Crustal reworking during orogeny: an active-system Himalayan perspective
P. Zeitler, A. Meltzer (Lehigh University), with co-PIs C.P. Chamberlain (Standford), J. Blum (Michigan), P.O. Koons (Maine), D. Craw (Otago), J. Armbruster, L. Seeber (LDEO) S. Park, R. Mackie (Riverside), W.S.F. Kidd (Albany), P. Le Fort and A. Pecher (Joseph-Fourier), J. Shroder, M. Bishop (Nebraska), J. Quade (Arizona) M.Q. Jan, M. Khattak, M.A. Khan, S. Hamidullah (Peshawar).

Overview of the Nanga Parbat Project.
Our goal in this multidisciplinary project was to use the excellent exposure of lower-crustal rocks at Nanga Parbat to understand how continental crust is petrologically and structurally overprinted during collision. We used studies in seismology, magnetollurics, structure, geochronology, petrology, geochemistry, and geomorphology to assess the tectonic processes responsible for the young igneous and metamorphic activity and rapid exhumation observed at the massif.

Detailed structural mapping and patterns of cooling ages show the massif to be a crustal-scale pop-up structure marked by active brittle faults and older shear zones into which relatively older granitoids were emplaced (Figures 15, 19; Schneider et al., 1999a, 1999b; Edwards et al., 2000). No evidence for significant extensional exhumation is present. Developed in Proterozoic basement, the massif was first overprinted by an earlier Himalayan metamorphism as shown by the presence of an ~18 Ma anatectic granite (Schneider et al., 199b). Widespread young granite dikes 1-3 Ma in age are confined to the topographically high core of the massif (Schneider et al., 1999c), where low-P/high-T cordierite/K-feldspar gneisses (Poage et al., in press) with metamorphic ages of ~3 Ma occur. Pervasive upper-crustal fluid flow occurs in the core region, as do steep thermal gradients of 60¾C/km or more as shown by fluid inclusions for the top 3 km of the crust. A sharp lower cut-off in microseismicity, bowed upward 3 km beneath the summit region, indicates the brittle-ductile transition is shallow at ~2-5 km bsl (Figure 20a; Meltzer et al., in review). This pattern, together with tomographic results showing very low Vp and Vs and higher attenuation throughout the crust in the region below the core of the massif, is consistent with rapid advection of hot crust at 5 mm/yr as indicated by petrological and geochronological data (Figures 15, 19, 20d-g; Meltzer et al., 2001; Sarker and Meltzer, in review). This also suggests that the primary flow path of material into the massif is from depth rather than along a shallow detachment. MT and seismic data rule out large magma bodies as the cause of the observed petrological anomaly; variable seismic waveforms indicate small-scale heterogeneities in the shallow crust related to fluids (either partial melt or aqueous; Figure 20c), whereas the MT data surprisingly show the lower crust to be atypically resistive (Figure 20f; Park and Mackie, 1997, 2000; Meltzer et al., in review). These structural, geophysical, and petrological anomalies occur in a bulls-eye pattern around the summit massif and are associated with focused exhumation and concentrated strain (Zeitler et al., 2001).

Using a coupled thermal/mechanical/erosional model (Koons et al., 2002), we can show that in a deforming orogen, local rheological variations such as those that arise from deep and rapid incision (as seen at Nanga Parbat (Shroder and Bishop, 2000)), can strongly rearrange the regional strain pattern to one which departs significantly from that predicted using uniform rheological parameters and simple velocity boundary conditions (Figure 21). A weak area will focus particle paths such that movement of material within the orogen will be concentrated into the weaker zone. A positive feedback develops in which advection of material into this weakened zone results in concentrated exhumation and corresponding advection of isotherms, further weakening the upper crust. One outcome of such erosional/mechanical coupling is the development of very large mountains of relatively limited spatial extent perched atop hot, thin, weak crust.

The manifestations of such localized deformation are many (Figure 21c). Focusing of strain and rapid exhumation lead to substantial metamorphic and structural overprinting of the crust, as high-temperature lower-crustal rocks, possibly already partially molten, are isothermally decompressed and brought into communication with surface waters. The thermal and petrological anomalies associated with this exhumation and concentration of strain can be thought of as a tectonic aneurysm. This model provides an integrating framework for understanding the diverse suite of observations that we and other workers have collected at Nanga Parbat. Our view is that rapid erosion and excavation of a deep gorge by the Indus River has focused strain and triggered development of a tectonic aneurysm in crust initially weakened by thickening and having very high radioactive heat production. The emplacement of granites during recent erosional exhumation (~1-5 Ma), development and exposure of young low-pressure granulite-facies metamorphic rocks (3 Ma in age), development of structural relief via antiform growth and thrusting, formation of a vigorous metamorphic/meteoric hydrothermal system, shallow brittle to ductile transition, and generally hot resistive crust are all consistent with advection of deep crustal material into a relatively weak crustal zone.

We suggest that at Nanga Parbat (see Figure 21c), initial high-pressure metamorphic equilibration and volatile release occur within a high-strain zone located to the south of the massif. In this zone, mid- to deep-crustal metamorphic recrystallization occurs as non-equilibrium transitions driven by steep velocity gradients affecting the rheology of the crust. Rocks having passed through the high-strain zone have been dehydrated and contain insufficient connected fluid phase to serve as an electrical conductor so that the crust beneath the massif is resistive. Rapid decompression of water-poor, dehydrated gneisses generates vapor-absent granitic melts which are emplaced upwards into the massif (Zeitler and Chamberlain, 1991; Butler et al., 1997). These dry gneisses then pass into a lower-strain zone beneath the summit, advecting isotherms. During uplift the gneisses undergo decompression reactions, generating the lower-pressure and high-temperature granulite assemblages preserved in the center of the massif.

Our model also sheds light on the relationship between fluid flow and the evolution of the brittle/ductile transition at Nanga Parbat. Less than 1 m.y. after exhumation begins, a near-steady-state thermal pattern is established (Koons, 1987). Steep geothermal gradients are produced (approaching 100°C/km near the surface), elevating the 400°C isotherm and the position of the brittle/ductile transition to ~2-5 km depth bsl, as suggested by our petrologic, fluid inclusion, and seismologic observations at Nanga Parbat (Winslow et al., 1994; Craw et al. 1994; Figure 20g). Rock moving through this thermal boundary passes from predominantly ductile to predominantly brittle mechanical behavior. The thermal boundary layer is a dynamic zone of fluctuating ductile/brittle strain as fluid pressure cycles between lithostatic and hydrostatic. Brittle failure within this boundary layer can occur as shear failure at elevated fluid pressure, or as local hydrofracturing, and provides the source of some of our recorded microseismicity. At the elevated temperatures of the lower part of the thermal boundary (~450°C; Craw et al., 1994, 1997), the few fluids available undergo phase transformations from supercritical fluid in the ductile rock at lithostatic pressure to dry steam as the pressure is released to hydrostatic pressure. The total amount of fluid at the thermal boundary base is minor, with little cooling potential, and has both the thermal and isotopic signature of the rock. The upper part of the thermal/mechanical boundary layer is one of high geothermal gradients with vigorous free and forced convection systems connected to the surface. Fluids are dominantly meteoric (Chamberlain et al., 1995). Fluids recharge along high altitude shear zones in the massif and discharge along marginal shear zones within boiling hot springs (Craw et al., 1994; Chamberlain et al. 1995). Driving forces for fluid flow are a combination of topographic and thermal gradients in the upper 3-5 km.

Specific Outcomes.
Our goal was to document and better understand the processes that have recently placed a strong structural and metamorphic overprint onto the Precambrian rocks of the massif. A significant portion of this project involved the comprehensive characterization of the massif’s structure, petrological conditions, and geochronology, to document the extent of the anomalous conditions in the core of the massif and to determine the relationships of various data sets, as described above. This effort was essential. Prior to this study virtually all data from Nanga Parbat came from a single ~10 km traverse on the north side of the massif, augmented by a few studies along the Indus gorge. Our MT data and our dense seismic array have provided the first close-up look at an actively deforming metamorphic massif, linking the current state of the lower crust and deformation regimes in the crust with the petrologic and other observations made on surface exposures. Given the resemblance that Nanga Parbat shows to classic gneiss domes, our aneurysm model provides both a genetic mechanism and a tectonic setting for these features often seen in older mountain belts.