Pulsed channel flow in Bhutan

We summarize our results from Bhutan and interpret the Greater Himalaya Sequence (GHS) of Bhutan, together with a portion of the underlying Lesser Himalaya Sequence, in the context of recently published channel flow models. For the GHS rocks now exposed in Bhutan the depth for beginning of muscovite dehydration melting (approximately 7508C at 11 kbar) and associated weakening of these rocks is constrained by geobarometry to be at about 35–45 km. The location of initial melting was down-dip and over 200 km to the north of Bhutan. Melt was produced and injected into ductilely deforming metamorphic rocks as they were extruded towards the south between the Main Central Thrust (MCT) and the South Tibetan Detachment zones. The lateral flow of low viscosity rocks at these depths occurred under southern Tibet between 22 Ma and 16 Ma. Subsequently, the channel rocks decompressed from 11 to 5 kbar (from 35 km to a depth of 15 km), but maintained high temperatures, between about 16 Ma and 13 Ma. The data from Bhutan are consistent with channel flow models if there were several pulses of channel flow. The first, between 22 and 16 Ma, produced the rock seen in the lower half of the GHS of Bhutan. A second pulse, which is cryptic, is inferred to have led to the uplift and exhumation of the MCT zone. A third, in central Bhutan, is exposed now as the hanging wall of the Kakhtang thrust, an out-of-sequence thrust that was active at 12–10 Ma. The latter two pulses likely broke around a plug at the head of the first pulse that was formed where the melt in the channel had solidified. The geodynamic models of Beaumont et al. (2001, 2004) and Jamieson et al. (2004) show that channel flow in the Himalaya was a likely consequence of building the Tibetan Plateau, and exhumation of the channel material was a consequence of focused erosion at the Himalaya front in concert with the channel flow. Jamieson et al. (2004) interpret most metamorphic and structural data of the central Himalayas in the context of a basic model called HT-1. We here review data from Bhutan (Fig. 1) that support a model of channel flow for formation of the Greater Himalaya Sequence (GHS). Inconsistencies between the basic model and our data can be rationalized by considering pulses of channel flow rather than the single, spatially and temporally continuous episode of channel flow as used in the basic model. The basic geodynamic model for the Himalaya calls for steady channel flow of thermally weakened rock in a channel from the Miocene to the present. This weakening is inferred to be due to the onset of melting in the midto lower crust; a few per cent of melt in a rock is known to lower the viscosity by several orders of magnitude (Rosenberg & Handy 2005). Because rock within the flowing channel, in comparison to the bounding plates, is relatively weak due to the presence of melt, when this melt solidifies (as leucogranite at about 7008C at 5 kbar; L of Fig. 2) the rock becomes stronger. Thus, the head, or the rheological tip, of the low viscosity portion of the channel is defined by the isotherm for this transition. The channel in the basic geodynamic model flowed from under southern Tibet to under the Himalaya. Under Tibet, the channel includes most of the middle crust, some 30–40 km of a 70 km thick crust. At its southern limit, the model channel narrows as it approaches the mountain front at the southern edge of the Tibetan Plateau. There, focused surface erosion leads to rapid removal of rock and to exposure of the channel as the GHS. The temperature of the channel is buffered by the presence of melt and solid phases. These phases are leucogranite melt, the remaining reactants of the melt-producing reaction, and other products of mica dehydration of metasedimentary rocks, which are in equilibrium at 11–12 kbar at 750– 8008C (muscovite þ albite þ quartz 1⁄4 aluminium silicate þ potassium feldspar þ leucogranite melt; Fig. 2). According to the basic model, the sedimentary protolith was entrained in the mid-crust as the Indian plate underthrust Tibet (Jamieson et al. 2006). The Indian lower crust and lithospheric mantle were subducted while the Late Proterozoic From: LAW, R. D., SEARLE, M. P. & GODIN, L. (eds) Channel Flow, Ductile Extrusion and Exhumation in Continental Collision Zones. Geological Society, London, Special Publications, 268, 415–423. 0305-8719/06/$15.00 # The Geological Society of London 2006. to Mesozoic sedimentary cover accreted in the middle of the thickening crust. The depth to the resulting channel corresponds to 8–12 kbar. This is the range of maximum model pressures reported from metamorphosed pelitic rocks of the GHS along the entire Himalayan chain (e.g. Brunel & Kienast 1986; Hubbard 1989; Inger & Harris 1992; Macfarlane 1995; Vannay & Hodges 1996; Vannay & Grasemann 1998; Neogi et al. 1998; Ganguly et al. 2000; Catlos et al. 2001; Daniel et al. 2003; Dasgupta et al. 2004; Harris et al. 2004). We note here that it is exceedingly difficult to extract accurate pressures (and temperatures) from the Himalayan metamorphic rocks. Davidson et al. (1997) and Daniel et al. (2003) showed that thermobarometry of rocks that are not fully characterized by using X-ray composition maps likely produces unreliable results. This is because of the near-universal chemical disequilibrium due to variable response of mineral compositions to reheating, to decompression at high temperature, and to rapid cooling. Furthermore, the uncertainty due to choice of mineral compositions outweighs analytical or thermodynamic model uncertainties. Accordingly, where, in this paper, we report the pressure in the proposed channel to be 11–12 kbar, we recognize an uncertainty that ranges from 8 to 14 kbar. And we estimate an uncertainty range from 4 to 6 kbar in our reporting of a pressure of 5 kbar. Nevertheless, maximum pressures at peak temperatures reported from the GHS along the Himalaya range only between 8 and 12 kbar. The ‘uniform’ pressure of 8–12 kbar is predicted by flow of a subhorizontal channel south from the mid-crust under Tibet. That is, the rocks of the channel do not originate from extreme depths at the base of the Tibetan or Himalayan crust (about 25 kbar). The origin of the GHS rocks from the ‘same’ depth means that the amount of exhumation depends on how far the channel has progressed Fig. 1. Geological map of Bhutan and surrounding areas, after Grujic et al. (2002) and references therein, with modifications based on continuing mapping by D. Grujic and students, and by Bhutanese colleagues. K, Klippen of Chekha Formation; KT, Kakhtang thrust; MBT, Main Boundary Thrust; MCT, Main Central Thrust; MFT, Main Frontal Thrust; STD, South Tibetan detachment. A–A’ indicates line of schematic section in Figure 5. L. S. HOLLISTER & D. GRUJIC 416