The Northeast Japan arc, a mature volcanic arc with a back-arc marginal basin (Japan Sea), is located on a convergent plate boundary along the subducting Pacific plate and the overriding North American plate. From a compilation and analysis of stratigraphy, radiometric age and data on erupted magma volumes, 176 eruptive episodes identified from 69 volcanoes so far, indicate that notable changes in eruption style, magma discharge rates and distribution of eruptive centres occurred around 1.0 Ma. Before ca.1.0 Ma, large-volume felsic eruptions were dominant, forming large calderas in the frontal arc, a region of low crustal strain rate. After ca. 1.0 Ma to the present, the calc-alkaline andesite magma eruptions in the frontal and rear arcs, synchronous with crustal shortening characterized by reverse faulting, resulted in stratovolcano development along narrow uplifted zones. Although, it is widely assumed that magma cannot rise easily in a compressional setting, some of the magma stored within basal sills could be extruded where N–S-trending uplifted mountains bounded by reverse faults formed since about ca.1.0 Ma.

Felsic magmatism in the southern part of Himachal Higher Himalaya is constituted by Neoproterozoic granite gneiss (GGn), Early Palaeozoic granitoids (EPG) and Tertiary tourmaline-bearing leucogranite (TLg). Magnetic susceptibility values ($\lt$3 ×10$^{−3}$ SI), molar Al$_2$O$^3$/(CaO+Na$_2$O+K$_2$O) ($\geq$1.1), mineral assemblage (bt–ms–pl–kf–qtz ± tur ± ap), and the presence of normative corundum relate these granitoids to peraluminous S-type, ilmenite series (reduced type) granites formed in a syncollisional tectonic setting. Plagioclase from GGn (An$_{10}$–An$_{31}$) and EPG (An$_{15}$–An$_{33}$) represents oligoclase to andesine and TLg (An$_2$–An$_{15}$) represents albite to oligoclase, whereas compositional ranges of K-feldspar are more or less similar (Or$_{88}$ to Or$_{95}$ in GGn, Or$_{86}$ to Or$_{97}$ in EPG and Or$_{87}$ to Or$_{94}$ in TLg). Biotites in GGn (Mg/Mg+Fe$^t$ = 0.34–0.45), EPG (Mg/Mg+Fe$^t$ = 0.27–0.47), and TLg (Mg/Mg+Fe$^t$ = 0.25–0.30) are ferribiotites enriched in siderophyllite, which stabilised between FMQ and HM buffers and are characterised by dominant 3Fe$\rightleftarrows$2Al, 3Mg$\rightleftarrows$2Al substitutions typical of peraluminous (S-type), reducing felsic melts. Muscovite in GGn (Mg/Mg+Fe$^t$ = 0.58–0.66), EPG (Mg/Mg+Fe$^t$ = 0.31−0.59), and TLg (Mg/Mg+Fet = 0.29–0.42) represent celadonite and paragonite solid solutions, and the tourmaline fromEPG and TLg belongs to the schorl-elbaite series, which are characteristics of peraluminous, Li-poor, biotite-tourmaline granites. Geochemical features reveal that the GGn and EPG precursor melts were most likely derived from melting of biotite-rich metapelite and metagraywacke sources, whereas TLg melt appears to have formed from biotite-muscovite rich metapelite and metagraywacke sources. Major and trace elements modelling suggest that the GGn, EPG and TLg parental melts have experienced low degrees (∼13, ∼17 and ∼13%, respectively) of kf–pl–bt fractionation, respectively, subsequent to partial melting. The GGn and EPG melts are the results of a pre-Himalayan, syn-collisional Pan-African felsicmagmatic event, whereas the TLg is a magmatic product of Himalayan collision tectonics.