Abstract

A central debate for the evolution of the Himalayan orogen is how the Greater Himalayan Crystalline complex in its core was emplaced during the Cenozoic Indo-Asian collision. Addressing this problem requires knowledge of the structural relationship between the South Tibet detachment fault (STD) and the Main Central thrust (MCT) that bound these rocks from above and below. The fault relationship is exposed in the Himachal Himalaya of northwestern India, where they merge in their updip direction and form a frontal branch line that has been warped by subsequent top-to-the-southwest shear deformation. To elucidate how the two major crustal-scale faults evolved in the western Himalaya, we conducted integrated geologic research employing field mapping, pressure-temperature (P-T) analyses, U-Pb zircon geochronology, trace and rare earth element (REE) geochemistry, and thermochronology. Our field study reveals complex geometric relationships among major thrusts with large-magnitude shortening within each thrust sheet. Three successive stages of top-to-the-southwest thrust development are recognized: (1) imbricate stack development, (2) translation of large thrust sheets along low-angle detachments and backthrusting along the STD, and (3) development of duplex systems via underplating. This kinematic process can be quantified by our new analytical data: (1) P-T determinations show 7–9 kbar and 450–630 °C conditions across the STD. The lack of a metamorphic discontinuity across the fault is consistent with a backthrust interpretation. (2) U-Pb zircon geochronology yields ca. 830 Ma and ca. 500 Ma ages of granitoids in the MCT hanging wall, ca. 1.85 Ga ages of granitic gneisses in both the MCT hanging wall and footwall, and 8–6 Ma ages of granitic pegmatites in the MCT footwall. These ages help define regional chronostratigraphy, and the youngest ages reveal a previously unknown intrusion phase. (3) Trace element and REE geochemistry of 1.85 Ga, 830 Ma, and 500 Ma granitoids are characteristic of remelted continental crust, constraining the protolith tectonic setting. (4) U-Pb geochronology of detrital zircon reveals that siliciclastic sedimentary sequences above the STD, below the MCT, and between these two faults have similar age spectra with Neoproterozoic youngest age peaks. This result implies that the STD and MCT each duplicated the same stratigraphic section. (5) Th-Pb geochronology of monazite included in MCT hanging-wall garnet yields Paleozoic and early Tertiary ages, indicating Paleozoic and early Tertiary metamorphism in these rocks. (6) The 40Ar/39Ar thermochronology of the K-feldspar from southern MCT hanging-wall rocks evinces cooling below 220–230 °C ca. 13–19 Ma or later, constraining the thrust development history. We use these results to derive a tectonic model of crustal shortening across the Himachal Himalaya involving early thickening, tectonic wedging emplacement of the Greater Himalayan Crystalline complex between the MCT and STD, and continued growth of the Himalayan thrust wedge by accretion of thrust horses from the Indian footwall.

INTRODUCTION

The first-order architecture of the Himalayan orogen is expressed by two major north-dipping faults bounding a high-grade complex in the orogenic core (e.g., Argand, 1924; LeFort, 1996; Yin and Harrison, 2000; DeCelles et al., 2002; Yin, 2006). A central issue with regard to the Cenozoic Himalayan development is how the metamorphic core, the Greater Himalayan Crystalline complex, has been emplaced to its current position (LeFort, 1975; Burchfiel and Royden, 1985; Grujic et al., 1996; Nelson et al., 1996; Webb et al., 2007). Resolving this issue requires knowledge of the kinematic history of each bounding fault, i.e., the Main Central thrust (MCT) below and South Tibet detachment (STD) above, and the structural relationship between these faults.

The regional significance of the MCT as a major Cenozoic shortening structure has been recognized since the classic work of Heim and Gansser (1939) (Fig. 1) (see also Le Fort, 1975; Upreti, 1999; Hodges, 2000; DiPietro and Pogue, 2004). In contrast, the STD was discovered much later (Burg et al., 1984; Burchfiel et al., 1992). The STD is generally north-dipping, features alternating top-to-the-south and top-to-the-north shearing, and juxtaposes the largely low-grade Tethyan Himalayan Sequence on top of the Greater Himalayan Crystalline complex (e.g., Patel et al., 1993; Hodges et al., 1996). Excepting the top-to-the-south shear indicators, these records are consistent with a normal fault interpretation. The apparent presence of a major normal fault within the contractional orogenic setting of the Himalaya has led to intense debate over the tectonic origin and dynamic role of the STD (e.g., Burg et al., 1984; Burchfiel and Royden, 1985; Yin, 1989; Hodges et al., 1992, 1996; Brown and Nazarchuk, 1993; Patel et al., 1993; Yin et al., 1994, 1999; Lee et al., 2000; Grujic et al., 2002).

Current hypotheses for the emplacement of the Greater Himalayan Crystalline complex offer different solutions to this problem (Fig. 2). Vertical wedge extrusion models show the STD as a normal fault at the crust of a Coulomb orogenic wedge (e.g., Burchfiel and Royden, 1985; Grujic et al., 1996). Models of southward middle-crustal channel flow interpret the STD as a backstop normal fault allowing the extrusion of channel rocks linked to focused denudation along the Himalayan topographic front (e.g., Nelson et al., 1996; Beaumont et al., 2001). In tectonic wedging models, the STD acts largely as a subhorizontal backthrust off of the MCT, with its top-to-the-north shear surfacing as the Great Counter thrust system (Webb et al., 2007). These competing models for the emplacement of the Greater Himalayan Crystalline complex make different predictions (Table 1) (Fig. 2). First, the wedge extrusion model (Burchfiel and Royden, 1985) requires local extension over the highest region of the Himalaya and suggests that slip may be focused along a preexisting lithologic contact (Burg et al., 1984; Burchfiel and Royden, 1985). Second, both wedge extrusion and channel flow models require rapid erosion of the Tethyan Himalayan Sequence and exposure of the Greater Himalayan Crystalline complex during the main motion along the MCT and STD in the Early and Middle Miocene (Nelson et al., 1996; Beaumont et al., 2001; Hodges et al., 2001). In contrast, the tectonic wedging model predicts that the Tethyan Himalayan Sequence was preserved above the Greater Himalayan Crystalline complex during STD motion. Third, the wedge extrusion model predicts the STD and MCT merge downdip to the north, the channel flow model predicts them to be largely subparallel, and the tectonic wedging model predicts them to merge updip to the south.

As the relationship between the STD and MCT is central in differentiating these models, field tests must be conducted in regions where their relationships can be established. This requirement motivates our geologic investigation of the Himachal Himalaya in northwest India (Fig. 1), where contact relationships between the MCT and STD have been proposed and locally tested (Thakur, 1998; Yin, 2006; Webb et al., 2007). This region loosely marks the transition from the western to the central Himalaya, which are in part distinguished by drastically different preservation of the Tethyan Himalayan Sequence above the STD. In the central Himalaya, the Tethyan Himalayan Sequence has been largely eroded away, leaving the MCT and STD exposed as subparallel structures above and below the Greater Himalayan Crystalline complex (e.g., Hodges et al., 1996; DeCelles et al., 2001). In contrast, the Tethyan Himalayan Sequence is well preserved in the western Himalaya. In the Himachal Himalaya, an along-strike variation of MCT juxtaposition (i.e., from a Greater Himalayan Crystalline complex over Lesser Himalayan Sequence relationship in the east to a Tethyan Himalayan Sequence over Lesser Himalayan Sequence relationship in the west) can be directly observed (Fig. 3) (Thakur, 1998; Steck, 2003; DiPietro and Pogue, 2004). Besides this advantage, the stratigraphic units above and below the MCT are correlative in Himachal (e.g., Miller et al., 2001; Myrow et al., 2003), allowing assessment of the original configuration of the northern Indian margin before the Cenozoic Indo-Asian collision. This correlative stratigraphic relationship across the MCT in the Himachal Himalaya contrasts strongly with the geology of the central Himalaya of Nepal, where rock units across the MCT differ drastically in age and provenance (Parrish and Hodges, 1996; DeCelles et al., 2000).

In this study we compiled a regional geologic map that combines information from existing literature and our new geological and analytical data collected via structural, geochemical, thermobarometric, and U-Th-Pb geochronologic and 40Ar/39Ar thermochronologic analyses of key areas and critical samples. We integrate our new results into a tectonic model that shows that the construction of the Himachal Himalaya was mainly accomplished by footwall accretion and vertical thrust stacking of Proterozoic strata of the northern Indian passive margin sequence, consistent with a tectonic wedging emplacement of the Greater Himalayan Crystalline complex.

LITHOLOGIC UNITS

Major lithologic units in the study area include the Cretaceous and Cenozoic Sub-Himalayan Sequence, the Proterozoic and Cambrian Lesser Himalayan Sequence, the high-grade Greater Himalayan Crystalline complex, and the Neoproterozoic to Mesozoic Tethyan Himalayan Sequence (Figs. 3 and 4; Table 2). The Lesser Himalayan Sequence, Greater Himalayan Crystalline complex, and Tethyan Himalayan Sequence are structurally divided as MCT footwall rocks, rocks encased by the MCT below and the STD above, and rocks structurally above the STD, respectively (e.g., Hodges, 2000; Yin, 2006). We describe the lithologic units briefly herein; for an expanded description, see Appendix 1.

The Sub-Himalayan Sequence consists of lower shallow-marine strata and upper continental deposits separated by an Oligocene unconformity (Table 2). Sub-Himalayan Sequence rocks depositionally overlie rocks of the Lesser Himalayan Sequence and correlative rocks at the base of the Himalayan foreland basin (e.g., Powers et al., 1998).

Four subunits are distinguished within the Lesser Himalayan Sequence: (1) the Neoproterozoic–Cambrian Outer Lesser Himalayan Sequence in the hanging walls of the Krol and Tons thrusts, (2) the Paleoproterozoic–Neoproterozoic Damtha and Deoban Groups in the hanging wall of the Bilaspur thrust and the footwalls of the Tons and Berinag thrusts, (3) the Paleoproterozoic Berinag Group in the hanging wall of the Berinag thrust, and (4) the Paleoproterozoic Munsiari Group, dominantly in the hanging wall of the Munsiari thrust (Table 2) (Figs. 3 and 4).

The Greater Himalayan Crystalline complex is ∼7–9 km thick and consists of paragneiss, schist, and orthogneiss intruded by minor Tertiary leucogranites concentrated mostly in its upper 2–3 km (Table 2). An inverted metamorphic field gradient is observed across a complete section of these rocks along the Sutlej River, progressing from garnet-staurolite–bearing rocks at the base to migmatitic rocks near the top (Vannay and Grasemann, 1998).

The Tethyan Himalayan Sequence is dominated by the Neoproterozoic–early Cambrian Haimanta Group, early Paleozoic granites, Cambrian Parahio Formation, and overlying Paleozoic–Mesozoic strata (Table 2). Its basement is likely the Paleoproterozoic Baragaon gneiss in the MCT shear zone directly below the unit (Bhanot et al., 1978; Miller et al., 2000; this study). The Haimanta Group is garnet grade across its basal 1–3 km of section, and the grade decreases upsection across the Tethyan Himalayan Sequence.

REGIONAL TECTONIC FRAMEWORK

First-order structures in the Himachal Himalaya are represented by a stack of northern rooted thrusts, many of which are folded (Fig. 3). The main fault zones and fault systems include, from southwest to northeast, (1) the Main Frontal thrust, (2) the Sub-Himalayan thrust zone, (3) the Bilaspur-Palampur thrust system, (4) the Krol-Mandi thrust system, (5) the MCT, (6) the Tons thrust, (7) the Berinag thrust, (8) the Munsiari thrust, (9) the STD, (10) the Tethyan Himalayan fold-and-thrust belt including the Mata nappe, and (11) the Great Counter thrust system (Figs. 3 and 4). Excepting the STD and the Great Counter thrust system, all major structures are southwest-directed thrusts. Several large structural culminations are also exposed in the map area due to folding of major thrusts and the development of low-angle detachment faults. We refer to these structures as (1) the Narkanda half-window, (2) the Uttarkashi half-window, (3) the Kullu window, (4) the Leo Pargil dome, and (5) the Tso Morari gneiss dome (Fig. 3). We briefly describe this regional tectonic framework herein; for an expanded description, see Appendix 2.

MCT Footwall Structures

The Sub-Himalayan thrust zone is bounded by the Main Frontal thrust below and the Krol and Mandi thrusts above (Fig. 3). The Main Frontal thrust places Neogene–Quaternary strata over the modern Indo-Gangetic Plain deposits; the Mandi and Krol thrusts place Lesser Himalayan Sequence rocks over Paleogene–Quaternary strata of the Sub-Himalayan Sequence (Kumar et al., 2006; Srikantia and Sharma, 1976; Powers et al., 1998; Raiverman, 2000).

The Tons thrust, exposed along the Sutlej River near Shimla and across the southern margin of the Uttarkashi half-window, places the Outer Lesser Himalayan Sequence over the Deoban and Damtha Groups (Fig. 3) (e.g., Valdiya, 1980; Célérier et al., 2009a). The Munsiari thrust can be traced along most of the central Himalayan orogen (e.g., Upreti, 1999; Yin, 2006; Searle et al., 2008; see discussions in Célérier et al., 2009a, 2009b), and is referred to as the MCT I in Nepal (Bordet et al., 1972; Arita, 1981; Harrison et al., 1998). The thrust crops out in the Kullu window and the Uttarkashi half-window, where it places the Munsiari Group (Wangtu and Jeori gneiss) over the Berinag Group (Figs. 3, 4, 6B, and 6C) (e.g., V.P. Sharma, 1977; Valdiya, 1980; Vannay et al., 2004). The Berinag thrust appears in both the hanging wall and footwall of the Munsiari thrust, where it juxtaposes Berinag Group rocks over the Wangtu gneiss and the Damtha Group, respectively (Fig. 3) (e.g., V.P. Sharma, 1977; Valdiya, 1980; Vannay et al., 2004; Célérier et al., 2009a).

MCT

The MCT is classically defined as the tectonic boundary between the Greater Himalayan Crystalline complex above and the Lesser Himalayan Sequence below (e.g., Heim and Gansser, 1939; Le Fort, 1996; Hodges, 2000; Yin, 2006). This definition is uncertain because of debates over which local lithological units should be attributed to which tectonic unit, and it is circular, since the Greater Himalayan Crystalline complex and Lesser Himalayan Sequence are defined by their bounding structures (see discussion by Upreti, 1999). Nonetheless, for the Himachal Himalaya most workers share a consensus interpretation of the MCT as a continuous, folded, southwest-directed thrust shear zone as much as 2 km thick that was active in the Early and Middle Miocene and has a largely established map trace (as shown in Fig. 3) (e.g., Thakur and Rawat, 1992; Frank et al., 1995; Steck, 2003; Vannay et al., 2004). This interpretation is based on congruent lithology, strain concentration, metamorphic grade, and thermochronologic ages along the mapped shear zone (e.g., Frank et al., 1995; Grasemann et al., 1999; Vannay et al., 2004; Thiede et al., 2005). The MCT hanging-wall rocks show variation; at its northeasternmost trace along the Sutlej River, the MCT underlies a well-established inverted metamorphic sequence that is universally acknowledged as Greater Himalayan Crystalline complex rocks (Fig. 3) (e.g., Vannay and Grasemann, 1998; Hodges, 2000; Steck, 2003; DiPietro and Pogue, 2004; Yin, 2006). Conversely, it has long been recognized that the MCT hanging-wall rocks to the west of Mandi (i.e., west of ∼31°50′N, 77°E) display a right-way-up metamorphic field gradient to chlorite zone conditions and are structurally continuous with the Tethyan Himalayan Sequence to the northeast (Fig. 3) (Frank et al., 1995; Fuchs and Linner, 1995; Thakur, 1998; Steck, 2003; DiPietro and Pogue, 2004; Yin, 2006). Therefore the Himachal region requires a relaxation of the MCT definition as the boundary between the Greater Himalayan Crystalline complex and the Lesser Himalayan Sequence.

Based on the change in hanging-wall rocks, we divide the MCT into northern and southern segments. The northern MCT juxtaposes the Greater Himalayan Crystalline complex over the Lesser Himalayan Sequence; the southern MCT places the Tethyan Himalayan Sequence and the Baragaon gneiss over the Lesser Himalayan Sequence (Fig. 3). The intersection line of the STD and the MCT marks the boundary between the two segments of the MCT to the north and south (Thakur, 1998; Yin, 2006; Webb et al., 2007). In subsequent text, we refer to the “MCT zone” if the ∼2 km thickness of the shear zone is relevant.

In the map area the MCT is folded and displays large full and half-windows and isolated klippes (Fig. 3). Cutoff relationships in the MCT footwall suggest that the thrust cuts upsection to the southwest in its transport direction. However, the presence of a large footwall ramp along the MCT in the map area raises the question of whether the Lesser Himalayan Sequence strata in the footwall ramp were horizontal when the MCT was initially cutting across them. As Cretaceous–Eocene beds overlie the Shimla Group and Tal Formation of the younger Outer Lesser Himalayan Sequence units in the south and the Deoban-Damtha strata in the north (Srikantia and Sharma, 1976; Valdiya, 1980) (Fig. 3), the Proterozoic–Cambrian sequence of the northern Indian passive margin must have been tilted to the south prior to the emplacement of the MCT hanging wall in the region.

Minimum displacement along the MCT in the map area is ∼115 km, determined from the northernmost and southernmost exposures of the fault. Early to Late Miocene activity on the northern segment of the MCT has been inferred from U-Th monazite-inclusion dating, 40Ar/39Ar muscovite cooling ages, and zircon fission track ages from the Greater Himalayan Crystalline complex hanging-wall rocks (Fig. 4A) (Walker et al., 1999; Schlup, 2003; Vannay et al., 2004; Thiede et al., 2005). The portion of the MCT across the Kullu and Uttarkashi windows must have ceased motion in the Late Miocene when these windows were developed and caused folding of the MCT, as indicated by cooling ages (Vannay et al., 2004; Thiede et al., 2005; Caddick et al., 2007; Chambers et al., 2008). However, the relationship does not preclude the southernmost MCT linking younger thrusts in the Lesser Himalayan Sequence to continue its motion after the Middle Miocene.

STD

The STD juxtaposing the Tethyan Himalayan Sequence over the Greater Himalayan Crystalline complex can be traced continuously from the central Himalaya to the northern end of the Kullu window (Figs. 1 and 3) (Burg et al., 1984; Burchfiel et al., 1992; Choudhuri et al., 1992; Thakur, 1998; Dèzes et al., 1999è; Jain et al., 1999; Wyss et al., 1999; Steck, 2003; DiPietro and Pogue, 2004; Yin, 2006). Here the STD is folded and overturned within the southwest-verging Phojal anticline, and the overturned fault trace extends back to the southeast and intersects the MCT on the north and south sides of the eastern Kullu window and at the northwestern rim of the Uttarkashi window (Fig. 3) (Thakur, 1998; Webb et al., 2007). The MCT-STD branch line trends to the northwest, parallel to the strike of the orogen. It is largely buried to the northwest and eroded to the southeast. In the Himachal Himalaya the fault features both top-to-the-northeast and top-to-the-southwest structures, including S-C fabric, normal drag shear bands, σ-type porphyroclasts, and asymmetric isoclinal to tight folds, across a 300–600-m-thick shear zone (Choudhuri et al., 1992; Jain et al., 1999; Vannay et al., 2004; our observations).

The Zanskar shear zone represents a possible northwestern extension of the STD (Figs. 1 and 3) (e.g., Searle, 1986; Herren, 1987; Patel et al., 1993; Dèzes et al., 1999; Epard and Steck, 2004). The northwest-trending, northeastern segment of this shear zone has been interpreted as (1) an along-strike continuation of the STD, connected in map view to the right-way-up STD across the Himachal Himalaya (e.g., Searle et al., 1988; Jain et al., 1999), or (2) as part of an STD window (Thakur, 1998; Dèzes, 1999; Yin, 2006). We follow the second interpretation because of the established continuity of the Tethyan Himalayan Sequence from Chamba to the Tso Morari, i.e., weakly deformed rocks cross the path of the proposed STD–Zanskar shear zone map-view connection (e.g., Frank et al., 1995; Fuchs and Linner, 1995). Because the Zanskar STD window is warped but not overturned, it follows that the Phojal anticline is a local structure along the northern margin of the Kullu and Uttarkashi windows that does not extend to the Zanskar region.

The amount of displacement along the STD is difficult to determine because rocks above and below the fault share the same protoliths in this region. Also, there is no pronounced metamorphic offset across the fault: metamorphic isograds are subparallel to the STD (Fig. 5). Therefore, despite the significant evidence for shear deformation across the STD zone, local lithologic and petrologic information do not constrain the magnitude of fault offset. The fault may have initiated in the Eocene as a basal décollement of the Tethyan Himalayan fold-thrust belt (Wiesmayr and Grasemann, 2002). Early to Middle Miocene 40Ar/39Ar muscovite ages in the immediate STD hanging wall and footwall suggest that plastic deformation along the STD ceased in that period (Walker et al., 1999; Vannay et al., 2004; Thiede et al., 2005).

Most Tethyan Himalayan Sequence strata northeast of the right-way-up STD are involved in the dominantly top-to-the-southwest Tethyan Himalayan fold-thrust belt within the Spiti Synclinorium (Figs. 3 and 4B) (e.g., Frank et al., 1995; Fuchs and Linner, 1995; Vannay and Steck, 1995; Steck et al., 1998; Wiesmayr and Grasemann, 2002). Similar relationships also occur across the Chamba Synclinorium in the northwest of our map area (e.g., Frank et al., 1995; Fuchs and Linner, 1995). To the north, Tethyan Himalayan Sequence rocks and suture zone rocks are thrust northeastward over the Indus molasse along strands of the latest Oligocene–Miocene Great Counter thrust system (e.g., Schlup et al., 2003; de Sigoyer et al., 2004). The Great Counter thrust system may represent the northern extension of the STD (e.g., Yin et al., 1994, 1999).

STRUCTURAL GEOLOGY

Thrusting is the dominant expression of contractional deformation in the footwall of the MCT. Although many of these thrusts are well defined locally, the relationships among the structures and deformation in individual thrust systems have not been studied in detail. Further issues include: (1) MCT-STD geometric relationships, (2) internal deformation of units to address the assumption of constant bed thickness in line-length balancing, and (3) variable deformation style and strain as functions of both lithology and structural positions relative to nearby faults. To address these issues, we conducted detailed geologic mapping and outcrop-scale observations across the Himachal Himalaya. We summarize the main results of our structural observations in the following.

Mandi-Bilaspur Thrust System Hanging-Wall Structures

Bedding in the Deoban and Damtha Groups is deformed by brittle structures, including dominantly southwest-verging parallel folds and dominantly southwest-directed brittle faults (Figs. 6B, 6D [stereoplots L3, L4, L6, L7], 7A, and 7B). Fold amplitudes range from <1 m to ∼100 m, with interlimb angles from 40° to 140°. Deoban rocks commonly contain as much as 3% carbonate veins generated from pressure solution during the development of axial cleavage, best displayed in the fold hinge zones. Carbonate rocks of the Deoban Group are deformed in a southwest-directed thrust duplex with the roof fault at the base of the Baragaon gneiss in the MCT shear zone (Fig. 7B). We also observed folded quartz veins in southwest-verging similar folds in Deoban carbonaceous phyllites, located within the 50-m-thick basal section of the Baragaon gneiss (Fig. 7C). The extensive thrusting and folding in the Narkanda and Uttarkashi half-windows suggest that the contractional deformation here is penetrative at the kilometer scale.

Tons Thrust System and Hanging-Wall Structures

The Tons thrust hanging wall displays large and distributed internal deformation. The bedding and phyllitic fabric in the Shimla Group in a zone of ∼500 m above the Tons thrust is pervasively folded in the Narkanda and Uttarkashi half-windows (Figs. 6B and 6D; stereoplots L7, L8). They are typically expressed by the formation of southwest-verging tight folds with wavelengths ranging from a few millimeters to a few meters (Fig. 7D). The fold development is spatially associated with slatey cleavage (Fig. 7E). Horizontal shortening of the highly folded strata locally exceeds 70% (Fig. 7E). Above this zone, deformation intensity increases upward toward the MCT zone, with bedding planes progressively transposed by slaty cleavage first and then schitosity near the MCT. Rocks at the highest structural levels of the Tons hanging wall exhibit greenschist facies metamorphism, expressed by the presence of pelitic phyllite, graphitic phyllite, and chlorite-pyrite schists. Tight to isoclinal southwest-verging folds, slaty cleavage, and phyllitic fabric with northeast-trending stretching lineations occur and are increasingly dominant upsection (Fig. 6B and 6D).

Berinag Thrust and Hanging-Wall Structures

Similar to the Bilaspur and Tons thrust hanging walls, internal deformation of the Berinag thrust sheet is penetrative. The Berinag thrust is expressed by a southwest-directed brittle structure in the southwestern Kullu window and Uttarkashi half-window to a top-to-the-southwest shear zone in the northeastern portions of the windows (Fig. 6; stereoplots L1, L2, L3, L6). Hanging-wall quartzites of the Berinag Group are partially recrystallized at the southwestern limits of the exposed thrust sheet, where preserved sedimentary structures include cross-bedding (Fig. 7F). Here the Berinag thrust hanging wall shows large-magnitude internal shortening, with tight and/or isoclinal meter-scale folds of quartzite cross-bedding indicating ∼64% horizontal shortening (Fig. 7G). The bulk of the Berinag thrust sheet within the Kullu window is schistose with north-northeast–trending stretching lineations defined by white mica (Fig. 6). Folds of schistosity are dominantly southwest verging and tight to isoclinal in cross-section view. At the eastern end of the Kullu window an ∼30-m-thick band of Berinag quartzite occurs between concordant contacts with Wangtu orthogneiss below and Greater Himalayan Crystalline complex paragneiss above. The quartzite here is completely recrystallized, with grain sizes as large as 5 mm; all three units display moderately northeast-plunging stretching lineations and top-to-the-southwest mylonitic fabric. Therefore deformation intensity within the Berinag thrust sheet increases to the northeast.

Munsiari Thrust and Hanging-Wall Structures

Penetrative deformation occurs throughout the Munsiari thrust hanging wall, with the exception of late crosscutting undeformed granites (Figs. 6B–6D; stereoplots L2, L5). The Munsiari thrust juxtaposes orthogneiss atop Berinag Group quartzites along a 1–2-km-thick top-to-the-south to top-to-the-south-southwest ductile shear zone containing S-C mylonitic fabrics. South-southwest–directed brittle structures overprint the ductile structures. These structures include Riedel shears, cataclasite, slickenfibers, and (in the Uttarkashi half-window) a >50-m-thick schuppen zone comprising 2–15-m-thick horses of quartzite, granitic schist, and granitic gneiss. Gneissic fabric across the Munsiari Group is characterized by a dominantly north-northeast–trending stretching lineation (Figs. 6B–6D). Gneissic foliation in the Jeori gneiss unit exposed in the Sutlej River Valley (∼15 km northeast of Rampur) is penetratively deformed by kink folds with centimeter- to meter-scale wavelengths and ∼90° interlimb angles (Fig. 7H). Along a 10 km stretch of the Sutlej River in the eastern Kullu window (centered on the town of Wangtu), unfoliated pegmatitic granites crosscut the foliation of the Wangtu granitic gneiss (Fig. 7I). The Late Miocene ages of these granites provide a lower limit for the development of the gneissic fabric (see following Geochronology discussion). These granites are the only unit within the Munsiari thrust hanging wall, and indeed within the Lesser Himalayan Sequence, for which we did not observe outcrop-scale deformation.

MCT and Hanging-Wall Structures

We divide the MCT into northern and southern portions based on Greater Himalayan Crystalline complex and Tethyan Himalayan Sequence hanging-wall lithologies, respectively. This division is structurally defined by the overturned South Tibet detachment, which is folded with the southwest-verging Phojal anticline. In this interpretation, the southern MCT hanging wall is continuous with the STD hanging wall. Therefore determining the deformation style and strain of the southern MCT hanging wall, the northern MCT hanging wall, and the STD hanging wall provides key tests of the interpretation. We review our main structural observations from each area in the following.

Southern MCT and Hanging-Wall Structures

Observations of deformation in the MCT hanging wall south of the Kullu window are limited because (1) most preserved rocks are near the base of the thrust sheet, with structurally high rocks largely eroded away, and (2) despite this knowledge, the undulatory folding of the thrust sheet limits our ability to tightly constrain structural elevation above the MCT base across much of the thrust sheet. This portion of the MCT is defined by a 1–2 km top-to-the-southwest shear zone featuring S-C fabrics with northeast-trending stretching lineations defined by biotite and feldspar, normal drag shear bands, ultramylonites, σ-type porphyroclasts of feldspar, brittle-ductile synthetic and antithetic microfaulting of feldspar, and tight to isoclinal folds [Figs. 6A, 6B, 6D (stereoplots TG1, T5, T6, T7, T8), 8A, and 8B]. The base of the shear zone is largely coincident with the garnet isograd within an inverted metamorphic field gradient (Fig. 5). Where the MCT juxtaposes Haimanta rocks directly above Outer Lesser Himalaya rocks (at Narkanda and along the Shimla klippe), mylonitic garnet schist of the MCT zone concordantly overlies pelitic phyllite, chlorite-pyrite schists, and graphitic phyllite, and graphitic schist. Here shear bands and isoclinal folds persist across ∼400 m structurally below the MCT zone. As previously noted, a brittle fault juxtaposes Baragaon gneiss over Deoban strata at the base of the MCT zone. The geometry of R1 shears and meter-scale duplexing along the brittle fault support a top-to-the-southwest sense of motion along this fault (Fig. 7B). The contact between the Baragaon gneiss and the overlying Haimanta Group is concordant and locally folded (Fig. 8B). In summary, the southern MCT is dominated by top-to-the-southwest ductile fabrics, with southwest-directed brittle deformation occurring along the MCT contact between the Baragaon gneiss and the Deoban Group.

The southern MCT hanging wall is dominated by ductile fabrics, with minor brittle deformation. The dominant structural fabrics are shallow to moderate foliations defined by micas and subparallel to locally preserved bedding, northeast-trending stretching lineations, crenulation cleavage, and largely southwest-verging tight to isoclinal folds of schistosity [Figs. 6 (stereoplots TG1, T5, T6, T7, T8) and 8C]. These fabrics are not entirely penetrative: they are ubiquitous in Haimanta metasedimentary rocks and Baragaon gneiss, but not throughout early Paleozoic granites. The early Paleozoic granite-Haimanta contacts are concordant, and the granite is foliated, lineated, and even locally mylonitized ∼5–30 m from the contact (Fig. 8D). However the internal fabric of the granite ranges from weakly foliated to unfoliated (Fig. 8E). A top-to-the-southwest shear zone developed across an ∼50-m-thick layer of ca. 830 Ma (see Geochronology discussion) mylonitic granitic gneiss exposed ∼4 km north of Narkanda. We named this the Audi shear zone, based on a local village, and interpret it as continuous with shear fabrics developed at the base of ca. 830 Ma rocks near Naura (Figs. 3, 4, and 6B). For purposes of comparison with the northern MCT hanging wall and the STD hanging wall, the key observation is that the penetrative nature of deformation in the southern MCT hanging wall is not complete: early Paleozoic granites are largely undeformed. We also observed top-to-the-southwest, 0.2–1-m-thick, brittle fault zones (Fig. 8F). In Haimanta rocks, the faults are developed along graphite-rich layers and display a steel blue color. Limited exposure prevented us from tracing such faults farther than ∼50 m.

The transect of the thrust sheet along the Beas River from Mandi to Larji presents the best opportunity to observe the structurally high portions of the southern MCT hanging wall, because the sheet is here folded in a single syncline and the preserved core of this syncline preserves rocks that are ∼4.5 km above the MCT base (Figs. 3, 4, and 6B). The northeastern flank has ultramylonitized Baragaon granite at the MCT base. To the southwest, overlying rocks ∼1.5 km thick, are garnet schists with top-to-the-southwest S-C fabrics, northeast-trending stretching lineations, tight to isoclinal southwest-verging folds, and 1%–5% quartz veins that are folded consistently with the schist. At higher structural levels, rocks show decreases in deformation intensity and metamorphic grade. The next ∼1.5 km farther up structural section has <1% quartz veins, poorly developed stretching lineations, and tight folding, and the top of this portion marks the right-way-up garnet isograd. At the core of the syncline, folds are tight to open, stretching lineations are rare, and foliations defined by micas are parallel to original bedding. The southwestern flank of the syncline shows similar patterns and includes a large early Paleozoic granite body that is partially undeformed. The overall structural pattern of the southern Beas transect across the MCT hanging wall reveals a decrease in deformation intensity concomitant with a decrease in metamorphic grade from the bottom to the preserved top of the thrust sheet.

Northern MCT and Hanging-Wall Structures

In contrast to the southern MCT hanging wall, the northern MCT hanging wall is penetratively deformed at all observed outcrops (Figs. 6A–6D; stereoplots G1, G2, TG1, G4, G5). The northern exposure of the MCT at the eastern end of the Kullu window is defined by an ∼2-km-thick, east-northeast–dipping, top-to-the-southwest ductile shear zone with S-C fabric and normal drag shear bands. The lithologic contact of Greater Himalayan Crystalline complex and underlying Lesser Himalayan Sequence lithologies is concordant, and the shear zone extends ∼0.5 km below the contact. The Greater Himalayan Crystalline complex here and across the entire Himachal Himalaya displays mica-defined foliation, northeast-trending stretching lineation, tight to isoclinal folds (which are largely southwest and west verging) of schistosity, S-C fabric, and normal drag shear bands ranging from centimeter to meter scale in length (Figs. 6 and 8G–8I). Folds in southern Greater Himalayan Crystalline complex sections along the Beas and Pabbar Rivers commonly mirror characteristics associated with the large-scale Phojal anticline: southwest vergence, ∼20°–40° interlimb angles, folding of existing foliation, and development of a new moderately northeast-dipping axial planar cleavage overprinting shallow, dominantly northeast-dipping foliation.

STD and Hanging-Wall Structures

The STD occurs in two primary geometries across the Himachal Himalaya: subhorizontal to shallowly northeast-dipping right-way-up fault strands in the right-way-up limb of the Phojal anticline, and moderately northeast-dipping overturned fault strands in the overturned limb of the Phojal anticline (Figs. 3, 4, and 6). As described in our early work (Webb et al., 2007), the key evidence for the Phojal folding and concomitant overturning of the STD comprises (1) overturned top-to-the-northeast shear structures including extensional shear bands and σ-type porphyroclasts (Fig. 9A), (2) folded lithologies and contacts including the Haimanta schist–Greater Himalayan Crystalline complex gneiss contact and Haimanta marker lithologies of graphitic quartzite and garnet-hornblende calc-silicate schist (Figs. 9B, 9C), and (3) folding of metamorphic isograds, which could cross structures but appear to parallel the STD across the Himalaya (Fig. 5). By extrapolation discussed in our early work, we interpret that the Phojal anticline hinge plane and the overturned strand of the STD must both intersect the MCT across the eastern Kullu window, pass across the Pabbar River northwest of Rohru, and intersect the MCT again along the Uttarkashi half-window. The only readily accessible transect across such structures is along the Pabbar River (Figs. 6B–6D; stereoplots T8, G5). Our mapped position of the STD in this valley was determined on the basis of lithology, metamorphic grade, and folding consistent with the Phojal anticline and STD. We identified top-to-the-southwest shear structures, but not top-to-the-northeast shear structures, in the mapped position of the STD. However, exposure here is so spotty that such evidence of the 300–600-m-thick shear zone may well be covered. From south to north, rocks change from graphitic and garnet schist with garnet-hornblende calc-silicate schist marker layers to leucosome-bearing paragneiss, consistent with a Haimanta to Greater Himalayan Crystalline complex change. Along the mapped hinge zone of the Phojal anticline, west- and southwest-verging tight to isoclinal overturned folds with 10–50 m amplitudes are exposed on cliff sides; these folds may reflect the larger scale Phojal anticline.

The right-way-up STD shear zone across the region has both top-to-the-southwest and top-to-the-northeast shear structures [Figs. 6A–6D (stereoplots T1, T2, T3, T9) and 9D]. Above the shear zone, Haimanta metasedimentary rocks are foliated (foliation defined by micas and subparallel to commonly preserved bedding), commonly lineated, and foliation is folded within southwest- and northeast-verging tight to isoclinal outcrop-scale folds (Figs. 6A–6D; stereoplots T1, T2, T3, T9). Early Paleozoic granites are strongly deformed within the STD shear zone, but above it these are weakly deformed and undeformed where observed (∼15 km southwest of Manali, ∼5 km southwest of Morang along the Sutlej River, and ∼7 km east of Sangla). Apophyses extend from these granites, crosscut foliation, and are locally warped (Fig. 9E). These relationships demonstrate that some foliation developed in the Haimanta above the STD prior to early Paleozoic granitic crystallization, but at least some internal deformation of these rocks outlasted this intrusion. The largest fold in the STD hanging wall is the Tandi syncline, a northeast-verging overturned syncline with an ∼2 km amplitude defined by the contact of the Haimanta Group and the Tandi Group and by sedimentary layering, which is particularly well preserved within the Tandi Group (Figs. 9F, 9G). This may represent a large drag fold consistent with top-to-the-northeast deformation in the STD thrust sheet. The section of the STD hanging wall exposed along the upper Sutlej contains folds from outcrop scale (generally isoclinal to tight) to ∼100 m amplitude scale (generally tight to open); the entire region from near Morang to the Spiti-Sutlej confluence forms a broad synclinorium (Figs. 3, 4A, and 6C). At the north end of this synclinorium (near Puh), a southwest-dipping sequence has pelitic schist overlying fine-grained psammitic paragneiss, which overlies kyanite- and sillimanite-bearing migmatitic paragneiss of the Leo Pargil dome. The paragneiss may represent the Greater Himalayan Crystalline complex, but the southern transition to lower grade Haimanta metasedimentary rocks is uncertain. Our reconnaissance observations reveal C′ cleavage indicating a top-to-the-southwest sense of shear within the fine-grained paragneiss; we speculate that this zone may contain the STD.

In summary, the overall deformation pattern within the STD hanging wall matches that of the southern MCT hanging wall: above the respective shear zones, deformation intensity decreases with increasing structural levels, and early Paleozoic granites are partially undeformed.

GEOCHEMISTRY

Major and trace element geochemistry data were obtained for felsic plutonic rocks of different crystallization ages in order to characterize their sources, evolution, and tectonic settings. Five samples of the ca. 1.85 Ga Baragaon granitic gneiss (AW 9–4-03 3, AW 9–24–03 5, AW 9–26–03 1B, AW 9–27–03 1B, AW 9–29–03 6A), one sample of ca. 830 Ma granitic gneiss (AY 9–3-03 12A), and two samples from early Paleozoic granite (AW 9–4-03 6A, AW 9–7-03 5A) were analyzed at Actlabs using X-ray fluorescence and inductively coupled plasma– mass spectrometry (ICP-MS) (Figs. 6B and 10; Supplemental File 33; for age constraints, see Geochronology discussion). The analyzed samples are (1) granite (excepting sample AW 9–26–03 1B, a granodiorite) (Supplemental File 3A [see footnote 3]) and (2) calc-alkaline (excepting sample AW 9–29–03 6A, which is tholeiitic) (Supplemental File 3B [see footnote 3]). Percent SiO2 ranges from 65.3 to 73.1 and molar Al2O3/(Na2O + CaO + K2O) values between 1.38 and 2.04 show that the rocks are highly peraluminous (Supplemental File 3D [see footnote 3]). Chondrite-normalized rare earth element (REE) data show relative enrichment in light REEs (LaN/YbN 7.47–31.29). The observed strong enrichment in lighter trace elements and REEs is indicative of a dominant continental crust source (Fig. 10). The samples are characterized by negative Eu anomalies (EuN/Eu*N 0.15–0.55) and negative Ta, Nb, P, and Ti anomalies. The negative Eu anomalies show the involvement of plagioclase in fractional melting. Negative Ta, Nb, P, and Ti anomalies suggest an arc setting for the original melts that gave rise to these rocks and/or the remelted continental crust component of these rocks. Elemental data plotted on granite discrimination diagrams (after Pearce et al., 1984) overlap the boundary between the syncollisional and within-plate granite fields and plot near the volcanic arc field (Supplemental File 3C [see footnote 3]). The two early Paleozoic samples plot exclusively within the syncollisional field, but data from similar granites across the northwest India Himalaya (15 samples from Miller et al., 2001) show more variation, overlapping the boundary between the syncollisional and within-plate granite fields, near the volcanic arc field. In summary, the results demonstrate that granitic rocks of distinct ages (i.e., ca. 1.85 Ga, ca. 830 Ma, and early Paleozoic) are all dominantly sourced from remelted continental crust and may reflect an arc setting here during or prior to granitic magma generation.

THERMOBAROMETRY

The presence of pelitic bulk compositions allowed us to acquire pressure and temperature estimates of metamorphic conditions in the basal section of the Tethyan Himalayan Sequence Haimanta Group in the southern MCT hanging wall, i.e., southwest of the overturned STD (Table 3; Supplemental File 44). This work is required for comparison of this section with the Greater Himalayan Crystalline complex and with the Tethyan Himalayan Sequence section above the STD; thermobarometric studies of both have been done (Epard et al., 1995; Vannay and Grasemann, 1998; Vannay et al., 1999; Jain et al., 1999; Walker et al., 1999; Chambers et al., 2009). We analyzed six garnet-bearing mica schists using the University of California, Los Angeles (UCLA) JEOL 8200 electron microprobe. Mineral compositions were measured for thermobarometric calculations and X-ray composition maps of garnet crystals were obtained to assess zoning patterns. For individual mineral analyses we used a 4 μm spot size (<1 μm spot size for garnet), a 15 kV accelerating voltage, and a current of 15 nA; for X-ray maps we used a current of 100 nA with 1–5 μm pixels and a dwell time of 30–50 ms per pixel. Programs AX and THERMOCALC 3.21 (Holland and Powell, 1998) were employed to calculate the activities of phases and the equilibrium Clapeyron slopes of exchange (garnet-biotite) and net transfer (garnet-biotite-muscovite-plagioclase) reactions.

Samples AW 9–21–04 2B and AW 9–5-05 2A were collected from schist with garnet, muscovite, biotite, and plagioclase with average compositions of An15 and An21, respectively (Table 3). Garnet in these samples preserves dominantly growth zoning, with decreasing Mn and increasing Fe and Mg from core to rim (Supplemental File 4A [see footnote 4]). Minor retrograde reactions may be recorded by thin (<15 μm) outermost rims of high Mn. All quartz is recrystallized. Data from the low-Mn inner rims of garnet crystals were used in thermobarometric calculations to retrieve pressure-temperature (P-T) conditions of 8.5 ± 2.0 kbar and 629 ± 127 °C and 8.8 ± 2.0 kbar and 617 ± 124 °C for two thin section areas of sample AW 9–21–04 2B, and 9.3 ± 2.0 kbar and 566 ± 120 °C for sample AW 9–5-05 2A (Table 3; Supplemental File 4B [see footnote 4]).

Samples AW 9–11–03 6 and AW 9–13–03 3 were collected from schist with garnet, muscovite, biotite, and plagioclase with average compositions of An12 and An4, respectively (Table 3). Garnet in these samples preserves growth zoning, with decreasing Mn and increasing Fe and Mg from core to rim (Supplemental File 4A [see footnote 4]). All quartz is recrystallized. Data from the low-Mn rims of garnet crystals were used in thermobarometric calculations to retrieve P-T conditions of 6.9 ± 1.7 kbar and 446 ± 97 °C for sample AW 9–11–03 6, and 9.6 ± 2.3 kbar and 535 ± 116 °C for sample AW 9–13–03 3 (Table 3; Supplemental File 4B [see footnote 4]). These results are consistent with expectations based upon the pelitic petrogenetic grid (Spear and Cheney, 1989). However, the low An plagioclase of sample AW 9–13–03 3 is not in the range of the barometer, which is not well calibrated below An9 (Ghent and Stout, 1981). Comparison with other results in this region suggests that retrieved pressure conditions for this sample likely overestimate pressure by <3 kbar, which would also result in a modest (<30 °C) overestimate of maximum temperature.

Sample AW 9–3-03 12 was collected from schist with garnet, muscovite, biotite, and plagioclase with an average composition of An24 (Table 3). Garnet in this sample is unzoned in Fe, Mg, and Mn (Supplemental File 4A [see footnote 4]). All quartz is recrystallized. Garnet rim compositions were used to determine P-T conditions of 7.2 ± 1.6 kbar and 494 ± 101 °C (Table 3; Supplemental File 4B [see footnote 4]).

The collection site of sample AW 9–29–03 12 is at the top of the garnet zone in the right-way-up metamorphic sequence observed along the Beas River (Figs. 5 and 6B) (see also Epard et al., 1995); this sample is therefore expected to yield the lowest maximum P-T conditions of this sample suite. It was collected from schist with garnet, muscovite, biotite, and plagioclase with an average composition of An1 (Table 3). Garnet in this sample is unzoned in Fe, Mg, and Mn (Supplemental File 4A; see footnote 4). Most quartz is recrystallized, but ≤1 mm relict quartz grains occur as porphyroclasts wrapped around by the dominant quartz-mica fabric. Retrieved P-T conditions are 13.1 ± 2.9 kbar and 480 ± 105 °C (Table 3; Supplemental File 4B [see footnote 4]). These results are consistent with expectations based upon the pelitic petrogenetic grid (Spear and Cheney, 1989), but the very low An plagioclase is outside the calibrated range of the barometer (Ghent and Stout, 1981). Comparison with other regional results suggest that retrieved pressure conditions for this sample likely overestimate maximum pressure by ∼5–7 kbar, which would also result in a modest (<50 °C) overestimate of maximum temperature.

The new temperature and pressure determinations are consistent with metamorphic conditions of the structurally lowest levels of (1) the Lesser Himalayan Sequence Jeori crystalline rocks, (2) the Greater Himalayan Crystalline complex, and (3) the Tethyan Himalayan Sequence Haimanta Group to the north of the STD (e.g., Epard et al., 1995; Vannay and Grasemann, 1998; Vannay et al., 1999; Caddick et al., 2007; Chambers et al., 2008, 2009). The decrease in metamorphic grade at increasing structural levels is inconsistent with the famous inverted metamorphic field gradient of the Greater Himalayan Crystalline complex (Vannay and Grasemann, 1998; Vannay et al., 1999).

GEOCHRONOLOGY

U-Th-Pb Zircon Geochronology: Granitic Rocks

We conducted U-(Th)-Pb spot dating of zircon from granitoid and leucogranite samples using the Cameca 1270 ion microprobe at UCLA to constrain the lithostratigraphy and timing of deformation above, below, and within the MCT shear zone. We briefly describe our results here; Appendix 3 contains a description of our methods and an expanded report of our results. Granitoid ages generally increase with structural depth: the MCT hanging wall features early Paleozoic and ca. 830 Ma granites and granitic gneiss, the MCT zone contains ca. 830 Ma and ca. 1.85 Ga granitic gneiss, and the MCT footwall has ca. 1.85 Ga granitic gneiss. Dates of leucogranite crosscutting foliation in the Leo Pargil dome and pegmatite crosscutting foliation of the Wangtu gneiss provide minimum ages for these fabrics and age information on the thermal conditions represented by their crystallization.

To characterize the age and affinities of granitic rocks in the MCT hanging wall, we analyzed 53 spots on 38 zircon crystals from 5 samples (Fig. 11); 2 granitic samples yield early Paleozoic dates: AW 9–7-03 5A, from granite with 5 cm feldspars ∼5 km east of Mandi (Figs. 6B and 10F), and AW 9–3-04 7A, collected from an ∼1-m-thick mylonitic granite sill ∼20 km north of Kullu in the overturned limb of the Phojal anticline (Fig. 6A). Calculated weighted mean ages of 238U/206Pb dates from concordant data of these two samples are 470.6 ± 5.3 Ma (AW 9–7-03 5A; 2σ, mean square of weighted deviates, MSWD = 1.9) and 473 ± 26 Ma (AW 9–3-04 7A; 2σ, MSWD = 0.73). We interpret the weighted mean ages as the granite crystallization ages. Sample AY 9–3-03 12A was collected from mylonitic augen gneiss of the Audi shear zone. Results for this sample are reversely discordant and interpreted to record the protolith crystallization ca. 830 Ma (see Appendix 3). Sample AW 9–4-03 6A was collected from mylonitic augen gneiss ∼1 m structurally above a sheared top-to-the-southwest basal contact with Haimanta Group metasedimentary rocks (Fig. 6B). Results with a range of dates spanning the Paleoproterozoic to the early Paleozoic and variable concordance and/or discordance are interpreted to reflect zircon inheritance ca. 840 Ma or earlier, a granitic crystallization event between ca. 840 Ma and the early Paleozoic, and early Paleozoic metamorphic overgrowth of zircon rims. Sample AW 9–24–04 4C represents tourmaline-bearing leucogranite that crosscuts bedding preserved in amphibolite-grade metasedimentary rocks of the Leo Pargil dome (Fig. 6C). Acquired dates are variable, discordant, and extend from the Precambrian to the Tertiary. The youngest (238U/206Pb) date is 50.4 ± 2.4 Ma (1σ). These data likely reflect the ages of inherited zircon prior to Tertiary leucogranite crystallization.

We analyzed 71 spots on 51 zircon crystals from 7 samples to characterize the age and affinities of mylonitic augen gneiss within the MCT shear zone (Fig. 11). Analyses of six of these samples are consistent with protolith crystallization ca. 1.85 Ga (samples AW 9–4-03 3, AW 9–24–03 5, AW 9–26–03 1B, AW 9–27–03 1B, AW 9–29–03 6A, AW 9–17–05 2; Figs. 6B, 6C). These results help define the age and spatial extent of the ca. 1.85 Ga Baragaon gneiss within the MCT shear zone (although the affinity of sample AW 9–17–05 2 is ambiguous; see Appendix 3). Analyses of sample AW 9–18–04 1D (Fig. 6B) yield a 207Pb/206Pb weighted mean age of 831 ± 19 Ma (2σ, MSWD = 0.69), which we interpret as the protolith crystallization age. This is the southernmost MCT shear zone sample, and it provides a geometric limit on the Baragaon gneiss.

We analyzed two samples from a single site in the MCT footwall: AW 9–18–05 1B and AW 9–18–05 1C were collected from Wangtu gneiss and an undeformed pegmatitic felsic dike that crosscuts foliation of the Wangtu gneiss, respectively (Figs. 6C and 11). Discordant results for sample AW 9–18–05 1B are consistent with previous dates of ca. 1.85 Ga for the crystallization of the granitic protolith of the Wangtu gneiss (Table 2). Four spot analyses on different zircon crystals from sample AW 9–18–05 1C yield Late Miocene 238U/206Pb ages, with a weighted mean age 7.4 ± 1.2 Ma (2σ, MSWD = 5.1) (Supplemental File 55). Late Miocene 232Th/208Pb monazite ages were also obtained for this sample (see Th-Pb Monazite Geochronology discussion). We interpret the Late Miocene dates from the undeformed dike as crystallization ages. Therefore the dikes result from partial melting and recrystallization during the concluding phases of the Late Miocene deformation and metamorphism of the Munsiari Group.

U-Th-Pb Zircon Geochronology: Detrital Zircon

We conducted U-Th-Pb geochronology of detrital zircon grains from 16 samples to test models for the composition and correlation of sequences carried by major thrust faults. All ages are spot ages collected using the laser ablation multicollector (LA-MC) ICP-MS facility at the University of Arizona (Tucson) and the UCLA ion microprobe Appendix 4 contains a description of our methods. Results from the Lesser Himalayan Sequence Shimla Group, the Greater Himalayan Crystalline complex, and the Tethyan Himalayan Sequence Haimanta Group all span the Proterozoic. In contrast, results from the Lesser Himalayan Sequence Berinag and Jeori Groups are dominantly Paleoproterozoic and older.

MCT Footwall

Because the Shimla Group of the Outer Lesser Himalayan Sequence has been alternately considered Paleoproterozoic and Neoproterozoic (see preceding), zircon grains from a siltstone sample of this unit (AW 9–15–04 4) were analyzed to constrain the depositional age (Fig. 12M). Sparse data (only 23 suitable analyses) show a concentration of 11 dates ranging from 600 Ma to 850 Ma, indicating a late Neoproterozoic maximum depositional age for the sampled layer.

Berinag Group quartzites were investigated to constrain their depositional age and examine along-strike and regional map correlations. Previous workers have alternately correlated quartzites between the Wangtu gneiss and the Greater Himalayan Crystalline complex on the northeastern flank of the Kullu window as Berinag Group or Greater Himalayan Crystalline complex rocks (cf. K.K. Sharma, 1977; Vannay and Grasemann, 1998). Therefore we analyzed detrital zircon from this setting and from the thick Berinag Group quartzites southwest of the Munsiari thrust. Samples AW 9–22–04 9A and AW 9–22–04 9C are from an ∼50-m-thick, highly recrystallized quartzite layer between Wangtu orthogneiss and Greater Himalayan Crystalline complex paragneiss at the eastern end of the Kullu window, whereas sample AW 9–22–04 1B is a quartzite from the Berinag Group near Rampur. These three samples all yield youngest age peaks between ca. 1.88 Ga and ca. 1.94 Ga and are dominated by Paleoproterozoic ages, consistent with late Paleoproterozoic maximum depositional ages. The two Berinag samples from just below the Greater Himalayan Crystalline complex paragneiss have distinct age distributions, one with substantial population of dates as old as ca. 3.0 Ga and the other showing a single, dominant ca. 1.88 Ga age peak. Similarly, our Berinag sample from south of the Munsiari thrust has a range of ages from ca. 1.9 to 3.0 Ga, whereas 6 of 7 ages from a similarly located sample from Richards et al. (2005) have ca. 1.87 Ga dates. The similar age populations obtained from quartzites of the northeastern and southwestern Kullu window support grouping these rocks into the Berinag Group. The two age patterns obtained in both regions suggest that the Berinag Group may contain layers with distinct sources.

Sparse younger dates were obtained from the Berinag samples from the eastern end of the Kullu window: AW 9–22–04 9C yields 5 scattered younger ages, with 238U/206Pb ages of 969 ± 26 Ma, 846 ± 8 Ma, 572 ± 39 Ma, 513 ± 14 Ma, and 469 ± 3 Ma, and AW 9–22–04 9A likewise yields 238U/206Pb ages as young as the Phanerozoic. The concordia plot of sample AW 9–22–04 9A shows that the young ages define a clear discordia line from ca. 1.85 to 1.9 Ga to the Cenozoic (Supplemental File 5B [see footnote 5]), and a single discordant approximately Cenozoic age has a high U/Th ratio. These results are consistent with minor zircon growth and Pb loss during Cenozoic metamorphism, so these young dates are not considered detrital ages. AW 9–22–04 9C data with Paleoproterozoic 207Pb/206Pb ages also show discordia lines trending toward the late Phanerozoic, but the five Neoproterozoic–early Paleozoic analyses listed above plot near or on the concordia curve. Our preferred interpretation is that the five young dates reflect Pb loss and not detrital ages, but if these represent real detrital ages then sample 9C could represent Paleoproterozoic to Cenozoic Himalayan strata. In summary, we interpret the data for the Berinag samples to indicate a maximum depositional age of ca. 1850 Ma and Pb loss related to Cenozoic metamorphism. Deposition soon after ca. 1850 Ma is consistent with the ca. 1.8 Ga age of interlayered and crosscutting metabasalt (Miller et al., 2000).

We performed detrital zircon geochronology of the Jeori gneiss in order to test a proposed correlation of these rocks with the MCT hanging-wall rocks to the south of the Kullu window (i.e., the Jutogh thrust hanging wall of Richards et al., 2005). The age patterns obtained for Jeori paragneiss samples (AW 9–25–04 8 and AW 9–25–04 9) span ca. 1.9 to 3.0 Ga. Age distributions of the Jeori samples and Berinag samples, excepting only Berinag sample AW 9–22–04 9A, are indistinguishable at the 95% confidence level according to the Kolmogorov-Smirnov statistic (Fig. 13; Supplemental File 5b [see footnote 5]). However, the minimum age peaks of the two Berinag samples and the Jeori rocks are distinct: the Berinag samples show younger minimum age peaks ca. 1.92 Ga and scattered ca. 1.8 to 1.94 Ga ages for the Berinag versus ca. 1.94 Ga and ca. 1.96 Ga for the Jeori). These data suggest a maximum depositional age of ca. 1.94 Ga for the protoliths of the Jeori metasedimentary samples. The lack of a ca. 1.85 Ga signal suggests that the Jeori rocks were deposited prior to crystallization of the Wangtu granitic protolith. The proposed correlation of Jeori gneiss with the MCT hanging-wall rocks south of the Kullu window is not supported by these data, because as detailed herein, these latter rocks yield Neoproterozoic detrital zircon age peaks.

MCT Hanging Wall

We analyzed zircon grains from the MCT hanging wall, focusing on the metasedimentary rocks of the southern MCT hanging wall (Figs. 12 and 13; Supplemental File 5 [see footnote 5]). Our mapping suggests that these are Tethyan Himalayan Sequence Haimanta Group rocks, but alternative models show these as Greater Himalayan Crystalline complex rocks, a combination of Tethyan Himalayan Sequence and Greater Himalayan Crystalline complex rocks, or divided into multiple Greater Himalayan Crystalline complex and Lesser Himalayan Sequence thrust sheets (e.g., Bhargava et al., 1991; Thakur and Rawat, 1992; Frank et al., 1995; DiPietro and Pogue, 2004; Richards et al., 2005).

Analyses of MCT footwall rocks (described above), a Tethyan Himalayan Sequence Haimanta Group sample from a site widely acknowledged to be in the Tethyan Himalayan Sequence Haimanta Group strata above the STD (e.g., Thakur and Rawat, 1992; Frank et al., 1995; Vannay and Steck, 1995; Jain et al., 1999), and a Greater Himalayan Crystalline complex sample provide bases of comparison for proposed correlations of the southern MCT hanging wall with the other thrust sheets. The Tethyan Himalayan Sequence Haimanta sample, AW 8–31–04 5, was collected from a quartz-mica phyllite in Keylong (Fig. 6A). We obtained 98 zircon dates from this sample, yielding well-constrained dates ranging from ca. 550 Ma to ca.3000 Ma (Figs. 12 and 13; Supplemental File 5 [see footnote 5]). There is no dominant age probability peak, but the largest are ca.940 Ma and a broad peak ca. 2560 Ma, with additional peaks ca.560, ca.770, and ca. 890 Ma (Fig. 12F). The Greater Himalayan Crystalline complex sample, AW 9–22–04 10B, was collected from Sutlej River section to the east of the Kullu window (Fig. 6C). Spot ages on 105 zircon grains yield well-constrained dates ranging from ca. 770 Ma to ca. 2600 Ma, with broad peaks ca. 820 Ma and ca. 2520 Ma (Fig. 12L). Detrital zircon results for the metasedimentary rocks of the southern MCT hanging wall (AW 9–5-03 8, AW 9–18–03 12, AW 9–26–03 3, AW 9–27–03 1A, AW 9–19–04 10, AW 9–27–04 5) have broadly similar age ranges and dominant age peaks to these two samples, with well-constrained ages ranging from ca. 550 Ma to ca. 3500 Ma.

The broadly shared detrital age pattern of the Haimanta, Greater Himalayan Crystalline complex, and southern MCT hanging-wall samples includes a Neoarchean through Proterozoic span of ages with two dominant age peak zones, one from ca. 2600–2500 Ma and the other across the Neoproterozoic. One sample, AW 9–27–03 1A, does not prominently show either age peak zone. It is a sample of phyllonite with microscopic garnet that occurs <1 m above the (sharp, mylonitic, folded) contact with ca. 1.85 Ga Baragaon gneiss in the southern MCT hanging wall (Figs. 6B and 8B). The 37 ages for this sample span the Proterozoic, but the only significant age peak is a diffuse ca. 1.8 Ga peak, suggesting that the Baragaon rocks were a dominant source for these zircon grains. The original contact may have been sedimentary; an unconformity between correlative lithologies is observed in the Pakistan Himalaya (DiPietro and Isachsen, 2001).

The consistent age populations of southern MCT hanging-wall metasedimentary samples support hypotheses that group these rocks in a single unit. As the age populations match results for a recognized Tethyan Haimanta Group sample (AW 8–31–04 5), these data support field evidence suggesting that all of these rocks are Tethyan Himalayan Sequence Haimanta Group. Similar age ranges for the Greater Himalayan Crystalline complex mean that these data alone cannot distinguish between the Haimanta Group and the Greater Himalayan Crystalline complex. The detrital age patterns and protoliths of the Greater Himalayan Crystalline complex, Haimanta Group, and Shimla Group of the Outer Lesser Himalayan Sequence are all similar, suggesting that these sequences may restore as a single stratigraphic layer prior to Phanerozoic tectonism.

Th-Pb Monazite Geochronology

Because monazite commonly crystallizes during metamorphism, Th-Pb monazite geochronology can be used to date metamorphic events (e.g., Harrison et al., 1997; Catlos et al., 2001); monazite is also commonly used to date igneous rocks (e.g., Harrison et al., 1995). Th-Pb analyses of monazite grains were obtained via the Cameca 1270 ion microprobe at UCLA using a beam current of 10–15 nA focused to a size of 15–30 μm in diameter. Th-Pb ratios were determined using a calibration curve based primarily on ThO2/Th versus Pb/Th from monazite standard 554 (45 ± 1 Ma; Harrison et al., 1995). Monazite grains in metamorphic rocks were dated in situ, allowing direct examination of the textural relationships between monazite grains and metamorphic fabrics. These monazite grains were commonly very small (<15 μm in diameter) and were identified using the scanning electron microscope in backscatter mode.

Despite examining thin sections of ∼40 garnet schist samples of the MCT hanging wall southwest of the Kullu window, only 4 samples contained sufficient monazite to warrant analysis: AW 9–3-03 12, AW 9–11–03 4, AW 9–11–03 6, and AW 9–13–03 3 (Fig. 6B; Table 4). The first three samples yielded 13 latest Proterozoic to early Mesozoic dates from monazite grains included in garnet, and in the matrix. For these samples, ages range from ca. 600 Ma to ca. 200 Ma, with 1σ analytical errors of as much as ∼33 m.y. Dates do not all overlap within 2σ for any one sample, even after matrix monazite dates are excluded. The range of ages may result from remaining difficulties in dating of pre-Cenozoic monazite (Peterman et al., 2006). Sample AW 9–13–03 3 yielded Cenozoic dates. The only monazite grain included in garnet yielded the oldest spot date, 41.5 ± 7.2 Ma (1σ), while younger spot ages from matrix monazite grains range from 39.9 ± 6.5 to 28.4 ± 2.3 Ma (1σ).

Monazite grains were analyzed from a leucogranite intruding the upper levels of the Greater Himalayan Crystalline complex (AW 9–23–04 3) and an undeformed pegmatite intruding the Wangtu (described above, AW 9–18–05 1C). Greater Himalayan Crystalline complex sample AW 9–23–04 3 ages range from the Paleozoic to the Cenozoic; the Cenozoic dates are 26.8 ± 1.5, 20.1 ± 0.9, and 19.9 ± 1.3 Ma (1σ). AW 9–18–05 1C had a tight range of ages spanning from 6.59 ± 0.37 to 5.87 ± 0.37 Ma (1σ).

These results show that the MCT hanging-wall metasedimentary rocks southwest of the Kullu window underwent early Phanerozoic and mid-Cenozoic garnet-grade or greater metamorphism. Data for sample AW 9–13–03 3 show that the Tethyan Himalayan Sequence Haimanta Formation above the MCT underwent garnet-grade metamorphism ca. 41.5 Ma, and that this metamorphism may have persisted through ca. 28.5 Ma. The early Phanerozoic ages add to a growing body of evidence for Himalayan metamorphism at that time (e.g., Gehrels et al., 2003). We interpret the range of ages from the leucogranite sample (AW 9–23–04 3) to reflect inheritance; the youngest ages of ca. 27–20 Ma indicate the crystallization age. The ages of ca. 6.6–5.8 Ma obtained for the pegmatite intruding the Wangtu crystalline rocks (sample AW 9–18–05 1C) reflect either crystallization or cooling indicating early uplift along the Munsiari thrust and/or related thrusts.

THERMOCHRONOLOGY

We acquired 40Ar/39Ar thermochronologic data to constrain the cooling history of the MCT hanging wall to the southwest of the Kullu window; ∼4–25 mg of biotite, muscovite, and potassium feldspar were separated, hand-picked, and irradiated along with sanidine from the 27.8 Ma Fish Canyon Tuff (Cebula et al., 1986: Renne et al., 1994) for 15–45 h at University of Michigan and McMaster University. The sanidine and the unknowns were step-heated in a double vacuum furnace and isotopic compositions of the released gas were measured using a VG 1200 mass spectrometer at UCLA. Data reduction was accomplished via the in-house program AGECAL.EXE, and uncertainties were calculated at the 1σ level. Results do not include uncertainties in J-factors or decay constants. Multidiffusion domain modeling (Lovera et al., 1989) of potassium feldspar data for four samples (AW 9–4-03 6A, AW 9–19–03 3, AW 9–24–03 5, and AW 9–27–03 1B) was conducted to extract thermal history information. Representative mica and feldspar results for a single sample are shown in Figure 14; age spectra, thermal history results, and data tables are in Supplemental File 66.

Excess argon affects these samples: only one muscovite spectra displays a plateau-type pattern, many biotite plateau ages are older than the range of the muscovite spectra for the same sample, and half of the individual analyses along K-feldspar spectra are older than mica ages for the same sample. Nonetheless, these data provide meaningful information on the cooling history of the MCT hanging wall.

Maximum Ar mica cooling ages for the MCT hanging-wall rocks above the Baragaon gneiss must be younger than ca. 40 Ma, because the age of a monazite included in garnet from sample AW 9–13–03 3, a garnet schist in these rocks, is ca. 41.5 Ma. Some mica cooling ages for these rocks are much older (AW 9–3-03 12, AW 9–7-03 5A), suggesting the presence of excess Ar, others are roughly equivalent to ca. 40 Ma (AY 9–3-03 12A, AW 9–4-03 6A), and the youngest mica age is 31.6 ± 0.6 Ma (muscovite from sample AW 9–4-03 6A). This suggests a maximum age of cooling below ∼375 °C of ca. 30 Ma for the MCT hanging-wall rocks above the Baragaon gneiss. For the Baragaon gneisses, biotite cooling spectra for samples AW 9–4-03 3, AW 9–19–03 3, AW 9–26–03 1B, and AW 9–27–03 1B provide a maximum age for cooling below ∼325 °C of ca. 50 Ma.

The K-feldspar age spectra, representing samples from the ca. 830 Ma gneiss, the Mandi granite, and the Baragaon gneiss, all show a single consistent pattern. With more than ∼40% of the 39Ar released, the individual analyses give ages much older than mica ages for these rocks and the ca. 41.5 Ma included monazite age, so the spectra after this point yield no interpretable age information. However, between 0% and ∼40% gas release, all of the spectra form a roughly parabolic pattern, the minimum of which may be taken as a maximum age plateau. The minimum age along these parabolic age spectra is ca. 15–20 Ma or younger. We used the multidomain diffusion model (Lovera et al., 1989) to model the minima of these age spectra parabolas, with resultant thermal histories that essentially represent a point in the cooling path. This point is consistently at ∼220–230 °C at 13–19 Ma, representing the maximum time that these rocks were at these temperatures. The coherent cooling pattern in the Baragaon gneisses and the overlying Haimanta Group likely indicates that these rocks behaved as a single tectonic unit during Middle Miocene cooling.

DISCUSSION

Our mapping documents variations in structural style across thrust hanging walls in the Himachal Himalaya, showing large-magnitude internal shortening and penetrative deformation across almost all layers. U-Th-Pb geochronology of igneous and detrital zircon reveals layers that can be correlated across major structures including the MCT and STD, i.e., ca. 1.85 Ga orthogneiss and Neoproterozoic siliciclastic layers. Additional geochronologic and geochemical analyses show ca. 1.85 Ga, ca. 830 Ma, and ca. 470 Ma granitic rocks derived primarily from remelted continental crust, and reveal a previously unknown phase of granitic pegmatite intrusion ca. 8–6 Ma. These pegmatites are undeformed and crosscut Wangtu gneiss foliation, providing a minimum age for this fabric; a similar record of early Paleozoic granite crosscutting foliation in the STD hanging wall indicates fabric development here prior to the Cenozoic Himalayan orogeny. Combined P-T results and Th-Pb monazite geochronology show that garnet-producing metamorphism occurred in the early Phanerozoic and middle Cenozoic in the southern MCT hanging wall. We discuss the implications of these findings for the applicability of current Himalayan tectonic models to the Himachal region, the pre-Cenozoic stratigraphic framework, and the Cenozoic tectonic evolution.

Viability of Himalayan Tectonic Models in the Himachal Himalaya

Proposed Himalayan tectonic models make distinct predictions for the affinity of high-grade rocks, fault kinematics, exhumation history, along-strike changes in structural geometry and stratigraphy, and metamorphism (Fig. 2; Table 1; see review in the Introduction).

Most Himalayan tectonic models show the Greater Himalayan Crystalline complex as derived from the India plate, but in early channel flow models (Nelson et al., 1996; Hodges et al., 2001) these rocks represent Tibetan material that has tunneled southward into the Himalaya. Our detrital zircon results reveal similar age populations in the Greater Himalayan Crystalline complex, Haimanta Group of the lower Tethyan Himalayan Sequence, and the Outer Lesser Himalaya (Figs. 12 and 13). Because the Haimanta Group and Outer Lesser Himalaya are indisputably of India plate origin, these data support the interpretation that the Greater Himalayan Crystalline complex comprises Indian materials.

Wedge extrusion and early channel flow models show the STD as a solely top-to-the-north shear zone, whereas channel flow models that incorporate asymmetric thrust extrusion (e.g., Beaumont et al., 2004) and the tectonic wedging model show alternating top-to-the-north and top-to-the-south kinematics along this structure. Our results and previous investigations show alternating kinematics along the STD in the Himachal Himalaya (Choudhuri et al., 1992; Jain et al., 1999; Vannay et al., 2004; Webb et al., 2007).

The exhumation of the Greater Himalayan Crystalline complex occurs during STD activity (Early and Middle Miocene) in wedge extrusion and channel flow models, but after STD activity (post–Middle Miocene) in the tectonic wedging model. The overturning of the southernmost STD within the Phojal anticline requires deformation of the leading edge of the Greater Himalayan Crystalline complex after cessation of motion along the STD, so these frontal Greater Himalayan Crystalline complex rocks could not be eroded during STD activity. Furthermore, previous studies indicated that the leading edge of the Greater Himalayan Crystalline complex remains buried in parts of the Himachal Himalaya, an interpretation that has been repeatedly determined for the western Himachal geology (e.g., Thakur and Rawat, 1992; Frank et al., 1995; Fuchs and Linner, 1995; Thakur, 1998; DiPietro and Pogue, 2004; Yin, 2006).

The wedge extrusion and early channel flow models predict along-strike continuity of the three major Himalayan tectonic units (Lesser Himalayan Sequence, Greater Himalayan Crystalline complex, Tethyan Himalayan Sequence) and bounding faults (MCT, STD). In contrast, the tectonic wedging model shows a southern termination of the Greater Himalayan Crystalline complex where the MCT and STD merge, such that the southern MCT hanging wall is continuous with the STD hanging wall. The metamorphic pattern predicted by the models corresponds to their overall structural geometry: for the wedge extrusion and channel flow models, the MCT hanging wall consists of only Greater Himalayan Crystalline complex rocks and thus displays the famously inverted metamorphic field gradient of the Greater Himalayan Crystalline complex, whereas the tectonic wedging model predicts an MCT hanging wall with the inverted Greater Himalayan Crystalline complex metamorphic pattern in the north and the right-way-up Tethyan Himalayan Sequence metamorphic pattern to the south.

Critical tests of these predictions were provided by investigation of the southern MCT hanging wall and comparison with the northern MCT and STD hanging walls. Some results do not distinguish these three sheets. (1) Protolith lithologies are similar for all three hanging walls, and detrital zircon age populations from them show similar age spectra spanning from ca. 550 Ma to ca. 3500 Ma (Figs. 12 and 13). (2) U-Pb zircon and Th-Pb monazite geochronology confirm that early Paleozoic granites intrude the southern MCT hanging wall (Fig. 9; Table 4). Granitic rocks of this age intrude the STD hanging wall (e.g., Miller et al., 2001) and are also associated with the Greater Himalayan Crystalline complex where the original contacts are sheared (Frank et al., 1977). (3) P-T determinations of ∼450–630 °C and ∼7–9 kbar obtained for the southern MCT hanging wall (Table 3) are consistent with the metamorphic conditions of the structurally lowest levels of the Greater Himalayan Crystalline complex and the STD hanging wall (e.g., Vannay and Grasemann, 1998; Vannay et al., 1999; Chambers et al., 2009). (4) Th-Pb monazite dating and crosscutting relationships indicate the preservation of early Paleozoic metamorphic fabrics in the southern MCT hanging wall and STD hanging wall (Fig. 9E; Table 4; see also Marquer et al., 2000; Chambers et al., 2009), but records of metamorphism of this age are reported from the Greater Himalayan Crystalline complex in other parts of the Himalaya (e.g., Gehrels et al., 2003). Other results support the correlation of the southern MCT hanging wall with the STD hanging wall. (1) In contrast to the penetratively deformed Greater Himalayan Crystalline complex, both the southern MCT and the STD hanging walls feature undeformed and weakly deformed granitic rocks (Figs. 8E and 9E). (2) Metamorphic grade increases upsection across the Greater Himalayan Crystalline complex along the well-studied Sutlej section at the east end of the Kullu window, matching the famous Greater Himalayan Crystalline complex inverted metamorphic field gradient observed across many Himalayan sections (Figs. 5 and 6C) (e.g., Vannay and Grasemann, 1998; Vannay et al., 1999). However, the southern MCT hanging wall only displays an inverted metamorphic field gradient within the ∼2-km-thick MCT shear zone (our mapping; see also Grasemann et al., 1999; Gregory, 2004). Metamorphic grade decreases with increasing structural elevation across the rest of the southern MCT hanging wall from garnet grade to conditions below biotite grade, matching the STD hanging wall (Fig. 5) (Epard et al., 1995; Chambers et al., 2009; this study). (3) No fault separates the rocks of the southern MCT hanging wall from rocks of the Chamba syncline area to the northwest, where the same lithologies are overlain by Permian–Jurassic Tethyan sedimentary rocks along a depositional contact (Figs. 3, 6A and 6B) (e.g., Frank et al., 1995).

In summary, tectonic wedging is the only tectonic model that accommodates all of the following findings for the Himachal Himalaya: detrital zircon results that indicate Indian crustal affinity of the Greater Himalayan Crystalline complex, alternating top-to-the-north and top-to-the-south STD kinematics, Greater Himalayan Crystalline complex exhumation only after cessation of STD motion, and the connected southern MCT and STD hanging walls with shared Tethyan affinity and pattern of metamorphic isograds. An important further consideration is that previous workers have inferred Early–Middle Miocene exhumation of the Greater Himalayan Crystalline complex on the basis of the detrital record without consideration of the similar evolution of the lower Tethyan Himalayan Sequence and the Greater Himalayan Crystalline complex (e.g., DeCelles et al., 1998; White et al., 2002). Because the Greater Himalayan Crystalline complex and the lower Tethyan Himalayan Sequence (the Haimanta Group and associated granites) are similar in protolith lithologies, isotopic constraints, zircon age populations, and even metamorphic grade, these criteria cannot be used to distinguish between these two units at isolated bedrock exposures or in the detrital record. In the Himalayan mountains, high-grade metamorphism (kyanite, sillimanite, migmatite) and structural setting may distinguish these units. In the detrital record, only the high-grade mineral record is sufficient, but it cannot distinguish between high-grade minerals from Tso Morari in the northernmost Himalaya (e.g., Schlup et al., 2003) and those from the Greater Himalayan Crystalline complex.

Pre-Cenozoic Stratigraphic Framework of Northern India

A key task in unraveling the Cenozoic tectonic history of the northwest India Himalaya is the construction of a pre-Cenozoic stratigraphic framework. This task is commonly done by stacking younger units atop older units and including intrusive bodies. However, the magnitude of thrusting in the Himalaya, e.g., at least ∼150 km along the combined distance of the Tons thrust and MCT, far exceeds the distance along which regular stratigraphy is maintained in the Indian craton (cf. Rao et al., 2000). Rather, ca. 2.5 Ga to Mesozoic bedrock is exposed over such distances. Since a high percentage of the deformed Himachal Himalaya rocks were Indian craton rocks (correlations discussed in the following), we made our pre-Cenozoic stratigraphic framework by first considering which fault-bound slices were at the surface as the India-Asia collision commenced and then determining the relative undeformed position of these slices from southwest to northeast.

Late Cretaceous and early Cenozoic rocks may be used as a proxy for the solid Earth's surface at the beginning of the collision, so sequences that are overlain by such rocks along sedimentary contacts are vertically constrained in the restored cross section. Three deformed sequences are thus constrained: (1) the Deoban and Damtha Groups, overlain by the latest Cretaceous–Middle Eocene Singtali-Subathu rocks; (2) the Outer Lesser Himalaya, also overlain by the latest Cretaceous–Middle Eocene Singtali-Subathu rocks; and (3) the Tethyan Himalayan Sequence, which contains Cretaceous strata (Figs. 3 and 4A). Of these three, the Deoban and Damtha Groups originated furthest to the southwest because (1) correlative Indian craton strata of the Vindhyan Supergroup underlie undeformed Sub-Himalayan stratigraphy beneath the Main Himalayan thrust (Powers et al., 1998); (2) Deoban rocks are deformed with Sub-Himalaya rocks in the frontal fold and thrust belt, e.g., Deoban rocks and depositionally overlying Subathu and Dharamsala rocks are thrust to the southwest along the Bilaspur thrust (e.g., Srikantia and Sharma, 1976); and (3) the Outer Lesser Himalaya moved southwest along the Tons thrust over Deoban and Damtha rocks locally capped by Singtali Formation and/or Subathu Formation rocks (e.g., West, 1939; Valdiya, 1980). Therefore the Tons thrust footwall rocks restore to the southwest relative to the hanging-wall rocks. Although the Outer Lesser Himalaya is thrust over the Sub-Himalaya, occupying the outer (i.e., farther south) portion of the Lesser Himalaya in much of the Himachal Himalaya, it represents a far-traveled thrust sheet relative to the Deoban and Damtha Groups. The Outer Lesser Himalaya originated to the southwest of the Tethyan Himalayan Sequence, as the northern Mesozoic sedimentary rocks of the latter represent the shelf and slope of the Indian Neotethys continental margin (e.g., Gaetani and Garzanti, 1991).

With the relative positions of the known surface fault slices thus constrained, we next correlate deformed rocks of matching age and lithology into single pre-deformation layers: First, our detrital zircon results and other data sets indicate that the Haimanta Group, Greater Himalayan Crystalline complex, and Lesser Himalayan Shimla Group represent deformed parts of a formerly continuous unit. These units share similar protoliths (Table 2), their Nd-Sr isotopic signatures match (see Supplemental File 2 [see footnote 2]), detrital zircon results from all three units show similar age spectra (Figs. 12 and 13), and the time window for deposition of all these units is narrow and equivalent, at the end of the Proterozoic. The older age limit is provided by the detrital zircon age data and ranges from ca. 550 to 800 Ma for different samples. The younger age limit is determined by the early Paleozoic intrusions and metamorphic events for both the Haimanta Group [our U-(Th)-Pb zircon and Th-Pb monazite geochronology] and the Greater Himalayan Crystalline complex (Walker et al., 1999; Miller et al., 2001), and overlying early Cambrian strata for both the Haimanta Group and the Shimla Group (e.g., Hughes and Jell, 1999; Myrow et al., 2006). Therefore a Neoproterozoic sedimentary sequence can be traced across both the MCT and STD. One significant distinction is that no evidence for early Paleozoic tectonism has been discovered in the Outer Lesser Himalaya, whereas both the Tethyan Himalayan Sequence Haimanta Group and the Greater Himalayan Crystalline complex record early Paleozoic granitic magmatism and garnet growth in metapelitic lithologies (our work; Frank et al., 1977; Miller et al., 2001; Gehrels et al., 2003). Since the Outer Lesser Himalayan rocks restore farther to the south than the other units, this difference may be accommodated if early Paleozoic tectonism was limited to the northernmost margin of India.

Second, we correlate the Baragaon granitic gneiss to the Wangtu granitic gneiss. The protolith of the discontinuous Baragaon granitic gneiss along the MCT shear zone crystallized ca. 1.85 Ga (Fig. 9). The Baragaon gneiss thus matches the Wangtu gneiss in age and lithology (Table 2; Fig. 9; Singh et al., 1994; Miller et al., 2000; Richards et al., 2005). This gives a second lithostratigraphic connection across the MCT. These rocks and the Jeori metasedimentary rocks are the oldest deformed units. We restore these to the base of the pre-collision stratigraphic framework.

Third, limited age constraints and lithologic similarities allow two additional tentative correlations: the Damtha Group may be laterally equivalent to the Berinag Group, and the upper, Neoproterozoic portion of the Deoban Group may be laterally equivalent to the Basantpur Formation of the Outer Lesser Himalaya (Figs. 12 and 13; Table 2).

The above considerations suggest a general model for a pre-deformation cross section involving a layer-cake stratigraphy gently tilted to the northeast, so that younger units progressively pinch out at the surface to the southwest (Fig. 15A). In a preferred model, we modify this layer-cake model such that the Deoban and Berinag-Damtha units pinch out to the northeast (Fig. 15B). The modifications are supported by the following considerations. (1) The Outer Lesser Himalaya is locally underlain by the quartzites and metavolcanics that may correlate with the Damtha-Berinag succession (the Darla volcanics, ∼25 km to the northwest of Shimla). (2) The Outer Lesser Himalaya is locally underlain by ca. 1.85 Ga mylonitic granitic gneiss just to the southeast of the area shown in Figure 3 (Célérier et al., 2009a); this gneiss correlates with the ca. 1.85 Ga Wangtu-Baragaon gneiss. Further support for this model includes the juxtaposition of the Baragaon granitic gneiss directly below the Haimanta Group along the base of the southern MCT hanging wall (Figs. 3 and 4). The Baragaon appears to also underlie the Greater Himalayan Crystalline complex in the same fashion in Himachal along the Tons River (Figs. 3 and 6C) and to the northwest along the Kishtwar window (Stephenson et al., 2001). At the Nanga Parbat and Indus Syntaxes of the Pakistan Himalaya, rocks correlative to the Outer Lesser Himalaya–Haimanta Group–Greater Himalayan Crystalline complex overlie ca. 1.85 Ga granites and older rocks along an unconformity (e.g., DiPietro et al., 1999; DiPietro and Isachsen, 2001; DiPietro and Pogue, 2004).

This reconstruction of relative pre-Cenozoic structural positions explains the curious thrusting of younger rocks atop older rocks in the Himachal Himalaya. In most fold and thrust belts, older rocks are thrust over younger rocks because deep rocks are thrust over shallow rocks and older rocks commonly occur deeper than the younger rocks (e.g., Dahlstrom, 1969). However, in the Himachal Himalaya the opposite relationship is commonly observed in the Himachal Himalaya, e.g., the northern portions of both the Tons thrust and the MCT place Neoproterozoic rocks atop Mesoproterozoic and/or Paleoproterozoic rocks. The pre-Cenozoic relative positions of the Deoban and Damtha Groups, the Outer Lesser Himalaya, and the Tethyan Himalayan Sequence shows that younger rocks originated at equal and greater depth in the northeast compared with older rocks in the southwest. Therefore in general, from northeast to southwest young rocks thinned and pinched out such that progressively older rocks were exposed at the surface at the start of Cenozoic Himalayan shortening.

Structural Origin of the Kullu and Uttarkashi Windows

The elongated dome may represent a fault-propagation fold, with the Munsiari thrust as the propagating fault. We speculate that the tip line of the Munsiari thrust may cut upsection and perhaps locally offset the MCT (as shown in Fig. 3 along the southern margin of the Kullu window; see also Fig. 4). We further infer that the Chaura thrust is folded with this anticlinorium (Figs. 3 and 4). The thrust has been mapped within 10 km of the Sutlej River along its northeastern trace (Singh and Jain, 1993; Jain et al., 2000; Chambers et al., 2008; our observations); the steep to overturned southwestern trace shown in Figures 3 and 4 is inferred. Although the Munsiari thrust generally dips shallowly or moderately to the northeast, the fault is locally folded. For example, the fault is folded dips along the Tons River near Sankri, such that it dips to the southwest at ∼1 km west of Sankri (Fig. 3) (Jain and Anand, 1988). The Munsiari fault is inactive, at least locally, as it is deformed by younger, deeper structures.

We interpret the structural package of deformed Munsiari Group rocks as an antiformal stack (Fig. 4). This interpretation is based upon our interpreted Chaura thrust geometry and the local deformation of the Munsiari thrust coupled with rapid, recent uplift of Munsiari Group rocks. The antiformal stack has six horses as drawn, the Jeori and Wangtu rocks representing two long horses. In this interpretation the Wangtu and Jeori rocks originate at the same structural level, and the total shortening across the stack would equal the length of all horses minus the deformed length, or ∼150 km.

The geometric relationship of the Tons thrust to the Berinag thrust and their kinematic histories are key factors for determining the structural evolution of the Uttarkashi half-window, but these are poorly understood. The Tons and Berinag thrusts may represent a single thrust that cut upsection to the south, thus transporting progressively younger hanging-wall rocks to the south (Fig. 16). Alternatively, these may be distinct structures that were juxtaposed by successive footwall accretion or out-of-sequence thrusting.

Kinematic Reconstruction

We present a schematic Cenozoic tectonic history of the Himachal Himalaya by progressively deforming our preferred pre-Cenozoic stratigraphic framework (Fig. 17). Contraction and crustal thickening across the Tethyan Himalayan Sequence and the associated burial and metamorphism of the Tethyan Himalayan Sequence and Greater Himalayan Crystalline complex was underway by 45–40 Ma and persisted through the Oligocene (Wiesmayr and Grasemann, 2002; Catlos et al., 2007; Chambers et al., 2009; our results). Emplacement of the Greater Himalayan Crystalline complex between the MCT and STD was accomplished by the end of the Middle Miocene, as cessation of plastic deformation along the STD in this period is recorded by 40Ar/39Ar muscovite ages in the immediate STD hanging wall and footwall (Vannay et al., 2004; Thiede et al., 2005). We show a tectonic wedging kinematic evolution for Greater Himalayan Crystalline complex emplacement; the preceding discussion details why this is the only possibility among current proposed tectonic models. Top-to-the northeast slip along the STD likely surfaced to the north along the GCT (Fig. 2). Up until the Miocene, the Himalayan fold-and-thrust belt is thin-skinned, but by the Middle Miocene the Baragaon gneiss was accreted from the Indian basement into the MCT zone.

Continued accretion of basement slices in the Late Miocene resulted in stacking of the Wangtu and Jeori thrust horses. The folding of the Phojal anticline at the leading edge of the Greater Himalayan Crystalline complex must have occurred prior to this significant downcutting of the Himalayan sole thrust. We also show the accretion of supracrustal thrust horses of the Tons thrust sheet (the Outer Lesser Himalaya) and the Berinag thrust sheet into the growing thrust wedge in the Late Miocene. The accretion of the Tons thrust sheet at this time offers an alternative to our early interpretation of Tons thrust emplacement in the Eocene–Oligocene (see Célérier et al., 2009a). Some structures deforming the Himalayan foreland basin during this period are preserved in the Bilaspur thrust hanging wall. In the Pliocene–Quaternary, additional basement slices are incorporated into the growing thrust wedge. These horses have yet to be exposed at the surface. The combination of basement horse stacking and out-of-sequence faulting along the Munsiari thrust warped the MCT and created the Kullu window. Likewise, the duplexing of Damtha and Deoban Group rocks in the Bilaspur thrust hanging wall warped the overlying Tons thrust and MCT hanging walls into the Narkanda half-window and Shimla klippe. Forward propagation of the Himalayan thrust wedge incorporated additional sections of the foreland basin into fold and thrust belts.

CONCLUSIONS

This work examines the viability of models for the emplacement of the Himalayan crystalline core (i.e., the Greater Himalayan Crystalline complex) between two north-dipping faults (i.e., the MCT and STD). The differing predictions of the main models, i.e., wedge extrusion (e.g., Burchfiel and Royden, 1985), channel flow/focused denudation (e.g., Beaumont et al., 2001), and tectonic wedging (e.g., Webb et al., 2007) (Table 2), were examined in the Himachal Himalaya, where the rocks carried along the MCT change from Greater Himalayan Crystalline complex rocks in the east to Tethyan Himalayan Sequence rocks in the west (e.g., Thakur, 1998; DiPietro and Pogue, 2004; Yin, 2006). Recent field mapping (Webb et al., 2007) shows that the MCT and STD intersect in this region, bounding the leading edge of the Greater Himalayan Crystalline complex. Southwest of this branch line, the MCT carries Tethyan Himalayan Sequence rocks directly over Lesser Himalayan Sequence rocks, as confirmed in this work by metamorphic variations and detrital zircon geochronology. This tectonic framework is consistent with a tectonic wedging emplacement of the Greater Himalayan Crystalline complex.

We determined stratigraphic correlations across the MCT and STD, leading to a new model of the pre–India-Asia collision stratigraphic framework of this region. The restored positions of various stratigraphic levels vary from north to south, such that younger rocks to the north were commonly deeper than older rocks to the south. These geometries explain the thrusting of younger rocks over older rocks in the Himalaya. The protoliths and detrital zircon records of Neoproterozoic rocks in the Lesser Himalaya, Greater Himalayan Crystalline complex, and Tethyan Himalaya are similar, suggesting that these units formed a single stratigraphic horizon prior to Cenozoic tectonism. Our new tectonic model shows three main phases in the Cenozoic construction of the Himachal Himalaya: early contraction and crustal thickening, emplacement of the Greater Himalayan Crystalline complex as a tectonic wedge between the MCT and STD, and accretion of thrust slices from the subducting Indian plate to the growing orogenic wedge.

APPENDIX 1. EXPANDED DESCRIPTION OF LITHOLOGIC UNITS

Major lithologic units in the study area are the Cretaceous and Cenozoic Sub-Himalayan Sequence, the Proterozoic and Cambrian Lesser Himalayan Sequence, the high-grade Greater Himalayan Crystalline complex, and the Neoproterozoic to Mesozoic Tethyan Himalayan Sequence (Figs. 3 and 4) (Table 2). These units were described briefly (see text discussion of Lithologic Units); expanded descriptions are provided here.

Sub-Himalayan Sequence

This sequence consists of lower shallow-marine strata and upper continental deposits; the two subunits are separated by an Oligocene unconformity (Table 2). The lower subunit is in depositional contact with the underlying Lesser Himalayan Sequence (see following) in the Main Central thrust (MCT) hanging wall ∼15 km north and ∼25 km west of Shimla (Pilgrim and West, 1928; West, 1939; Srikantia and Sharma, 1976) and in the footwall of the Tons thrust ∼50 km north and ∼20 km east-southeast of Dehra Dun (Jain, 1972; Bhargava, 1976; Valdiya, 1980) (Fig. 3). The relationships require that the MCT and Tons thrust have moved over the very top of the Proterozoic–Cambrian units of the Lesser Himalayan Sequence.

Lesser Himalayan Sequence

The Lesser Himalayan Sequence consists of (1) the Neoproterozoic–Cambrian Outer Lesser Himalayan Sequence in the hanging walls of the Krol and Tons thrusts, (2) the Paleoproterozoic to Neoproterozoic Damtha and Deoban Groups in the hanging wall of the Bilaspur thrust and the footwalls of the Tons and Berinag thrusts, (3) the Paleoproterozoic Berinag Group in the hanging wall of the Berinag thrust, and (4) the Paleoproterozoic Munsiari Group dominantly in the hanging wall of the Munsiari thrust (Table 2) (Figs. 3 and 4). Because these units are juxtaposed by faults, understanding the age relationships requires unraveling the history of the bounding faults. Here we describe the age and lithology of each unit.

Neoproterozoic–Cambrian Outer Himalayan Sequence

This 7.5-km-thick unit consists of, from bottom to top, the Darla volcanic unit, Basantpur Formation, Shimla Group, Krol Group, and Tal Formation (Table 2) (Valdiya, 1980; Kumar and Brookfield, 1987; Jiang et al., 2003). Its basement as observed in the Garhwal Himalaya southeast of our study area consists of 1.85 Ga augen gneiss (Célérier et al., 2009a) (Fig. 1). The Darla unit is only exposed near Shimla below the Basantpur Formation. The Basantpur Formation consists of limestone, siltstone, shale, and locally minor slate (Table 2). The Shimla Group comprises shale, siltstone, sandstone, minor graywacke, conglomerate, and tillite (Pilgrim and West, 1928; Valdiya, 1970; Srikantia and Sharma, 1976; Kumar and Brookfield, 1987). It has been correlated with the Paleoproterozoic Damtha Group in the MCT footwall (e.g., Valdiya, 1995) or alternatively with the Neoproterozoic Haimanta Formation in the MCT hanging wall (e.g., Kumar and Brookfield, 1987; Miller et al., 2001). As shown here, our U-Pb detrital-zircon dating supports correlation of the Shimla and Haimanta rocks. The Krol Group, dominated by dolostone and limestone, is below the Tal Formation; the latter comprises sandstone and siltstone and contains early Cambrian trilobites (Myrow et al., 2003).

Paleoproterozoic to Neoproterozoic Damtha and Deoban Groups

The Damtha Group comprises siliciclastic strata and the Deoban Group is dominated by carbonates with shale and siltstone in its uppermost section (Table 2). The latest Paleoproterozoic to Neoproterozoic age of the Deoban Group is constrained from fossils (mainly stromatolite); these rocks depositionally overlie the Damtha Group (Valdiya, 1995).

Paleoproterozoic Berinag Group

The Berinag Group consists dominantly of quartz arenite locally interbedded with metabasalt flows and minor slate and intruded by sills and dikes (e.g., V.P. Sharma, 1977; Valdiya, 1980; Miller et al., 2000). The metaigneous rocks underwent greenschist facies metamorphism (Miller et al., 2000).

Paleoproterozoic Munsiari Group

The unit consists of two subunits bounded by two thrusts at their bases: the upper Wangtu gneiss bounded by Chaura thrust and the lower Jeori gneiss bounded by the Munsiari thrust (e.g., Singh and Jain, 1993; Valdiya, 1995; Jain et al., 2000; Chambers et al., 2008). The Wangtu gneiss is dominated by orthogneiss with minor intercalated metasediments, whereas the Jeori gneiss is mostly paragneiss (Table 2). Undeformed pegmatitic felsic dikes with garnet and centimeter-scale muscovite books and feldspar locally crosscut the Wangtu gneissic foliation (Fig. 7B). Kyanite and sillimanite in metasediments are present in the Jeori and the lower part of the Wangtu gneisses (Fig. 5). The Munsiari rocks underwent Late Miocene prograde metamorphism and record an inverted temperature profile (∼600–700 °C) with a normal pressure gradient (∼9–7 kbar) (Vannay and Grasemann, 1998; Vannay et al., 1999, 2004; Caddick et al., 2007; Chambers et al., 2008). Apatite fission track ages as young as ca. 1 Ma and 40Ar/39Ar muscovite ages as young as ca. 5 Ma are recorded from the Munsiari rocks (Fig. 4A) (Jain et al., 2000; Thiede et al., 2004, 2005, 2009; Vannay et al., 2004), which are probably related to recent and possibly active development of a duplex system below the Kullu window.

Greater Himalayan Crystalline Complex

The Greater Himalayan Crystalline complex in the MCT hanging wall is ∼7–9 km thick and consists of paragneiss, schist, and orthogneiss intruded by minor Tertiary leucogranites concentrated mostly in its upper 2–3 km. Metasediments in the complex are younger than 830 Ma, as constrained by U-Pb detrital-zircon dating (Richards et al., 2005; our observations, see following) and orthogneiss was crystallized in the early Paleozoic, based on Rb-Sr whole-rock isochron dates (Frank et al., 1977).

An inverted metamorphic field gradient is observed along the Sutlej River, which is expressed by garnet-staurolite-bearing rocks at the base, kyanite and sillimanite in the middle, and migmatitic rocks near the top across the Greater Himalayan Crystalline complex (Fig. 5) (Vannay and Grasemann, 1998). Oxygen isotope thermometry indicates that peak temperatures at the base of the Greater Himalayan Crystalline complex are ∼600 °C and rise to ∼750 °C in the uppermost section; calculated pressures from the garnet-muscovite-aluminosilicate-plagioclase barometer are constant across the Greater Himalayan Crystalline complex at ∼8 kbar (Fig. 4A) (Vannay et al., 1999). Northwest of the Sutlej River near Tos, Manali, and Khoksar, the Greater Himalayan Crystalline complex rocks yield peak temperatures ranging from ∼550 to ∼700 °C and pressures ranging from ∼6 to ∼10 kbar (Epard et al., 1995; Jain et al., 1999; Walker et al., 1999; Wyss, 2000; Verma et al., 2005). The estimated temperatures from the northwestern area were calculated from garnet-hornblende and garnet-biotite pairs and may represent minimum values of the peak metamorphic conditions due to later retrograde metamorphism (Vannay and Grasemann, 1998).

The U-Pb dating of monazite grains from kyanite schist near Manali yields a discordia line on a concordia diagram with upper and lower intercepts of 498.6 ± 20.2 Ma and 30.1 ± 1.3 Ma, respectively (Walker et al., 1999). The older intercept age was likely induced by intrusion of Cambrian–Ordovician granites (Frank et al., 1977, 1995). Th-Pb spot ages for monazite grains from Greater Himalayan Crystalline complex metapelite along the Sutlej River range from 39.6 ± 2.8 Ma to 22.8 ± 0.4 Ma (E. Catlos, 2004, personal commun.); monazite Th-Pb spot ages of ca. 44–19.5 Ma were obtained from a similar section along the Bhagirathi River (Foster et al., 2000; Catlos et al., 2007). The Greater Himalayan Crystalline complex rocks in the study area were exhumed to below ∼400 °C in the Early Miocene, as indicated by 40Ar/39Ar muscovite ages, but the metamorphic complex did not cool down below 110–60 °C until ca. 2–4 Ma, as indicated by the young zircon and apatite fission track ages (Jain et al., 2000; Schlup, 2003; Vannay et al., 2004; Thiede et al., 2004, 2005, 2009).

Tethyan Himalayan Sequence

This sequence consists of the Neoproterozoic–early Cambrian Haimanta Group, Cambrian Parahio Formation, and remaining Paleozoic–Mesozoic strata above. The Haimanta Group comprises metasedimentary rocks and is intruded by early Paleozoic granites (Table 2) (e.g., Jäger et al., 1971; Frank et al., 1995; Miller et al., 2001; Chambers et al., 2009). Its basement is likely the Paleoproterozoic Baragaon gneiss in the MCT shear zone directly below the unit (Bhanot et al., 1978; Miller et al., 2000; this study). South of the MCT–South Tibet detachment (STD) branch line, metamorphic grades of the Haimanta Group decrease upward in the MCT hanging wall, whereas north of the branch line the Haimanta rocks show a similar upward decrease in metamorphic grade in the STD hanging wall. In both cases, the decreasing metamorphic grade is expressed by the presence of garnet schist in the basal section and phyllite and slate in the upper section, as well as the regional pattern of metamorphic isograds (Fig. 5). Grade progression within the shear zones is distinct: an inverted metamorphic field gradient occurs in the 2-km-thick MCT shear zone (Grasemann et al., 1999; Gregory, 2004; Law et al., 2010), whereas metamorphic patterns are right-way-up across the STD (Vannay and Grasemann, 1998). Estimated P-T conditions in the basal Haimanta Group are 5–8 kbar and 450–660 °C (Epard et al., 1995; Vannay and Grasemann, 1998; Vannay et al., 1999; Jain et al., 1999; Wyss, 2000; Chambers et al., 2009; our observations, see following). In contrast, illite crystallinity indicates that the upper Haimanta Group underwent metamorphism at temperatures of only 200–300 °C (Wiesmayr and Grasemann, 2002). Using Raman spectroscopy of carbonaceous material, Gregory (2004) obtained temperatures of 499 ± 5 °C near Mandi and 520 ± 4 °C and 529 ± 9 °C near Narkanda for garnet schist in the basal Haimanta Group.

The U-Pb monazite dating of Haimanta garnet schist suggests that it underwent prograde metamorphism at 34–27 Ma (Chambers et al., 2009). Sparse 40Ar/39Ar muscovite thermochronology yields Early to Middle Miocene ages in the basal portions of the Haimanta Group (Schlup, 2003; Vannay et al., 2004; Thiede et al., 2005; Chambers et al., 2009). Zircon fission track ages of the Haimanta Group rocks range from Late Triassic to Early Miocene, and apatite ages span the Middle Miocene to Pliocene (Schlup, 2003). Late Miocene to Pliocene apatite ages are dominantly from the lowermost part of the Haimanta Group (Lal et al., 1999; Schlup, 2003; Vannay et al., 2004; Thiede et al., 2009).

The Cambrian Parahio Formation is dominated by shale, siltstone, fine sandstone, and minor carbonates. It likely represents continuous sedimentation of the Haimanta Group (Myrow et al., 2006). The remaining Paleozoic and Mesozoic Tethyan Himalayan Sequence strata are above an Ordovician angular unconformity (e.g., Wiesmayr and Grasemann, 2002). In the Tandi-Chamba and Tso Morari areas (western and northeastern parts of our map area in Fig. 3), Cambrian–Carboniferous strata are missing, whereas Permian and younger strata directly overlie the Haimanta Group (Fig. 3) (e.g., Steck, 2003).

Nd and Sr Chemostratigraphy

Results from previous investigations of the Nd and Sr isotopic characteristics of pre-Cenozoic rocks from the Himachal Himalaya and other Himalayan regions show two largely distinct groups in Nd and Sr isotopic space (Table 2; Supplemental File 2 [see footnote 2]). These groups can be distinguished by age: Mesoproterozoic and older rocks yield εNd (500) <∼–14 and a broad range of 87Sr/86Sr (500) values, whereas Neoproterozoic and younger rocks yield εNd (500) >∼–14 and a narrow range of 87Sr/86Sr (500) values. The only exception is the Berinag Group metabasalts, which are consistent with the Neoproterozoic and younger group despite an age of ca. 1.8 Ga (Supplemental File 2 [see footnote 2]) (Bhat and Le Fort, 1992, 1993; Miller et al., 2000). Data for Berinag Group quartzites are consistent with the Mesoproterozoic and older group of Himalayan rocks (Ahmad et al., 2000; Richards et al., 2005).

APPENDIX 2. EXPANDED DESCRIPTION OF REGIONAL TECTONIC FRAMEWORK

The first-order structures in the Himachal Himalaya were described, with emphasis on the Main Central thrust (MCT) and South Tibet detachment (STD) (see text discussion, Regional Tectonic Framework). Here we offer some additional information on those two major structures and comprehensive discussion of other structures.

Sub-Himalayan Thrust Zone

The Sub-Himalayan thrust zone in the Himachal Himalaya is bounded by the Main Frontal thrust (MFT) below and the Krol and Mandi thrusts above. The bounding faults of the thrust zone are not parallel; the northern Krol-Mandi thrust system defines the Dehra Dun reentrant in the south and the larger Kangra reentrant in the north (Fig. 3). The MFT places Neogene–Quaternary strata over the modern Indo-Gangetic Plain deposits and has an estimated Quaternary slip rate of 10 ± 6 mm/yr (Fig. 3) (Kumar et al., 2006). The Krol thrust places Proterozoic–Cambrian strata of the Outer Lesser Himalayan Sequence over Paleogene–Quaternary strata of the Sub-Himalayan Sequence, whereas the Mandi thrust juxtaposes Proterozoic strata of the Lesser Himalayan Sequence (Berinag, Deoban, Damtha units) over Paleogene–Quaternary strata of the Sub-Himalayan Sequence (Srikantia and Sharma, 1976; Powers et al., 1998; Raiverman, 2000). The Sub-Himalayan thrust zone comprises a southwestern subzone exposing Middle Miocene and younger strata and a northeastern subzone dominated by Early Miocene and older strata (Fig. 3). These subzones are separated by the Bilaspur and Palampur thrusts. The Bilaspur thrust along its southern trace carries Paleogene–Early Miocene strata (Subathu, Lower Dharamsala units) over Middle Miocene–Quaternary strata (Upper Dharamsala, Siwaliks units). The thrust cuts downsection of the hanging wall northwestward, exposing the Proterozoic Deoban Group and the Paleogene Subathu Formation. The slip along the Bilaspur fault is partitioned between the Mandi thrust and fault-fold systems linked to the Palampur thrust. The Palampur thrust places Paleogene–Miocene strata (Subathu, Dharamsala, Lower Siwalik units) over Middle Miocene–Quaternary strata (Siwaliks).

Internal structures of the southwestern subzone are complicated, expressed by southwest-directed thrusts, northeast-directed backthrusts, and tight folds at which thrusts terminate (Figs. 3 and 4A; Supplemental File 1A [see footnote 1]). The presence of backthrusts suggests the development of several thrust wedges in the thrust zones. As thrusts tend die out into folds, fault-propagation folds must also be present in the area. A minimum amount of shortening from the Main Frontal thrust to the Palampur thrust in the Kangra Reentrant is ∼23 km, which is equivalent to ∼22% shortening strain (Powers et al., 1998).

The Bilaspur-Palampur thrust systems and the Krol-Mandi thrust systems join in the northwest and southeast in map view, bounding the northeastern Sub-Himalayan thrust belt between. This map pattern is of a typical forward-propagating thrust duplex, with the Krol-Mandi thrust zone as an apparent roof fault and the Mandi-Bilaspur thrust zone as an apparent floor fault. However, a closer examination of the Krol-Mandi thrust zone and its footwall strata suggests that the Krol-Mandi thrust zone is an out-of-sequence structure. This conclusion is derived from the map relationship, where the Krol-Mandi thrust zone truncates beds in its footwall and locally cuts downsection in the thrust transport direction; the latter is a typical relationship for an out-sequence thrust, but contradicts the forward-propagating duplex system in which beds directly below the root thrust are parallel to the fault above (Boyer and Elliott, 1982; cf. Morley, 1988; Yin and Kelty, 1991).

Tons Thrust

The Tons thrust exhibits antiformal geometry and places the Outer Lesser Himalayan Sequence over the Deoban and Damtha Groups of the Lesser Himalayan Sequence; the latter are locally overlain by Cretaceous–Eocene strata (Fig. 3) (e.g., West, 1939; Srikantia and Sharma, 1976; Valdiya, 1980; Célérier et al., 2009a). The fault merges with the MCT to the north and the Krol thrust to the south along the Sutlej River. It may also link with the Berinag thrust in the southern Uttakashi area, as shown in Figure 3.

Berinag Thrust System

The Berinag thrust is exposed in the Kullu window where it appears in both the hanging wall and footwall of the Munsiari thrust (Fig. 3) (V.P. Sharma, 1977; Tewari et al., 1978; Valdiya, 1980; Srivastava and Mitra, 1994; Vannay and Grasemann, 1998; Vannay et al., 2004; Célérier et al., 2009a). In the Munsiari hanging wall the Berinag thrust is antiformal and places the Berinag Group over the Wangtu gneiss, whereas in the Munsiari footwall, the same fault exhibits synformal geometry and places the Berinag Group over the Damtha Group. As implied by this relationship, the Berinag thrust is offset by the underlying Munsiari thrust bounding the base of the Wangtu gneiss (Fig. 3).

In the Uttarkashi half-window, the southern exposure of the thrust places the Berinag Group over the Damtha Group and the northern exposure of the thrust places the Berinag Group over the Wangtu gneiss. The Berinag thrust merges with the MCT, as observed in the Kullu window, and the thrust is truncated by the Munsiari thrust, based on reinterpretation of the existing map (Jain and Anand, 1988) and our field observations. The systematic change in juxtaposition relationship from north to south across the Berinag thrust in the Kullu window and Uttarkashi half-window areas requires the fault to climb upsection from a detachment horizon along the Berinag-Wangtu contact to a north-dipping footwall ramp across the Berinag and Damtha strata to link with a higher footwall flat on top of the Damtha Group.

The linking of the Tons and Berinag thrusts in the westernmost Uttarkashi half-window led to the correlation of the Nagthat Formation in the Shimla Group (Table 2) with the Berinag Group (e.g., Valdiya, 1980; Srivastava and Mitra, 1994). This correlation was considered not feasible due to their different ages and chemostratigraphic signatures (e.g., Ahmad et al., 2000; Miller et al., 2000, 2001; Richards et al., 2005). The Shimla Group may have been deposited on top of the Berinag Group, and the two units together form the hanging wall of the linked Tons-Berinag thrust (as drawn in Fig. 3).

Munsiari Thrust

The Munsiari thrust is marked by a mylonitic shear zone involving an ∼1–2-km-thick section of Munsiari Group gneiss in the hanging wall and ∼1-km-thick Berinag Group in the footwall (Jain and Anand, 1988; Singh and Jain, 1993; Vannay et al., 2004; our observations). All observed mylonitic fabrics are consistent with a top-to-the-southwest sense of shear (Jain and Anand, 1988; Vannay et al., 2004; our observations). Distributed deformation also occurs in the Munsiari hanging wall and footwall. Jain and Anand (1988) reported an ∼1-km-thick ductile shear zone located ∼2 km below the fault with a top-to-the-southwest sense of shear. The Wangtu gneiss in the hanging wall is also extensively mylonitized or isoclinally folded, consistent with top-to-the-southwest shear (see also Jain and Anand, 1988; Singh and Jain, 1993; Vannay et al., 2004). Brittle fabrics (e.g., Riedel shears, cataclasite, and slickenfibers) overprinting the ductile fabrics in the Munsiari thrust zone also indicate a top-to-the-southwest sense of shear (Jain and Anand, 1988; Singh and Jain, 1993; Vannay et al., 2004; our observations).

Chaura Thrust

This fault divides the Munsiari Group in the eastern Kullu window and places the Wangtu gneiss over the Jeori gneiss. The fault merges with the Munsiari thrust in map view (Fig. 3). The Chaura thrust zone preserves S-C fabric and σ-type porphyroclasts that are consistent with top-to-the-southwest shear (Singh and Jain, 1993; Jain et al., 2000; Chambers et al., 2008; our observations).

Main Central Thrust

Surface Trace and Fault Kinematics

In the map area the MCT is folded and displays large full and half-windows and isolated klippes: major thrust windows include the Kullu window, and the Narkanda and Uttarkashi half-windows, and klippes include the Shimla, Lansdowne, and Almora klippes, and two small klippes ∼15 km southeast of Mussoorie (Figs. 3 and 6B). As discussed herein, the MCT in this region is a top-to-the-southwest mylonitic shear zone as much as 2 km thick (e.g., Grasemann et al., 1999; Vannay et al., 2004). Brittle deformation is restricted to the region along the Sutlej River, where the base of the MCT zone is a top-to-the-southwest brittle thrust placing the mylonitic Baragaon gneiss and metasedimentary rocks of the Haimanta Group over Deoban carbonates (Grasemann et al., 1999; our observations). Grasemann et al. (1999) investigated the mylonitic fabrics of the MCT zone within the Baragaon gneiss here, using fringe folds and quartz c- and <a> axes patterns to infer a decelerating strain path (i.e., early high-temperature simple shear deformation progressing to general shear deformation due to an increase in pure shear component as the shear zone cooled).

Hanging-Wall and Footwall Cutoffs and MCT Geometry in Cross-Section View

As noted herein, cutoff relationships in the MCT footwall suggest that the thrust cuts upsection to the southwest in its transport direction. Along the northern rims of the Kullu window and Uttarkashi half-window, the thrust places the Greater Himalayan Crystalline complex rocks over the Wangtu gneiss, which is considered to the basement of the Lesser Himalayan Sequence strata. Farther south, the MCT cuts progressively the Berinag Group in the southern Kullu window and the central Uttarkashi half-window, the Damtha and Deoban Groups in the Narkanda half-window and southernmost Uttarkashi half-window, the youngest Outer Lesser Himalayan Sequence below the Shimla klippe, and an MCT protrusion south of Naura directly east of the Shimla klippe (Fig. 3). The southernmost trace of the MCT along the northern limb of the Kangra reentrant merges with the Mandi thrust and places the Tethyan Himalayan Sequence directly over the Himalayan foreland-basin strata; MCT klippes occur along the leading edge of the preserved thrust sheet (Figs. 3, 4, and 6B). These footwall cutoff relationships suggest that the MCT origination was at least as deep as along the contact between the basement and the cover sequence of the northern Indian passive margin. Throughout its evolution, the MCT climbed up the entire section of the passive margin sequence and cut across the foreland strata. As the Baragaon gneiss in the MCT hanging wall and the Wangtu gneisses in the MCT footwall both represent the basement of the Indian craton, the MCT must be a basement-involved thick-skinned thrust in the Himachal Himalaya. This fault characteristic is remarkably similar to the MCT fault in the eastern Himalaya, where it also involves and duplicates the Indian basement rocks (Yin et al., 2010a, 2010b).

As discussed herein, the large footwall ramp along the MCT is not entirely due to the MCT cutting upsection: because Cretaceous–Eocene beds overlie progressively older units to the north (Fig. 3), the Proterozoic–Cambrian sequence of the northern Indian passive margin must have been tilted to the south prior to the emplacement of the MCT hanging wall. Some apparent MCT ramping merely reflects this earlier tilting.

The MCT along the northern limb of the Kangra reentrants truncates an east-trending anticline involving 11–7 Ma Late Miocene Middle Siwalik strata (Fig. 3) (Table 2), suggesting that the MCT must have been locally reactivated since ca. 7 Ma.

South Tibet Detachment

Surface Trace and Fault Kinematics

The trace and kinematics of the STD are discussed within the main body of the text. Two further notes: (1) another possible STD exposure occurs at the Leo Pargil dome, where high-grade rocks were exhumed along northeast-trending normal faults that initiated ca. 16–14 Ma, but the southern transition to lower grade Haimanta metasedimentary rocks is uncertain (Thiede et al., 2006). We speculate that the STD may be exhumed along the southern and possibly northern flanks of the Leo Pargil dome, such that the dome is cored by Greater Himalayan Crystalline complex rocks. (2) Although the northwest-trending Sarchu normal fault has been interpreted to link the STD mapped in the Himachal Himalaya in the east to the Zanskar shear zone in the west (e.g., Wyss et al., 1999), it is a steep brittle structure with total slip <2 km, making it an impossible candidate as an extension of the STD (cross section in Supplemental File 1A [see footnote 1]; Epard and Steck, 2004).

Great Counter Thrust System and Hanging-Wall Structures

The southern limb of the Spiti Synclinorium comprises the top-to-the-southwest Tethyan Himalayan fold-thrust belt, whereas its northeastern limb consists of a southwest-verging tight to isoclinal fold system commonly referred to as the Mata nappe (Figs. 3 and 4B) (Steck et al., 1998). The Mata nappe is tilted to the southwest and thrust to the southwest along its synclinal axial plane; this fault may daylight as the Dutung-Thaktote thrust (Steck et al., 1998). The Tso Morari ultrahigh-pressure (UHP) rocks underlie the Mata nappe. Evidence of UHP metamorphism is scarce; both the base of the Mata nappe and the UHP rocks record dominantly amphibolite facies metamorphism (e.g., Epard and Steck, 2008). Top-to-the-southwest ductile shear fabrics are dominant across the upper levels of the Tso Morari rocks through the lower levels of the Mata nappe (Epard and Steck, 2008). Both the UHP rocks and the Mata nappe carapace are domed within open and north-verging folds (e.g., Stutz and Steck, 1986; Steck et al., 1998). The northwest-trending subvertical Ribil fault uplifted these rocks relative to India-Asia suture zone rocks dominated by the Nidar ophiolite (e.g., Schlup et al., 2003; de Sigoyer et al., 2004). Farther north, the suture zone rocks are thrust northeastward over the Indus molasse along a strand of the latest Oligocene–Miocene Great Counter thrust system (GCT) (e.g., Schlup et al., 2003; de Sigoyer et al., 2004). The youngest structures in the Tso Morari region are north-trending normal faults that may link with northwest-trending right-lateral strike-slip faults (e.g., Epard and Steck, 2008).

UHP metamorphism is loosely constrained as ca. 55 Ma (e.g., de Sigoyer et al., 2000; Leech et al., 2005, 2007; O'Brien, 2006; cf. Gouzu et al., 2006), followed by amphibolite facies metamorphism and deformation of both the UHP rocks and the Mata nappe and exhumation of these rocks to the middle crust by ca. 40 Ma (de Sigoyer et al., 2000, 2004; Epard and Steck, 2008). Apatite fission track ages decrease from ca. 35 Ma in the south at the base of the Mata nappe to ca. 8 Ma across the Ribil fault (Schlup et al., 2003). The tectonic evolution of the Tso Morari UHP rocks and surrounding units remain uncertain (cf. Schlup et al., 2003; de Sigoyer et al., 2004; Yin, 2006; Epard and Steck, 2008; Beaumont et al., 2009). Following Yin (2006), we show the doming of the UHP rocks and motion along the Ribil fault as deformation within the GCT (Figs. 3 and 4B). This fault system overprints the India-Asia suture zone, juxtaposing Tethyan Himalayan Sequence rocks, suture zone sedimentary and ophiolitic rocks, and the Indus molasse atop the Ladakh batholith of the Asian plate (Yin, 2006; Epard and Steck, 2008). The GCT likely represents the northern extension of the STD (e.g., Yin et al., 1994, 1999).

Definition of the Main Boundary Thrust

Following the definition of Heim and Gansser (1939) that the Main Boundary thrust (MBT) places pre-Tertiary strata over Tertiary strata, we may designate the northern portion of the Bilaspur thrust to be the MBT. However, this definition breaks down for the same fault to the south, as its hanging-wall strata change from Proterozoic to Tertiary. The MBT definition of Heim and Gansser (1939) is also locally applicable to the Mandi-Krol thrust system and even to the MCT along the northern limb of the Kangra reentrant (Fig. 3). Furthermore, the folded Tons thrust places Proterozoic Lesser Himalayan Sequence strata over slivers of Cretaceous and Paleogene strata in its footwall along the southern margins of the Narkanda and Uttarkashi half-windows and in the southeast corner of our map area, qualifying this fault locally as MBT. The above examples illustrate the limitation of the traditional definition of the MBT that can lead to miscorrelation of different structures along strike as the same fault (see discussion in Yin, 2006). For this reason, we do not use the term Main Boundary thrust in our structural description of the Himachal Himalaya in this paper.

APPENDIX 3. EXPANDED REPORT OF U-Th-Pb GEOCHRONOLOGY OF ZIRCON FROM GRANITIC ROCKS

Results of U-(Th)-Pb spot dating of zircon from granitoid and leucogranite samples were summarized in the discussion U-Th-Pb Zircon Geochronology: Granitic Rocks. This appendix includes a description of our methods and an expanded report of our results.

Methods

Analytical procedures reported in Quidelleur et al. (1997) were followed to acquire spot dates using the Cameca 1270 ion microprobe at UCLA. This work was accomplished using an 8–15 nA O primary beam and an ∼25-μm-diameter spot size. U-Pb ratios were determined using a calibration curve based on UO/U versus Pb/U from zircon standard AS3 (age 1099.1 Ma; Paces and Miller, 1993), and isotopic ratios were adjusted for common Pb with a 204Pb correction from the model of Stacey and Kramers (1975). Data reduction was accomplished via the in-house program ZIPS 3.0.3 by Chris Coath.

Main Central Thrust Hanging Wall

We analyzed 53 spots on 38 zircon crystals from 5 samples to characterize the age and affinities of granitic rocks in the Main Central thrust (MCT) hanging wall. Sample AY 9–3-03 12A was collected from mylonitic augen gneiss of the Audi shear zone (Fig. 6B); 12 spot dates on 9 zircon crystals from this sample were obtained (Fig. 11; Supplemental File 5A [see footnote 5]). The analyses form a cluster ca. 830 Ma in concordia space (Fig. 11). Only two ages are reversely disconcordant within 1σ error ellipses, but all ellipse centers are reversely discordant. Consequently the weighted mean age calculated from 207Pb/206Pb dates is younger (809 ± 15 Ma, 2σ, mean square of weighted deviates, MSWD = 0.36) than the weighted mean age calculated from 238U/206Pb dates (853 ± 19 Ma, 2σ, MSWD = 6.8). We interpret these data to record the protolith crystallization age. It is likely that these data record the same crystallization event as the data of sample AW 9–18–04 1D, which has a well-determined age of ca. 830 Ma (see following).

Sample AW 9–4-03 6A was collected from mylonitic augen gneiss ∼1 m structurally above a sheared top-to-the-southwest basal contact with Haimanta Group metasedimentary rocks (Fig. 6B). We acquired 18 spot analyses on 9 zircon crystals from this sample (Fig. 11; Supplemental File 5A [see footnote 5]). The results resemble typical detrital zircon results, with a range of dates spanning the Paleoproterozoic to the early Paleozoic and variable concordance and/or discordance. A Neoproterozoic cluster of 5 concordant and reversely discordant dates acquired from spots on 3 zircon crystals yields a weighted mean age calculated from 207Pb/206Pb dates of 842 ± 22 Ma (2σ, MSWD = 0.85). Two early Paleozoic dates were obtained: a reversely discordant datum with a 207Pb/206Pb date of 469.7 ± 10.3 Ma (1σ) and a concordant datum with a 207Pb/206Pb date of 514.0 ± 28.2 Ma (1σ). The early Paleozoic analyses are of spots from the outer rims of analyzed zircon crystals, and spots from the core of these two zircon crystals yield Mesoproterozoic and Paleoproterozoic 207Pb/206Pb ages. Th/U ratios of the two early Paleozoic analyses are very low (<0.009), characteristic of zircon crystallizing in equilibrium with an aqueous metamorphic fluid (Ding et al., 2001). We interpret the dates older than ca. 840 Ma to represent inherited zircon, and we interpret the early Paleozoic rim ages to represent metamorphic overgrowth of the inherited zircon. The ca. 840 Ma dates may represent the time of granite crystallization, or inherited zircon if the granite crystallized later. The lower limit on the crystallization age is represented by the early Paleozoic age of metamorphism.

Two granitic samples yield early Paleozoic dates: AW 9–7-03 5A, collected from granite with 5 cm feldspars ∼5 km east of Mandi (Figs. 6B and 10F), and AW 9–3-04 7A, collected from an ∼1-m-thick mylonitic granite sill ∼20 km north of Kullu in the overturned limb of the Phojal anticline (Fig. 6A). We analyzed 14 spots on 11 zircon crystals from sample AW 9–7-03 5A (Fig. 11; Supplemental File 5A [see footnote 5]). Excepting one reversely discordant analysis, these data are concordant and yield a weighted mean age calculated from 238U/206Pb dates of 470.6 ± 5.3 Ma (2σ, MSWD = 1.9). Spot dates on 5 different zircon crystals from sample AW 9–3-04 7A have similar results: concordant data with a weighted mean age calculated from 238U/206Pb dates of 473 ± 26 Ma (2σ, MSWD = 0.73). For both samples, we interpret the weighted mean age as the granite crystallization age.

We analyzed four spots on different zircon crystals from sample AW 9–24–04 4C, which was collected from a tourmaline-bearing leucogranite that crosscuts bedding preserved in amphibolite-grade metasedimentary rocks of the Leo Pargil dome (Fig. 6C). The acquired dates are variable, discordant, and Precambrian–Tertiary (Fig. 11; Supplemental File 5A [see footnote 5]). The youngest (238U/206Pb) date is 50.4 ± 2.4 Ma (1σ). These data likely reflect the ages of inherited zircon prior to Tertiary leucogranite crystallization.

MCT Shear Zone

We analyzed 71 spots on 51 zircon crystals from 7 samples to characterize the age and affinities of deformed granitic rocks within the MCT shear zone. The southernmost sample, AW 9–18–04 1D, was collected from a 3-m-thick layer of mylonitic augen gneiss with 5 cm feldspars in the MCT zone on the western flank of the Uttarkashi half-window (Fig. 6B). Spot analyses on 7 different zircon crystals yield a 207Pb/206Pb weighted mean age of 831 ± 19 Ma (2σ, MSWD = 0.69), which we interpret as the protolith crystallization age.

Analyses of the remaining six samples are consistent with protolith crystallization ca. 1.85 Ga. Sample AW 9–24–03 5 was collected from a mylonitic augen gneiss along the northern margin of the Narkanda half-window (Fig. 6B). We analyzed 15 spots on 10 zircon crystals from this sample (Fig. 11; Supplemental File 5A [see footnote 5]), and obtained 207Pb/206Pb dates ranging from 1690 Ma to 1866 Ma. Older dates are concordant, and with decreasing age the analyses are increasingly discordant. We interpret the older ages to represent the crystallization age, which we estimate as 1847 ± 10 (2σ, MSWD = 1.2). This estimate is a weighted mean age of 207Pb/206Pb dates that excludes the dates younger than 1800 Ma (8 of 15 ages were used). The younger discordant analyses likely reflect the effect of subsequent thermal disturbance or disturbances of ca. 1.85 Ga zircon. Sample AW 9–27–03 1B was collected from mylonitic augen gneiss that is ∼12 km north of Narkanda and ∼1 m structurally below a concordant contact with Haimanta garnet mica schist (Figs. 6B and 7C). We analyzed 13 spots on 9 zircon crystals from this sample (Fig. 11; Supplemental File 5A [see footnote 5]). Excepting one concordant analysis with a ca. 2.2 Ga 207Pb/206Pb date, the remaining discordant, concordant, and reversely discordant analyses cluster near ca. 1.85 Ga on the concordia plot and describe a weak chord extending from this age to a lower intercept in the Phanerozoic. We used 11 of 13 207Pb/206Pb dates (excluding the ca. 2.2 Ga age and the youngest, discordant analysis) to calculate a weighted mean age of 1835 ± 14 Ma (2σ, MSWD = 2.8). We interpret the ca. 2.2 Ga date to represent inherited zircon, the 1835 Ma age as the protolith crystallization age, and the chord trend to reflect thermal disturbance or disturbances.

Three of the analyzed samples are from the southwestern flank of the Kullu window of the MCT. Sample AW 9–29–03 6A was collected from a strongly mylonitic augen gneiss and/or ultramylonite ∼5 km west of Larji, and samples AW 9–4-03 3 and AW 9–26–03 1B were collected from mylonitic augen gneiss ∼15 km southwest and ∼20 km west of Rampur, respectively (Fig. 6B). We analyzed 9 spots on 6 zircon crystals from AW 9–29–03 6A, 7 spots on 5 zircon crystals from AW 9–4-03 3, and 15 spots on 9 zircon crystals from AW 9–26–03 1B (Fig. 11; Supplemental File 5A [see footnote 5]). Data from AW 9–29–03 6A and AW 9–4-03 3 plot as concordant clusters of ca. 1.85 Ga, with 207Pb/206Pb weighted mean ages of 1843 ± 10 (2σ, MSWD = 0.47) and 1842.4 ± 9.3 (2σ, MSWD = 0.66), respectively. We interpret these dates as protolith crystallization ages. Analyses of AW 9–26–03 1B zircon crystals have 207Pb/206Pb dates ranging from 1550 to 1865 Ma. The older dates are concordant, and increasingly younger 207Pb/206Pb dates are increasingly discordant. As with sample AW 9–24–03 5, we interpret the older ages to represent the crystallization age, which we estimate as 1840 ± 14 (2σ, MSWD = 3.1). This estimate is a weighted mean age of 207Pb/206Pb dates that excludes dates younger than 1800 Ma (10 of 15 ages were used). The younger discordant analyses likely reflect subsequent thermal disturbance or disturbances.

Sample AW 9–17–05 2 was collected from mylonitic augen gneiss along the northeastern flank of the Kullu window of the MCT, ∼10 km north-northeast of Wangtu (Fig. 6C). As mapped by Vannay and Grasemann (1998), the MCT zone here contains Wangtu gneiss at the base below a <100-m-thick discontinuous layer of recrystallized orthoquartzite commonly interpreted as Berinag quartzite, which in turn is succeeded by the Greater Himalayan Crystalline complex. Sample AW 9–17–05 2 likely represents discontinuous augen gneiss at the base of the Greater Himalayan Crystalline complex. However, we were unable to confirm the orthoquartzite exposure here and therefore allow that the sample could represent the top of the Wangtu gneiss. We analyzed 5 spots on different zircon crystals from this sample (Fig. 11; Supplemental File 5A [see footnote 5]); 4 analyses yield discordant to concordant data that plots near ca. 1.85 Ga on concordia; the 207Pb/206Pb dates of these analyses yield a weighted mean age of 1863 ± 17 (2σ, MSWD = 2.6). The fifth analysis is strongly discordant. A concordia line determined from all 5 analyses yields a 1897 ± 27 Ma upper intercept and a 470 ± 120 Ma lower intercept (2σ, MSWD = 0.49). We interpret the Paleoproterozoic dates to reflect ca. 1.85 Ga protolith crystallization, and the lower intercept to record younger thermal disturbance. The ca. 470 Ma date of the lower intercept matches the crystallization age of MCT hanging-wall granitic rocks (see following), so it is possible that this age records the same early Paleozoic tectonic episode.

The cumulative results from the MCT zone help define the age and spatial extent of the ca. 1.85 Ga Baragaon gneiss. Previous dating of the Baragaon gneiss yielded conflicting Rb-Sr dates of ca. 1.43 Ga and ca. 1.84 Ga (Bhanot et al., 1978; Miller et al., 2000); our results support the latter age. MCT zone granites to the south are distinct from this unit, as shown by the ca. 830 Ma age obtained for the southernmost sample AW 9–18–04 1D.

MCT Footwall

Samples AW 9–18–05 1B and AW 9–18–05 1C were collected from Wangtu gneiss and an undeformed pegmatitic felsic dike that crosscuts foliation of the Wangtu gneiss, respectively (Fig. 6C). Spot analyses on 3 different zircon crystals from sample AW 9–18–05 1B are strongly discordant, yielding a weak chord with intercepts at 123 ± 280 Ma and 1911 ± 390 Ma (2σ, MSWD = 6.6) (Fig. 11; Supplemental File 5A [see footnote 5]); 4 spot analyses on different zircon crystals from sample AW 9–18–05 1C yield Late Miocene 238U/206Pb ages, with a weighted mean age 7.4 ± 1.2 Ma (2σ, MSWD = 5.1) (Fig. 11; Supplemental File 5A [see footnote 5]). The AW 9–18–05 1B upper intercept age is consistent with previous dates of ca. 1.85 Ga for the crystallization of the granitic protolith of the Wangtu gneiss (Table 2), and the lower intercept age may record the Late Miocene metamorphism of the Munsiari Group. We interpret the Late Miocene dates from the undeformed dike as crystallization ages.

APPENDIX 4. U-Th-Pb GEOCHRONOLOGY OF DETRITAL ZIRCON: METHODS

Results of U-(Th)-Pb spot dating of detrital zircon were summarized in the discussion U-Th-Pb Zircon Geochronology: Detrital Zircon. Data for 16 samples were obtained using the laser ablation–multicollector–inductively coupled plasma–mass spectrometry (LA-MC-ICP-MS) facility at the University of Arizona, Tucson, and the UCLA ion microprobe. Procedures for ion microprobe zircon geochronology follow description in Appendix 3 (Methods); the following discussion pertains to the LA-MC-ICPMS analyses. These involve ablation of zircon with a New Wave–Lambda Physik DUV193 Excimer laser (operating at a wavelength of 193 nm) using a spot diameter of 25–50 μm. The ablated material is carried in helium into the plasma source of a GV Instruments Isoprobe, which is equipped with a flight tube of sufficient width that U, Th, and Pb isotopes are measured simultaneously. All measurements are made in static mode, using 1011 Ω Faraday detectors for 238U, 232Th, 208Pb, and 206Pb, a 1012 Ω Faraday collector for 207Pb, and an ion-counting channel for 204Pb. Ion yields are ∼1.0 mv/ppm. Each analysis consists of one 20 s integration on peaks with the laser off (for backgrounds), 20 1 s integrations with the laser firing, and a 30 s delay to purge the previous sample and prepare for the next analysis. The ablation pit is ∼15 μm in depth.

For each analysis, the errors in determining 238U/206Pb and 206Pb/204Pb ratios result in a measurement error of ∼1%–2% (2σ) in the 238U/206Pb age. The errors in measurement of 207Pb/206Pb and 206Pb/204Pb also result in ∼1%–2% (2σ) uncertainty in age for grains that are older than 1.0 Ga, but are substantially larger for younger grains due to low intensity of the 207Pb signal. For most analyses, the crossover in precision of 238U/206Pb and 207Pb/206Pb ages occurs at 0.8–1.0 Ga.

Common Pb correction is accomplished by using the measured 204Pb and assuming an initial Pb composition from Stacey and Kramers (1975) (with uncertainties of 1.0 for 206Pb/204Pb and 0.3 for 207Pb/204Pb). Our measurement of 204Pb is unaffected by the presence of 204Hg because backgrounds are measured on peaks (thereby subtracting any background 204Hg and 204Pb), and because very little Hg is present in the argon gas.

Interelement fractionation of Pb/U is generally ∼20%, whereas fractionation of Pb isotopes is generally ∼2%. In-run analysis of fragments of a large zircon crystal (generally every fifth measurement) with known age of 564 ± 4 Ma (2σ) is used to correct for this fractionation. The uncertainty resulting from the calibration correction is generally 1%–2% (2σ) for both 238U/206Pb and 207Pb/206Pb ages.

The LA-MC-ICPMS analytical data are tabulated in Supplemental File 5C (see footnote 5), and interpreted ages are shown on relative and cumulative age-probability diagrams (Figs. 12 and 13) (from Ludwig, 2003). Uncertainties shown in the table are at the 1σ level, and include only measurement errors. Interpreted ages for both LA-MC-ICPMS and ion microprobe analyses are based on 238U/206Pb for grains younger than 1000 Ma and on 207Pb/206Pb for grains older than 1000 Ma (with few exceptions noted in the Fig. 12 caption). This division at 1000 Ma results from the increasing uncertainty of 238U/206Pb ages and the decreasing uncertainty of 207Pb/206Pb ages as a function of age. Analyses that are >30% discordant (by comparison of 238U/206Pb and 207Pb/206Pb ages) or >5% reverse discordant are not considered. The relative age-probability diagrams show each age and its uncertainty (for measurement error only) as a normal distribution, and sum all ages from a sample into a single curve.

We thank editors Dennis Henry and Michael Williams and two anonymous reviewers for their constructive feedback and guidance. Stephanie Briggs, George Jarzebinski, Frank Kyte, Oscar Lovera, Gilles Peltzer, Axel Schmitt, and Victor Valencia provided technical advice and fruitful discussions. Field support from C.S. Dubey, Ashok Kumar, Subhash Kumar, Sandeep Saod and his team, Braj Sharma, Manoj Tiwari, and especially Paul Burgess, Prashant Dubey, and Courtney Gregory, is gratefully acknowledged. Discussions with Om Bhargava, Peter Bird, Jen Chambers, Joe DiPietro, Darrell Henry, Brian Horton, Nigel Hughes, David Jacobs, Arvind Jain, Remy Leger, Ryan McKenzie, Bill Moore, Alex Robinson, Yongwei Sheng, Justin Simon, Birendra Singh, Sandeep Singh, and Mike Taylor all led to improvements in this work. This research was a grant from by the Tectonics Program, U.S. National Science Foundation, and a start-up fund from Louisiana State University.

1Supplemental File 1. PDF file of sketch cross section along profile B-B′; see text Figure 3. The discontinuous graphitic quartzite marker lithology occurs at two levels in the Chandrabhaga River Valley. These occurrences are interpreted as two distinct stratigraphic horizons. Alternatively, these may reflect unrecognized kilometer-scale tight to isoclinal folds. Primary sources for this section are Powers et al. (1998) (sub-Himalayan thrust zone); Srikantia and Sharma (1976) (sedimentary Lesser Himalayan Sequence units); Frank et al. (1995) [Haimanta Group in the Main Central thrust (MCT) hanging wall]; Thakur (1998), Dèzes (1999), Yin (2006), and Webb et al. (2007) (STD, South Tibet detachment). If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00627.S2 or the full-text article on www.gsapubs.org to view Supplemental File 1.
2Supplemental File 2. PDF file of 87Sr/86Sr (500 Ma) vs. εNd (500 Ma) plots for the western and central Himalaya. In the Nepal Himalaya, Nd and Sr isotopic compositions and detrital zircon age distributions are different in the Lesser Himalayan Sequence (LHS), Greater Himalayan Crystalline complex (GHC), and Tethyan Himalayan Sequence (THS). It has been proposed that such data can be used to identify the structural setting of Himalayan strata (e.g., Parrish and Hodges, 1996). However, rocks of the same age in the Lesser Himalayan Sequence, Greater Himalayan Crystalline complex, and Tethyan Himalayan Sequence have the same isotopic compositions and detrital zircon patterns (e.g., Myrow et al., 2003; Richards et al., 2005). This suggests that distinctions from isotopic and detrital zircon signatures in Himalayan rocks cannot be used directly to infer structural setting. Such distinctions are nonetheless valuable for constraining age ranges of Himalayan strata and thus for making stratigraphic comparisons. Neoproterozoic and younger rocks are generally distinguishable from Mesoproterozoic and older rocks in 87Sr/86Sr (500 Ma) vs. εNd (500 Ma) space. (A) 87Sr/86Sr (500 Ma) vs. εNd (500 Ma) plot for western and central Himalaya rocks. Data for this plot are divided into three plots to ease tracking of data for individual units. (B) 87Sr/86Sr (500 Ma) vs. εNd (500 Ma) plot for Himachal Himalaya rocks. Data from Bhat and Le Fort (1992, 1993), Miller et al. (2000, 2001), and Richards et al. (2005). (C) 87Sr/86Sr (500 Ma) vs. εNd (500 Ma) plot for Kumaun and Nepal Himalaya rocks (Kumaun Himalaya—northwest India Himalaya to the east of Himachal Pradesh). Data from Ahmad et al. (2000), Deniel et al. (1987), France-Lanord et al. (1993), Inger and Harris (1993), and Prince (1999). (D) 87Sr/86Sr (500 Ma) vs. εNd (500 Ma) plot for western Himalayan (Nanga Parbat) syntaxis rocks. Data from Argles et al. (2003), Foster et al. (2000, 2002), Gazis et al. (1998), George et al. (1993), and Whittington et al. (1999). If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00627.S3 or the full-text article on www.gsapubs.org to view Supplemental File 2.
3Supplemental File 3. PDF file of geochemistry of granitoid samples. (A) QAP (quartz–alkali feldspar–plagioclase feldspar) diagram for granitoid samples generated using whole-rock major element data. CIPW norms were calculated for each sample and projected on the diagram. All samples plot within (or on the edge of) the granite field, with the exception of one Baragaon sample that plots within the granodiorite field. (B) AFM diagram for granitic samples; A = K2O + Na2O; F = Fe2O3 + FeO; M = MgO. Curve for calc-alkaline and thoeliitic division is from Kuno (1968). (C) Geochemical discrimination diagrams for granites after Pearce et al. (1984). VAG—volcanic arc granite; syn-COLG—syncollisional granite; ORG—ocean ridge granite; WPG—within-plate granite. (D) Table of geochemical data for granitoid samples. If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00627.S4 or the full-text article on www.gsapubs.org to view Supplemental File 3.
4Supplemental File 4. PDF file. (A) Mn zoning maps of garnet crystals from thermobarometry analyses. (B) Electron microprobe analyses for thermobarometry results. If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00627.S5 or the full-text article on www.gsapubs.org to view Supplemental File 4.
5Supplemental File 5. PDF file. (A) Ion microprobe U-(Th)-Pb zircon data table. (B) U-Pb concordia plots showing results of single spot analyses on zircon from metasedimentary and/or sedimentary rocks. Most data were acquired via laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) at the University of Arizona Laserchron Center; analyses labeled “ion microprobe” were acquired via the University of California Los Angeles ion microprobe. (C) LA-ICP U-(Th)-Pb zircon data. (D) Kolmogorov-Smirnov (K-S) statistic applied to the probability density functions of detrital zircon samples. If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00627.S6 or the full-text article on www.gsapubs.org to view Supplemental File 5.
6Supplemental File 6. PDF file of mica and K-feldspar 40Ar/39Ar thermochronology results for the Main Central thrust hanging wall to the southwest of the Kullu window. (A) Mica age spectra. (B) K-feldspar age spectra. (C) Thermal history results, from modeling of only that small (young) portion of the K-feldspar age spectra that may not be contaminated by excess Ar. (D) Mica data table. (E) K-feldspar data table. If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00627.S7 or the full-text article on www.gsapubs.org to view Supplemental File 6.