western Arunachal Himalaya using combined structural and metamorphic analyses of rocks of the Lesser and Greater Himalayan Sequences. Four thrust-bounded stratigraphic units, which from the lower to higher structural heights are (a) the Gondwana rocks and relatively weakly deformed metasediments of the Bomdila Group, (b) the tectonically interleaved sequence of Bomdila gneiss and Bomdila Group, (c) the Dirang Formation and (d) the Se La Group are exposed along the transect, Jira–Rupa–Bomdila–Dirang–Se La Pass. The Main Central thrust, which coincides with intense strain localization and the first appearance of kyanite-grade partial melt is placed at the base of the Se La Group.Five metamorphic zones from garnet through kyanite, kyanite migmatite, kyanite-sillimanite migmatite to K-feldspar-kyanite-sillimanite migmatites are sequentially developed in the metamorphosed low-alumina pelites of Dirang and Se La Group, with increasing structural heights. Three phases of deformation, D₁–D₂–D₃ and two groups of planar structures, S₁ and S₂ are recognized, and S₂ is the most pervasive one. Mineral growths in all these zones are dominantly late-to post-D₂, excepting in some garnet-zone rocks, where syn-D₁ garnet growths are documented. Metamorphic isograds, which are aligned parallel to S₂ were subsequently folded during D₃. The deformation produced plane-non-cylindrical fold along NW–SE axis.In the garnet-zone, peak metamorphism is marked by garnet growth through the reaction biotite + plagioclase → garnet + muscovite. An even earlier phase of syn-D₁ garnet growth occurred in the chlorite stability field with or without epidote. In the kyanite-zone metapelites, kyanite appeared via the pressure-sensitive reaction, garnet + muscovite → kyanite + biotite + quartz. Staurolite was produced in the same rock by retrograde replacement of kyanite following the reaction, garnet + kyanite + H₂O → staurolite + quartz. These reactions depart from the classical kyanite- and staurolite-isograd reactions in low-alumina pelites, encountered in other segments of eastern Himalaya. In the metapelites, just above the kyanite-zone, melting begins in the kyanite field, through water-saturated and water-undersaturated melting of paragonite component in white mica. Leucosomes formed through these reactions are characteristically free of K-feldspar, with sodic plagioclase and quartz as the dominant constituents. With increasing structural height, the melting shifts to water-undersaturated melting of muscovite component of white mica, producing an early K-feldspar + kyanite and later K-feldspar + sillimanite assemblages and granitic leucosomes.Applications of conventional geothermobarometry and average P–T method reveal near isobaric (at P  8 kbar) increase in peak metamorphic temperatures from 550 °C in the garnet-zone to >700 °C for K-feldspar-kyanite-sillimanite-zone rocks. The findings of near isobaric metamorphic field gradient and by the reconstruction of the reaction history, reveal that the described inverted metamorphic sequence in the western Arunachal Himalaya, deviates from the classical Barrovian-type metamorphism. The tectonic implication of such a metamorphic evolution is discussed.