Abstract

The South Tibetan fault system, a family of primarily extensional faults that separates the metamorphic core of the Himalaya (expressed as the Greater Himalayan sequence) from overlying, predominantly unmetamorphosed Tibetan sedimentary sequence units, has been mapped for over 2000 km coincident with the Himalayan range crest. In most areas, the immediate hanging wall of the South Tibetan fault system sole detachment consists of predominantly carbonate rocks of Lower Paleozoic age. However, in the Bhutan sector of the eastern Himalaya (∼89°E–92°E), the hanging wall of the sole structure is instead frequently mapped at the base of a metamorphosed, predominantly siliciclastic succession (the Chekha Formation), and the base of the overlying predominantly carbonate rocks (Pele La and Tang Chu Groups) is mapped as a less significant splay of the South Tibetan fault system. Unfortunately, poor exposures throughout central Bhutan make mapping and structural interpretation of these important contacts difficult, resulting in many disparities among geologic maps made by different research groups. The South Tibetan fault system in other parts of the Himalaya accommodates a significant metamorphic discontinuity that should also be apparent in Bhutan. Therefore, as an alternative approach, we have used the Raman spectroscopy on carbonaceous material (RSCM) thermometer to evaluate the evidence for a metamorphic discontinuity across both putative South Tibetan fault system structures.

RSCM thermometric data from 17 samples across three purported South Tibetan fault system klippen in central Bhutan (the Dang Chu, Ura, and Zhemgang klippen) suggest that the contact between the Chekha Formation and the underlying Greater Himalayan sequence is not a fault with large postmetamorphic displacement. We find no resolvable change in peak metamorphic temperature across the contact (∼560 °C in both the Chekha and Greater Himalayan sequence), but we see a 130–140 °C drop in paleotemperature across the higher contact between the Chekha Formation and overlying Pele La and Tang Chu Groups. This change coincides with a major change in structural style, from high-strain, leucogranite-bearing rocks below to large-scale recumbently folded marbles above. Together, the changes in deformational character and metamorphic grade suggest that the principal South Tibetan fault system detachment in Bhutan is the structural boundary between the Chekha Formation and the predominantly carbonate rocks above. The presence of a South Tibetan fault system detachment ∼80 km south of the main South Tibetan fault system fault trace at the crest of the Himalaya, with no match between correlative footwall and hanging-wall units along the direction of fault motion, implies large displacements on the South Tibetan fault system in the eastern Himalaya.

INTRODUCTION

Over the past few decades, the eastern Himalayan Kingdom of Bhutan has become increasingly accessible to foreign visitors, resulting in a flurry of geologic research that has added critical new information to our understanding of the Himalayan-Tibetan orogenic system (Carosi et al., 2006; Chakungal et al., 2010; Chambers et al., 2011; Cooper et al., 2012; Corrie et al., 2012; Daniel et al., 2003; Davidson et al., 1997; Edwards et al., 1999; Edwards and Harrison, 1997; Gansser, 1983; Grujic et al., 2002, 2006, 2011; Hughes et al., 2011; Kellett et al., 2009, 2010; Kellett and Grujic, 2012; Long and McQuarrie, 2010; Long et al., 2011a, 2011b, 2011c; Stüwe and Foster, 2001; Swapp and Hollister, 1991; Tobgay et al., 2012). However, research progress has been hindered by the dense vegetation and shortage of roads (and, in turn, road-cut outcrops) throughout most of the country. Only in the areas near the Tibetan border are outcrops sufficient to tightly constrain geologic mapping. As a result, there are still many disparities among geologic maps of Bhutan made by different research groups. One outstanding issue regards the position, character, and displacement of the principal basal (or “sole”) detachment of the South Tibetan fault system, a family of primarily extensional faults that crops out for over 2000 km along the length of the Himalayan range crest (Burchfiel et al., 1992; Burg and Chen, 1984; Hodges et al., 1992; Pognante and Benna, 1993; Searle et al., 1997; Searle, 1999).

Basal low-angle detachments of this system typically mark a metamorphic discontinuity between high-grade metamorphic gneisses and anatectites of the Himalayan core below and lower-grade or unmetamorphosed strata above (Burchfiel et al., 1992). Recent mapping in the central Bhutan Himalaya (Figs. 1 and 2) suggests that there may be multiple detachments of the South Tibetan fault system there (e.g., Carosi et al., 2006; Edwards and Harrison, 1997; Grujic et al., 2002, 2006, 2011; Kellett et al., 2009, 2010; Kellett and Grujic, 2012; Long and McQuarrie, 2010; Long et al., 2011a, 2011c), an observation similar to that made in several other parts of the orogen, where structurally higher detachments are typically marked by tectonite fabrics but not by significant metamorphic discontinuities (Burchfiel et al., 1992; Hodges et al., 1994, 1996; Searle and Godin, 2003; Searle, 1999).

The role of the extensional South Tibetan fault system in the development of the Himalayan orogen is a matter of active debate, largely because constraining the magnitude of its displacement has proven to be difficult. Generally exposed along the Himalayan range crest, downdip exposures of South Tibetan fault system detachments that allow direct measurements of minimum displacement are rare. In a recent study in NW Bhutan, Cooper et al. (2012) traced the South Tibetan fault system from the range crest south for ∼65 km, suggesting large displacements on the system. The presence of South Tibetan fault system detachments even farther south in central Bhutan suggests that minimum displacements on the South Tibetan fault system could be even larger. Grujic et al. (2011), for example, mapped the South Tibetan fault system ∼100 km south of the range crest (Fig. 2B). Alternatively, mapping by Long and McQuarrie (2010) implies that the breakaway zone for the South Tibetan fault system is present in southern Bhutan. The structural offset between this breakaway and hanging-wall units to the north is at most 20 km, suggesting limited slip on the South Tibetan fault system.

The differences in mapping and interpretation of the South Tibetan fault system in central Bhutan suggest that further work is needed in order to constrain the position and thus the significance of the system there. Due to the poor exposures in Bhutan, which have hindered mapping of South Tibetan fault system structures thus far, we used the widely applicable Raman spectroscopy on carbonaceous material (RSCM) thermometer to evaluate the evidence for metamorphic discontinuities across the putative South Tibetan fault system structures mapped by previous workers.

STRUCTURAL SETTING

Himalayan-Tibetan Orogenic System

The Himalayan-Tibetan orogenic system is one of our planet’s most spectacular signatures of continent-continent collision. Thought to have initiated in the early Eocene with the closure of the Neotethys Ocean (de Sigoyer et al., 2000; Guillot et al., 2008; Leech et al., 2005; Rowley, 1996), collision of the Indian and Eurasian plates created both the highest mountain range and the most expansive area of regional uplift on Earth: the Tibetan Plateau. The Himalayan sector of the orogenic system consists of four broad lithotectonic belts of contrasting metamorphic grade separated by a series of north-dipping faults (Gansser, 1964; Heim and Gansser, 1939; Hodges, 2000; Le Fort, 1975). From south to north, these are the Subhimalayan zone, Lesser Himalayan zone, Greater Himalayan zone, and Tibetan zone. Each of these zones constitutes a distinctive package of rocks, known as the Eocene to Lower Miocene Rawalpindi and Lower Miocene to Pleistocene Siwalik Groups of the Subhimalayan zone, the Proterozoic to Middle(?) Miocene Lesser Himalayan sequence, the Neoproterozoic to Ordovician Greater Himalayan sequence, and the Paleozoic to Eocene Tibetan sedimentary sequence (Acharyya and Sastry, 1979; Brasier and Singh, 1987; Brookfield, 1993; Burbank et al., 1997; Critelli and Garzanti, 1994; DeCelles et al., 1998; Gaetani and Garzanti, 1991; Gansser, 1983; Hodges, 2000; Najman et al., 1993, 1997; Parrish and Hodges, 1996; Singh et al., 1999; Stöcklin, 1980; Valdiya, 1980). Rocks of the Subhimalaya, Lesser Himalayan sequence, and Greater Himalayan sequence are separated by orogen-parallel contractional structures of the Main Boundary thrust system and the Main Central thrust system. In the Bhutanese Himalaya, an out-of-sequence thrust fault (the Kakhtang thrust) mapped by Gansser (1983) has been interpreted to have roughly doubled the thickness of the Greater Himalayan sequence (Davidson et al., 1997; Grujic et al., 2002) (Fig. 1), although the location, age, and displacement across this structure are poorly constrained. In contrast to its southern and lower boundary, the top of the Greater Himalayan sequence is bounded by extensional faults and shear zones of the South Tibetan fault system. The opposing-sense South Tibetan fault system and Main Central thrust system are both thought to have been active during the Miocene (Hodges et al., 1992, 1996; Hubbard and Harrison, 1989; Searle and Rex, 1989), implying that the South Tibetan fault system played an important role in exhuming the Greater Himalayan sequence metamorphic core.

The structurally highest and northernmost zone is represented by the Tibetan sedimentary sequence, which generally crops out north of the Himalayan range crest. The Tibetan sedimentary sequence consists of low-grade to unmetamorphosed sediments deposited on the northern passive continental margin of India (Gaetani and Garzanti, 1991). Although it is generally accepted that the Greater Himalayan sequence and the Tibetan sedimentary sequence are separated by the South Tibetan fault system throughout most of the Himalaya, the relationships are more controversial in regions south of the range crest, where the Tibetan sedimentary sequence occurs in a series of low-elevation outliers above Greater Himalayan sequence lithologies. In some cases, these contacts—typically poorly exposed—have been interpreted as detachments (presumably strands of the South Tibetan fault system), whereas they have been interpreted as depositional in others (Gehrels et al., 2003; Grujic et al., 2002; Robinson et al., 2006; Stöcklin, 1980). In Bhutan, both depositional and tectonic relationships have been reported for these outliers (e.g., Long and McQuarrie, 2010).

South Tibetan Fault System

The South Tibetan fault system was first recognized in central Nepal (Caby et al., 1983) and later described in southern Tibet (Burchfiel et al., 1992; Burg and Chen, 1984) and northwest India (Herren, 1987; Searle, 1986; Valdiya, 1989). Although it consists of a variety of fault types including steeply dipping transfer faults (Wu et al., 1998) and low-angle, oblique faults with a significant component of strike-slip motion (Pêcher, 1991), most descriptions focus on the basal structure of the system, a low-angle, north-dipping fault and associated ductile shear zone commonly referred to as the South Tibetan detachment.

The presence of the South Tibetan fault system within a zone of continental collision has led to considerable debate regarding its initiation and role in the construction of the orogen (e.g., Hodges, 2000; Law et al., 2006, and references therein). Two principal models have been put forward to explain its existence. In the first, the South Tibetan fault system forms a collection of passive roof faults over an evolving contractional orogenic wedge (e.g., Robinson et al., 2006; Yin, 2006; Yin et al., 1994). In this case, the South Tibetan fault system has only a minor role in extrusion, with minimal displacement across the structure and little excision of material in the footwall. In the second, the Main Central thrust system and South Tibetan fault system are kinematically linked structures that collectively sustained Miocene southward extrusion of the metamorphic core of the Himalaya (Beaumont et al., 2001; Godin et al., 2006, and references therein; Grujic et al., 2002; Hodges et al., 2001; Nelson et al., 1996). This concept, often referred to as the “channel flow” model, implies that the South Tibetan fault system would have accommodated many kilometers of displacement and is responsible for excision of several kilometers of structural section within the Greater Himalayan sequence (Searle et al., 2006).

However, because of the geographic coincidence of many of the basal detachments with the Himalayan range crest and the relatively subdued relief north of the crest, most of these structures cannot be traced very far downdip, and their net displacements are derived from indirect geothermobarometric measurements (15–200 km: Cottle et al., 2007, 2011; Dézes et al., 1999; Searle et al., 2002, 2003; Walker et al., 1999) and studies of fault-related telescoped isograds (25–170 km: Herren, 1987; Law et al., 2011). Exceptions occur in the Mount Everest region of Nepal and the Mount Jomolhari region of NW Bhutan, where components of the South Tibetan fault system can be traced parallel to their slip vectors, with no match between correlative footwall and hanging-wall units for ≥34 km in Nepal (Carosi et al., 1998; Hodges et al., 1992) and ≥65 km in Bhutan (Cooper et al., 2012), implying minimum displacements comparable to ∼75–140 km minimum estimates for broadly contemporaneous S-directed slip on the Main Central thrust system in the eastern Himalaya (Yin, 2006; Yin et al., 2010).

The South Tibetan fault system footwall consists of high-grade (upper amphibolite facies) paragneisses and orthogneisses of the Greater Himalayan sequence with abundant leucogranite sills, dikes, and plutons. Near the top of the footwall, these rocks are strongly deformed, and most exposures contain well-developed S-C mylonites (Lister and Snoke, 1984) indicative of hanging-wall down-to-the north (normal-sense) shearing with varying degrees of oblique slip. In most mapped transects, the basal detachment carries unmetamorphosed Lower Paleozoic sedimentary rocks of the Tibetan sedimentary sequence in its immediate hanging wall (e.g., Herren, 1987 [India]; Hodges et al., 1993 [southern Tibet]). In two areas—the Annapurna Range of central Nepal (e.g., Brown and Nazarchuk, 1993) and the Everest region of eastern Nepal (e.g., Searle, 1999)—hanging-wall Tibetan sedimentary sequence units also experienced greenschist- to lower- or middle-amphibolite-facies metamorphism, but the South Tibetan fault system still marks a significant discontinuity in metamorphic pressure and temperature.

Studies of the South Tibetan fault system in Tibet (Burchfiel et al., 1992; Hodges et al., 1994) and Nepal (Hodges et al., 1996; Searle and Godin, 2003; Searle, 1999) show that wherever the basal detachment carries metamorphosed Tibetan sedimentary sequence rocks in its hanging wall, there is at least one major detachment of the South Tibetan fault system at a structurally higher level. These structures typically place stratigraphically younger Tibetan sedimentary sequence lithologies on older lithologies, or lower-grade (or unmetamorphosed) rocks on higher-grade rocks. The character of deformation along and above the South Tibetan fault system detachments depends on the hanging-wall lithology. Observations of the South Tibetan fault system in various localities in Nepal suggest that, when the basal detachment carries lower-amphibolite- or greenschist-facies rocks in its hanging wall, there is typically a relatively wide shear zone above and below the detachment but often also a relatively sharp brittle-ductile shear zone at the contact itself (e.g., Deorali detachment—Hodges et al., 1996; Lhotse detachment—Searle, 1999; Annapurna detachment—Vannay and Hodges, 1996). Both ductile and brittle fabrics are transposed into parallelism with the detachment, indicating syndetachment development. When the hanging wall is unmetamorphosed or weakly metamorphosed, there is a well-developed, relatively wide shear zone beneath the detachment and a usually pronounced (but sometimes thin) breccia zone at the contact. The breccia zone is oriented subparallel to the shear fabric in the footwall but the hanging wall–footwall contact is marked by an obvious cutoff of hanging-wall strata, sometimes at a very high angle. Leucogranites cut the basal South Tibetan fault system detachment in several well-studied areas (e.g., Hodges et al., 1996), but examples of them cutting the upper detachment (e.g., Guillot et al., 1994; Hodges et al., 1998) are extremely rare.

South Tibetan Fault System in Bhutan and Adjacent Areas of Tibet

In the first regional study of the South Tibetan fault system, Burchfiel et al. (1992) mapped two transects across the Greater Himalayan sequence–Tibetan sedimentary sequence boundary just north of Bhutan at Wagye La and Lhozag-La Kang (Fig. 1). In the Wagye La area, the contact is not exposed, but the topography and outcrop pattern indicate that it must dip shallowly northward, subparallel to well-developed S-C mylonitic planar fabrics in the footwall. The contact was interpreted by Burchfiel et al. (1992) as a segment of the basal detachment of the South Tibetan fault system, with classic Greater Himalayan sequence footwall units including amphibolite-facies orthogneisses and psammitic and pelitic schists, all intruded by leucogranite sills and dikes. The hanging-wall units (low-grade Ordovician marbles and phyllites) are themselves cut by a well-exposed upper detachment that carries unmetamorphosed Carboniferous–Permian limestones in its hanging wall.

At Lhozag La Kang, the principal South Tibetan fault system detachment has been deformed into ∼10-km-wavelength, upright folds and subsequently cut by steeply N-dipping, E-striking normal faults with relatively minor displacement (Burchfiel et al., 1992; Edwards et al., 1999). This detachment cuts and thus postdates the ca. 12.5 Ma Khula Kangri leucogranite pluton (Edwards and Harrison, 1997; Fig. 1). At Gonto La (Edwards et al., 1996; Fig. 1), the detachment also cuts an older, structurally lower South Tibetan fault system detachment that is intruded by the Khula Kangri pluton.

In the central latitudes of Bhutan, Gansser (1983) mapped synformal erosional remnants of Tibetan sedimentary sequence lithologies above Greater Himalayan sequence units far south of the main outcrop trace of the South Tibetan fault system along the Bhutan-Tibet border (Fig. 1). The structurally and stratigraphically highest units in the erosional remnants are Precambrian–Devonian(?), locally fossiliferous, low-grade calc-schists, calcarenites, and limestones of the Pele La Group and the Tang Chu Group (Hughes et al., 2011; Long and McQuarrie, 2010; Tangri and Pande, 1995), similar to basal Tibetan sedimentary sequence lithologies widespread in the Himalaya and southern Tibet (Dézes et al., 1999; Gansser, 1964; Hodges et al., 1996; Le Fort, 1975; Searle and Godin, 2003; Searle et al., 2003). The rocks mapped as Greater Himalayan sequence throughout Bhutan include Proterozoic–Ordovician(?) granitic and migmatitic orthogneiss, migmatitic metasedimentary rocks, schist, paragneiss, quartzite, and discrete marble bands, pervasively intruded by Miocene leucogranites (Bhargava, 1995; Davidson et al., 1997; Gansser, 1983; Grujic et al., 2002; Hollister and Grujic, 2006; Long and McQuarrie, 2010; Long et al., 2011c; Swapp and Hollister, 1991). There is little controversy regarding correlations of these rocks with Greater Himalayan sequence units elsewhere in the Himalaya. Of less certain affinity is a unit interposed between classically Greater Himalayan sequence and Tibetan sedimentary sequence units referred to as the Chekha Formation (Gansser, 1983; Jangpangi, 1974; Nautiyal et al., 1964; Tangri and Pande, 1995).

The Chekha Formation in Bhutan is composed of nonfossiliferous greenschist- to amphibolite-facies metapelite, paragneiss, augen gneiss, quartzite, and calc-silicate intruded by leucogranite sills and dikes (Cooper et al., 2012; Gansser, 1983; Grujic et al., 2002; Kellett et al., 2009; Long and McQuarrie, 2010; McQuarrie et al., 2008). Kellett et al. (2009), Kellett et al. (2010), and Kellett and Grujic (2012) correlated the Chekha regionally with Tibetan sedimentary sequence units in other parts of the Himalaya, notably the Everest Series and North Col Formation of eastern Nepal (Searle et al., 2003), the Annapurna Yellow Formation of central Nepal (Gleeson and Godin, 2006), and the Haimanta Group of NW India (Chambers et al., 2009). If this is the case, then age constraints for these other units suggest a probable Cambrian age for the Chekha Formation (Burchfiel et al., 1992; Carosi et al., 1999; Colchen et al., 1986; Frank et al., 1973; Lombardo et al., 1993; Mu et al., 1973; Myrow et al., 2009; Wang and Zhen, 1975). Other workers have used the position of the Chekha Formation at the base of the Tibetan sedimentary sequence and its lack of fossils to infer a Precambrian age (Gansser, 1983; Tangri and Pande, 1995) for the unit. Detrital zircon U-Pb age spectra from some Chekha samples are similar to those obtained for Tibetan sedimentary sequence samples, whereas others are more like those obtained for Greater Himalayan sequence samples collected in Bhutan (Gehrels et al., 2011; Hughes et al., 2011; Long and McQuarrie, 2010; McQuarrie et al., 2008). As Hughes et al. (2011) noted, additional stratigraphic constraints on the depositional age of the Chekha Formation are needed.

The contacts between the Chekha Formation and units above and below have been variably interpreted. Gansser (1983) described the Greater Himalayan sequence–Chekha contact as conformable and noted that both the uppermost Greater Himalayan sequence pelitic units and the lowermost Chekha Formation schists contain distinct biotite porphyroblasts (cross-biotites) lying perpendicular to foliation but parallel to lineation. Grujic et al. (2002), on the other hand, found evidence for a diffuse top-to-the-north shear zone (width unconstrained) at the base of the Chekha Formation across Bhutan, and re-interpreted the Dang Chu (or Tang Chu), Ura, Zhemgang (or Black Mountain), and Sakteng (or Radi) exposures as klippen soled by the South Tibetan fault system basal detachment. Their main lines of evidence were: (1) top-to-the-north shear sense indicators at the top of the Greater Himalayan sequence and the base of the Chekha Formation; (2) the presence of migmatite and sillimanite in the Greater Himalayan sequence, which are absent from the Chekha Formation above; and (3) an up-section decrease in metamorphic grade within the Chekha Formation. However, despite these first-order observations across Bhutan, no discrete (meter-scale) brittle-ductile shear zone at the contact, upward increase in strain toward the contact, or definitive structural discordance between footwall and hanging-wall units has yet been described at this structural level (e.g., Carosi et al., 2006; Grujic et al., 2002; Kellett et al., 2009), contrary to observations of classic South Tibetan fault system detachments in other parts of the Himalaya (e.g., Burchfiel et al., 1992; Hodges et al., 1992, 1996; Pognante and Benna, 1993; Searle et al., 1997; Searle, 1999; Vannay and Hodges, 1996).

The ambiguity in the Greater Himalayan sequence–Chekha contact is exemplified by the contrasting mapping of two different groups working in Bhutan, which shows little agreement on either the position or the nature of the contact. Long and McQuarrie (2010), for example, largely followed the original mapping by Gansser (1983), and agreed with Grujic et al. (2002) that the base of the Chekha Formation in the Dang Chu, Ura, and Sakteng klippen was a top-to-the-north shear zone of the South Tibetan fault system. However, in contrast to Grujic et al. (2002), they suggested that interfingering of Greater Himalayan sequence and Chekha units at the base of the Zhemgang klippe in southern Bhutan indicated a depositional contact there (Figs. 2A, 3, and 4). Grujic et al. (2011), on the other hand, mapped the Chekha to a far greater extent throughout central Bhutan, combining the Dang Chu and Zhemgang klippen into a single entity, and extending the Ura klippe northward, where it is cut by the Kakhtang thrust (KT; Fig. 2B). A direct comparison between the maps of the two groups (Fig. 2) suggests that the large areal extent of the Chekha mapped by Grujic et al. (2011) corresponds closely to the extent of the Greater Himalayan sequence metasedimentary unit mapped by Long and McQuarrie (Fig. 4). This again reinforces the ambiguity between the two lithologic units and suggests that new data are needed to understand the structural relationship between them.

The difference in mapping of the South Tibetan fault system in central Bhutan has implications for the magnitude of displacement on the South Tibetan fault system. The interpretation by Long and McQuarrie (2010), that the base of the Zhemgang klippe is a conformable contact between Greater Himalayan sequence and Tibetan sedimentary sequence units, but the base of the Ura klippe is a strand of the South Tibetan fault system, led them to argue that the breakaway zone for the South Tibetan fault system must lie in between the two (Figs. 2A and 3A). If this is the case, then it limits slip on the South Tibetan fault system to a maximum of only 20 km, and reduces the significance of extensional faulting as a major orogen-building process. In contrast, mapping of South Tibetan fault system detachments (including at the base of the Zhemgang klippe) by Grujic et al. (2011) as far as 100 km south of the main South Tibetan fault system trace along the Himalayan range crest, with no observed breakaway zone, means that South Tibetan fault system hanging-wall-on-footwall relationships can be traced in the direction of slip for ∼100 km, implying large displacements on this system.

Thermobarometric studies across the Greater Himalayan sequence–Chekha contact in Bhutan are limited. Studies by Davidson et al. (1997) and Daniel et al. (2003) found the Greater Himalayan sequence to have reached peak metamorphic temperatures of 600–750 °C and pressures of 8–10 kbar. In a detailed thermobarometric study across the Ura klippe based on silicate mineral compositions, Kellett et al. (2010) inferred a change in temperature across the Greater Himalayan sequence–Chekha contact but saw no discernible change in pressure. On closer inspection, their data show a similarly large spread in pressure-temperature (P-T) conditions for both the Chekha Formation (576–730 °C and 6.9–8.7 kbar) and the Greater Himalayan sequence (560–789 °C and 8.0–9.1 kbar), indicating no metamorphic discontinuity across this contact. Their results are similar in the Jomolhari region of NW Bhutan (referred to by the authors as the Lingshi klippe), where one sample from the base of the Chekha Formation gives P-T conditions of 721 °C and 8.7 kbar, while six samples from the Greater Himalayan sequence give a spread of 622–787 °C and 6.2–10.9 kbar. Thermobarometric data from a recent study across the Greater Himalayan sequence–Chekha contact at the base of the Zhemgang klippe by Corrie et al. (2012) support the interpretation of Long and McQuarrie (2010), i.e., that it is a conformable contact, noting a gradual change in peak temperature and pressure across the contact from ∼540–620 °C and 9 kbar in the Greater Himalayan sequence ∼2 km from the contact to 550 °C and 7.5 kbar throughout the Chekha Formation. In the only other RSCM study in Bhutan to date, Kellett and Grujic (2012) obtained peak RSCM temperatures from Chekha and Tibetan sedimentary sequence rocks of the Linghsi klippe that show little variation, with a consistently low peak temperature of ∼300 °C. By combining these RSCM data with the P-T data of Kellett et al. (2010), Kellett and Grujic (2012) inferred a gradual change in temperature across the Greater Himalayan sequence–Chekha contact, which they ascribed to a diffuse shear zone at this structural level. However, the substantial drop in temperature to ∼300 °C is ∼600 m above this contact, suggesting a more significant offset at a structurally higher position.

The nature of the contact between the Chekha Formation and the overlying indisputable Tibetan sedimentary sequence (Pele La Group and Tang Chu Group) is also unclear. The contact is exposed in the Mount Jomolhari region in NW Bhutan and in the Dang Chu klippe in central Bhutan. (Note that this is typically referred to as the Tang Chu klippe following Gansser [1983]. We have decided not to use this name because the Tang Chu river actually lies to the east near Ura. To avoid confusion we use Dang Chu klippe after the local Dang Chu river [Fig. 2].) Gansser (1983), Carosi et al. (2006), and Kellett and Grujic (2012) mapped the contact as conformable, but Edwards et al. (1996), Hollister and Grujic (2006), and Chambers et al. (2011) interpreted it as a detachment of the South Tibetan fault system. In NW Bhutan (Fig. 1), Cooper et al. (2012) mapped recumbently folded fossiliferous marbles of the Tibetan sedimentary sequence above amphibolite-facies metapelites, calc-silicates, and leucogranites of the Chekha Formation. The abrupt change in structural style across the contact between these two units together with the stark change in lithology and metamorphic grade led the authors to interpret this contact as a detachment of the South Tibetan fault system.

In the Dang Chu klippe, Gansser (1983) mapped two isolated exposures of the Tibetan sedimentary sequence lying above the Chekha Formation. In the more accessible northern exposure, the transition from Chekha Formation to Tibetan sedimentary sequence units of the Pele La Group and Tang Chu Group is marked by a dramatic change in structural style from foliated metapelites and quartzites to recumbently folded calc-silicates and marbles (Figs. 5A–5C). Just to the east of the southern Tibetan sedimentary sequence exposure mapped by Gansser (1983), Hughes et al. (2011) identified Cambrian brachiopod and trilobite fossils in siliciclastic and carbonate units of the Pele La Group. Although this location has been mapped by other researchers as part of the Greater Himalayan sequence (Grujic et al., 2002; Kellett et al., 2009, 2010; Long and McQuarrie, 2010; Long et al., 2011c; Fig. 2A) and the Chekha Formation (Grujic et al., 2011; Fig. 2B), we join Hughes et al. (2011) as interpreting these fossiliferous outcrops as part of the Tibetan sedimentary sequence and have extended the southern exposure of this unit in the Dang Chu klippe eastward to include this locality (Fig. 1).

Above the Chekha Formation in the center of the Zhemgang klippe (Fig. 5E), Long and McQuarrie (2010) mapped the Maneting Formation, a biotite-garnet bearing phyllitic unit (Fig. 5F) of the Pele La Group (Tangri and Pande, 1995). Based on an observed up-section transition from Chekha quartzite to Maneting phyllite and interfingering of the two lithologies, they interpreted the contact between the Chekha and Maneting Formations to be conformable (Figs. 2A and 5). Thermobarometric data from Corrie et al. (2012) agree with this interpretation, suggesting a steady decrease in peak P-T conditions across the Chekha-Maneting contact, with no evidence for a structural break.

Because structural studies alone do not seem sufficient to determine the nature of the Greater Himalayan sequence–Chekha and Chekha–Tibetan sedimentary sequence contacts in the central latitudes of Bhutan, we applied thermometric techniques to evaluate the evidence for a metamorphic discontinuity across them. Although conventional pelitic thermobarometers are easily applied to many Greater Himalayan sequence rocks, Chekha Formation and Tibetan sedimentary sequence rocks typically contain less suitable high-variance mineral assemblages. As a consequence, we focused our studies on the establishment of peak metamorphic temperatures through the more widely applicable Raman spectroscopy on carbonaceous material (RSCM) method. This relatively new technique (Aoya et al., 2010; Beyssac et al., 2002a, 2002b; Rahl et al., 2005) has become very popular in recent years and has been applied to rocks from several sectors of the Himalayan orogen (Beyssac et al., 2004; Bollinger et al., 2004; Célérier et al., 2009; Cottle et al., 2011; Kellett and Grujic, 2012). The popularity of the RSCM thermometer stems from its applicability to rocks of many bulk compositions, the fact that it is apparently independent of metamorphic pressure (unlike most of the commonly used metamorphic thermometers for amphibolite-facies metamorphic rocks), and its resistance to retrograde resetting during protracted or polyphase metamorphism.

RSCM THERMOMETRY

Carbonaceous material is a common constituent of metasedimentary rocks, deriving from the solid-state metamorphic transformation of original organic material (Buseck and Huang, 1985). During diagenesis and metamorphism, this carbonaceous material experiences progressive structural organization until it transforms into graphite. The degree of organization is independent of pressure but strongly dependent on temperature, such that the carbonaceous material can be used as an indicator of metamorphic grade (Beyssac et al., 2002a, 2002b; Rietmeijer and Mackinnon, 1985; Wopenka and Pasteris, 1993). Beyssac et al. (2002a) demonstrated that peak metamorphic temperature (T) can be estimated in the range 330–650 °C with a nominal uncertainty of ±50 °C (1σ) by measuring the peak area ratio (R2) of characteristic carbonaceous material bands (D1, D2, G; Fig. 6) in the Raman spectrum and using these values as an input parameter into the equation: T(°C) = –445 R2 + 641.

Rahl et al. (2005) devised an alternative calibration of the RSCM thermometer that extends its range to 100–700 °C. In this calibration, peak metamorphic temperature is calculated from both the peak area ratio (R2) of Beyssac et al. (2002a) and the peak height ratio (R1) of carbonaceous material bands D1 and G. Temperature is calculated using the equation: T(°C) = 737.3 + 320.9 R1 – 1067 R2 – 80.638 R12. However, both of these calibrations were made using a micro-Raman system with a 514-nm-wavelength laser. At Arizona State University, we use a 532 nm laser, which results in a slightly, but systematically larger R2 ratio than that of a 514.5 nm laser (Aoya et al., 2010). To account for this difference, Aoya et al. (2010) derived a new 532 nm laser calibration in which the temperature is calculated using the equation: T(°C) = 221.0 R22 – 637.1 R2 + 672.3, where the R2 ratio derives from the original Beyssac et al. (2002a) calibration. The Aoya et al. (2010) calibration is valid for samples in the range 340–655 °C, and we used this for all of our RSCM calculations.

Sampling and Analysis

We collected 17 samples for RSCM analysis across the Dang Chu klippe, the Ura klippe, and the Zhemgang klippe, encompassing rocks of the Greater Himalayan sequence, Chekha Formation, and Tibetan sedimentary sequence (Fig. 2). Lithologies include paragneiss, pelitic schist, calc-silicate, slate, phyllite, and marble (Table 1). Laser Raman analyses of carbonaceous material were made on microprobe-quality polished petrographic thin sections. In order to avoid variations in mineral orientation and anisotropy on the Raman spectra (Beyssac et al., 2002a; Katagiri et al., 1988), thin sections were cut normal to foliation and parallel to stretching lineation (when present).

Measurements were made using a custom-built Raman spectrometer in the LeRoy Eyring Center for Solid State Science at Arizona State University. The sample was excited using a Coherent Compass laser, with power controlled using neutral density filters. The laser was focused onto the sample using a 50× Mitutoyo objective, and the signal was discriminated from the laser excitation with a Kaiser laser band-pass filter followed by a Semrock edge filter. The system has a spectral resolution of 3.5 cm−1 using a 1200 g/mm grating and a spatial resolution of <1 μm with the 50× objective lens. In order to avoid any mechanical disruption of the carbonaceous material from the thin section making and polishing process (Beyssac et al., 2003), the laser was typically focused on carbonaceous material situated beneath the surface of a transparent grain of quartz or calcite (see GSA Data Repository item A1). The data were collected using an Acton 300i spectrograph and a back-thinned Princeton Instruments liquid nitrogen–cooled CCD (charge-coupled device) detector. Grains of carbonaceous material were analyzed with a 3 mW beam for 120 s over a spectral window of 1100–2000 cm−1. Depending on the abundance of carbonaceous material, between 15 and 25 grains were analyzed in each sample in order to evaluate the degree of in-sample heterogeneity. Peak positions, band areas, and band widths of the resulting Raman spectra were determined with the computer program PeakFit 4.12 (Systat Software Inc.).

Results

All temperatures were calculated using the 532 nm laser calibration of Aoya et al. (2010) and are given in Table 1. For comparison, we also calculated temperatures with the Beyssac et al. (2002b) and Rahl et al. (2005) 514 nm laser calibrations, which gave results in close agreement (GSA Data Repository item B [see footnote 1]). Examples of Raman spectra for each sample are shown in Figure 6 together with R2 values and calculated temperatures. Photographs of representative carbonaceous material grains from selected samples can be found in Data Repository item A (see footnote 1).

In Table 1, the variation in R2 within each sample is indicated by the standard deviation (1σ). Carbonaceous material heterogeneity can result from differences in the original organic material, variations in the structure of the carbonaceous material, the influence of the mineral matrix (e.g., shielding of carbonaceous material within porphyroclasts), or the composition of metamorphic fluids (Beyssac et al., 2002a, 2002b; Large et al., 1994). The average variation in R2 for the 17 samples is 0.095, which corresponds to a temperature difference of ±50 °C. Sample FB132 has the highest variation in R2, at ±0.122, which corresponds to a temperature difference of ±75 °C.

Temperatures calculated using the calibration of Aoya et al. (2010) are reported as standard means of multiple measurements from each sample. The internal uncertainty on our analytical procedures is reflected by the variation in temperature within each sample, and is reported as 1 standard deviation on the mean. However, for each individual value of R2, there is also an associated external uncertainty on the calculated temperature of ±50 °C stemming from the original calibration of carbonaceous material organization against independent P-T data (Beyssac et al., 2002a). Therefore, in order to report a complete and more accurate uncertainty, we added our internal and external uncertainties in quadrature before dividing by the square root of the number of analyses per sample. Final temperatures are thus reported at 2 standard errors of the mean (Table 1; GSA Data Repository item B [see footnote 1]).

Thirteen samples from Chekha and Greater Himalayan sequence units give very consistent temperatures, with an error-weighted mean average of 560 ± 2 °C (2SE) (Figs. 2, 7, and 8). The only change in peak temperature is seen in four samples across the Dang Chu klippe. Two foliated calc-silicates on the NW edge of the klippe, samples FB64 and FB85, give slightly lower peak temperatures of 508 ± 33 °C and 489 ± 26 °C, respectively. The lowest temperatures are seen in sample FB28, a folded marble collected at Pele La within the northern Tibetan sedimentary sequence exposure (Fig. 5B), and sample FB77, a black slate located on the west side of the klippe in the Chekha Formation. These give temperatures 130–140 °C lower than the majority of the samples at 430 ± 30 °C and 420 ± 21 °C, respectively.

Comparison with GARB-GMBP Thermometry

In order to verify the temperatures calculated with the RSCM method, we conducted independent P-T calculations on three of the 17 samples. Samples FB07, FB125, and FB132 have a mineral assemblage of garnet + biotite + muscovite + plagioclase, permitting the application of the well-established GARB (garnet-biotite) exchange thermometer (Ferry and Spear, 1978) and GMBP (garnet-muscovite-biotite-plagioclase) net-transfer barometer (Ghent and Stout, 1981). In order to minimize sources of uncertainty in the thermobarometric calculations, we followed the approach of Cooper et al. (2010) by characterizing textural and geochemical relationships in detail and conducting multiple independent calculations on each sample. For more details, see GSA Data Repository item C (see footnote 1).

Mineral composition data were obtained with a JEOL JXA-8200 electron microprobe at the University of California, Los Angeles, and a Cameca SX50 electron microprobe at the University of Massachusetts. Thermobarometric calculations were made using THERMOCALC v. 3.33 software (Powell and Holland, 1988), and the latest version of the Holland and Powell data set (Holland and Powell, 1998). Activity-composition relationships were calculated using the AX program (Tim Holland: http://www.esc.cam.ac.uk/research/research-groups/holland/ax). Individual P-T calculations and representative mineral analyses can be found in the GSA Data Repository, items D and E (see footnote 1).

RSCM temperatures and GARB-GMBP temperatures and pressures for each sample are given in Table 2 for comparison. The results show that there is good agreement between the three independent temperature measurements within the limits of the uncertainties on both methods, and the pressure estimates on the three samples are also consistent, with an error-weighted mean pressure of 5.2 ± 0.5 kbar (2SE).

IMPLICATIONS FOR THE SOUTH TIBETAN FAULT SYSTEM IN CENTRAL BHUTAN

The lack of either a distinct discontinuity or a progressive change in temperature across the base of both the Dang Chu and Ura klippen, and similar temperatures in the center of the Zhemgang klippe, raise questions about the interpretation of the Greater Himalayan sequence–Chekha contact being a strand of the South Tibetan fault system. In contrast, the lower peak temperatures reached by four samples in the Dang Chu klippe point to a likely detachment between the Chekha Formation and overlying Tibetan sedimentary sequence sediments, an interpretation supported by the change in structural style observed in rocks above and below the contact (Fig. 5).

The interpretation that there is no structural discontinuity between the Greater Himalayan sequence and the Chekha Formation is consistent with the conclusion reached by Long and McQuarrie (2010), i.e., that the Chekha Formation of the Zhemgang klippe (which they interpret as part of the Tibetan sedimentary sequence) is in depositional contact with the Greater Himalayan sequence (Figs. 1 and 2A). However, we disagree with Long and McQuarrie (2010) regarding the broader tectonic significance of that observation. We suggest that the Chekha Formation and overlying Maneting Formation are in the South Tibetan fault system footwall, and that the absence of evidence for fault slip at the Greater Himalayan sequence–Chekha contact in the Zhemgang klippe is not surprising as a consequence. On the other hand, the samples that give low peak temperatures of ∼420–430 °C coincide closely with recumbently folded marbles, which exhibit a wholly different structural style to both the Chekha and Greater Himalayan sequence units below (Fig. 5). Although we have not found fossiliferous marbles in the northern Tibetan sedimentary sequence Dang Chu exposure, the fossils found to the south by Hughes et al. (2011) are from the Tibetan sedimentary sequence. If similar fossiliferous beds are present in the northern part of the klippe, they must lie at structurally higher elevations that have so far proven to be inaccessible. The strongly foliated calc-silicate samples FB64 and FB85, which give intermediate peak temperatures of ∼490–510 °C, are more difficult to interpret. Their high-strain fabrics suggest that they probably lie close to or within the South Tibetan fault system shear zone, and it is possible that they are basal units of the Tibetan sedimentary sequence that have been affected by shear heating, hydrothermal fluid flow, or upward transfer of heat from the footwall into the hanging wall.

Figure 8A shows our interpretation of the distribution of RSCM temperatures in central Bhutan. The Chekha Formation and Greater Himalayan sequence are combined as one sequence in the figure, although our data and observations do not speak to whether or not the two are separated by a major unconformity. We propose a different map pattern for the Tibetan sedimentary sequence in the Dang Chu area, which includes the four lower-temperature samples FB28, FB64, FB77, and FB85, and we interpret the contact between the Tibetan sedimentary sequence and structurally lower units to be the sole South Tibetan fault system detachment. The high-strain calc-silicate samples FB64 and FB85 are mapped at the base of the Tibetan sedimentary sequence, and we suggest that their slightly higher peak temperatures result from heating within the South Tibetan fault system shear zone. This interpretation is consistent with all outcrops we have seen in the area, but the quality of outcrop is so poor that detailed field confirmation of this map pattern is difficult.

If our interpretation is correct, it suggests that to truly understand the kinematics and displacement history of the South Tibetan fault system, we need to focus on this upper contact in the Dang Chu area, not on the previously mapped Greater Himalayan sequence–Chekha contacts. Our interpretation is inconsistent with the contention by Long and McQuarrie (2010) that the stratigraphic contact between Tibetan sedimentary sequence and Greater Himalayan sequence units in the Zhemgang klippe can be used to place a limit of ∼20 km on South Tibetan fault system displacement and thus the magnitude of putative channel flow. The presence of Tibetan sedimentary sequence units ∼80 km south of the Himalayan range crest suggests that displacement on the South Tibetan fault system may in fact be even greater than previously thought (Fig. 8B).

An alternative interpretation of the geology that fits with our data is that of Grujic et al. (2011), who mapped the Chekha Formation more extensively across Bhutan (Fig. 2B). According to their mapping, the majority of RSCM samples that give a consistent temperature of ∼560 °C are situated within the Chekha Formation. The exceptions are samples FB125 and FB132, which lie within the Greater Himalayan sequence on the edge of the Dang Chu klippe, and samples BT1134, BT1136, and BT1138, which lie within the Maneting Formation in the Zhemgang klippe. However, we do not favor their interpretation as we see no evidence for a discrete shear zone at any of the Greater Himalayan sequence–Chekha contacts mapped, and we have found paragneisses reasonably attributed to the Greater Himalayan sequence within areas mapped as Chekha Formation by Grujic et al. (2011) in both the Dang Chu and Ura klippen. Grujic et al. (2011) also mapped the Tibetan sedimentary sequence very differently in the Dang Chu klippe, with no clear explanation as to why. The folded marbles at Baylangdra (Fig. 5A) are mapped as part of the Tibetan sedimentary sequence, but at Pele La (Fig. 5B) they are mapped as Chekha Formation. This is inconsistent with both our temperature and structural data.

CONCLUSIONS

RSCM thermometry data from 17 samples combined with structural observations across three purported South Tibetan fault system klippen in central Bhutan suggest that current maps of this structure require revision. We find no change in peak metamorphic temperature across the contact between Chekha Formation rocks and underlying indisputable Greater Himalayan sequence units. Instead, we see a 130–140 °C drop in temperature across an upper contact between the Chekha Formation and Precambrian–Devonian(?) Tibetan sedimentary sequence sediments of the Pele La and Tang Chu Groups. We therefore see no reason to infer that the Chekha Formation and the Greater Himalayan sequence are separated by the basal strand of the South Tibetan fault system. We regard the upper contact between the Chekha Formation and indisputable Tibetan sedimentary sequence units as the sole South Tibetan fault system detachment, and suggest that future studies of the kinematics and displacement on this system should be focused on this upper contact. The lack of matching hanging-wall and footwall lithologic units or metamorphic grade in the direction of South Tibetan fault system motion suggests that displacement on the South Tibetan fault system may be as much as ∼80 km, not <∼20 km as suggested by Long and McQuarrie (2010). In light of the new data presented here, there is no clear evidence for a breakaway zone for the South Tibetan fault system in southern Bhutan.

This work was supported by U.S. National Science Foundation grant EAR-0838112 to Hodges. We thank Emmanuel Soignard of the LeRoy Eyring Center for Solid State Science at Arizona State University for his help with the Raman spectrometer, and Frank Kyte at the University of California, Los Angeles, and Michael Jercinovic at the University of Massachusetts for their assistance with the electron microprobe analyses. Field work would not have been possible without assistance from Kelin Whipple and Arjun Heimsath and the support of our friends and colleagues in Bhutan: Peldon Tshering (National Environment Commission), Ugyen Wanda (Department of Geology and Mines), Karma Choden and Ugyen Rinzen (Yangphel Adventure Travel). Detailed and constructive reviews by Editor Eric Kirby and two anonymous reviewers are gratefully acknowledged.

1GSA Data Repository item 2012332, Examples of analyzed carbonaceous material, complete Raman spectroscopy on carbonaceous material (RSCM) spectra data, and thermobarometric methods and data tables, is available online at www.geosociety.org/pubs/ft2012.htm, or on request from editing@geosociety.org or Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301, USA.