Tectonic control of Yarlung Tsangpo Gorge revealed by a buried canyon in Southern Tibet

See allHide authors and affiliations

Science  21 Nov 2014:
Vol. 346, Issue 6212, pp. 978-981
DOI: 10.1126/science.1259041


The Himalayan mountains are dissected by some of the deepest and most impressive gorges on Earth. Constraining the interplay between river incision and rock uplift is important for understanding tectonic deformation in this region. We report here the discovery of a deeply incised canyon of the Yarlung Tsangpo River, at the eastern end of the Himalaya, which is now buried under more than 500 meters of sediments. By reconstructing the former valley bottom and dating sediments at the base of the valley fill, we show that steepening of the Tsangpo Gorge started at about 2 million to 2.5 million years ago as a consequence of an increase in rock uplift rates. The high erosion rates within the gorge are therefore a direct consequence of rapid rock uplift.

Tibetan gorge avoids a tectonic aneurysm

Rapid tectonic uplift was responsible for the immense Tsangpo Gorge on the eastern edge of the Tibetan plateau 2.5 million years ago. Wang et al. found a buried canyon upstream from the gorge along the Yarlung Tsangpo River that began filling with sediments after sudden uplift. Drill cores of the buried canyon sediments show the same river gradient as found downstream of the gorge. The constant river gradient strongly suggests a rapid uplift event created the gorge, rather than river incision as previously believed.

Science, this issue p. 978

The topographic evolution of mountain ranges results from the competition between tectonic and erosive forces (13) and controls the organization of drainage and atmospheric-circulation systems (46). Although tectonics and erosion act in opposing directions, there may be feedbacks that couple the two (7, 8). Prominent candidates for such coupling comprise the syntaxes regions (9, 10), at either end of the Himalaya, where the two biggest rivers draining Tibet, the Indus and Yarlung Tsangpo, have cut deep gorges into very young metamorphic massifs (Fig. 1A) (1117). In the so-called tectonic aneurysm model, it has been proposed that rapid incision within these gorges has thermally weakened the crust and now sustains a positive feedback between uplift and erosion (9, 10), but how and when this happened remains elusive.

Fig. 1 River gorges in the Himalaya.

(A) Geographical overview of Tibet and the Himalaya, with internally drained areas (horizontally hatched), major rivers, and gorges (red and black points) located within steep mountainous areas—i.e., where 5-km-radius local relief is >2 km (red). Black polygons show the mountainous drainage basins of the Indus River in the west and the Yarlung Tsangpo-Brahmaputra River in the east. The Indus-Yarlung Tsangpo Suture Zone separates the Eurasian (north) from the Indian plate (south) and is strongly distorted in the Indus Gorge (IG) and Tsangpo Gorge (TG). (B) Oblique northeastward aerial view of the Yarlung Tsangpo upstream of the Tsangpo Gorge (from Google Earth, (C) Map of the eastern Himalayan syntaxis with the studied valley fill and drill core locations (red points). Contour lines show rapidly exhuming areas with young (<2 Ma) Zircon U/Th-He (orange), and Biotite 40Ar/39Ar ages (yellow) centered on the Tsangpo Gorge (16), which is marked by a red line.

Deeply incised gorges exist along the entire Himalaya and always coincide with very steep river gradients that have been related to zones of rapid rock uplift and incision (1820). The most spectacular and emblematic gorge in the Himalaya, and probably on Earth, is the Tsangpo Gorge, where the Yarlung Tsangpo drops by 2 km in elevation as it cuts through the ~50-km-wide eastern Himalayan syntaxis where erosion rates are exceptionally high (Fig. 2B). The contorted and generally steep course of the gorge is often considered evidence for relatively young capture of the Yarlung Tsangpo by the headward incising Brahmaputra River, from a former course connecting it to the Parlung and Yigong rivers, which themselves might have been connected to rivers farther to the east (5, 6, 9, 18). Provenance studies from the Himalayan foreland, however, indicate that a connection between the Yarlung Tsangpo and Brahmaputra River was already established before the Middle Miocene (2123), emphasizing the stability of the gorge. More recently, the discovery of extensive lake deposits that have accumulated behind a river-blocking glacial dam upstream of the Tsangpo Gorge (24) has spurred the idea that glacial damming during the Quaternary inhibited headward incision of the Brahmaputra River and might have contributed to initiating rapid rock uplift (25).

Fig. 2 Tsangpo Gorge and valley fill.

(A) Longitudinal river profile showing present-day elevation (black; red where passing through Tsangpo Gorge), location of drill cores with observed depth to bedrock (vertical black bars), estimated depth to bedrock from artificial neural network (yellow area), and reconstructed valley bottom before uplift of Tsangpo Gorge (dashed line). IYSZ, Indus-Yarlung Tsangpo Suture Zone. Inset figures show present-day valley bottom width (lower left) and simplified drill core stratigraphy (upper right), with grain size variations (see fig. S1 for details) and sample locations (red points). (B) Hillslope angles within a 10-km distance from the river (mean ± 1σ), specific stream power (12), and landslide erosion rates (17). (C) Mineral cooling ages (1113) within a 10-km distance from the river. See fig. S7 for spatial distribution of cooling ages.

Before entering its narrow gorge, the Yarlung Tsangpo flows for ~300 km across a broad alluvial plain that increases gradually in width toward the confluence with the Nyang River (Fig. 2A). Such downstream widening is typical for lakes that have drowned valleys upstream of dams. Five drillings recently conducted along the Yarlung Tsangpo (Fig. 1C) provide evidence of a thick sedimentary valley fill upstream of the gorge. The drillings were located near the valley center and reached bedrock at depths that increase from 70 m at the most upstream site (no. 1), to a maximum depth of 567 m (no. 3) at ~80 km upstream from the Nyang River confluence. Two more sites at distances of ~40 and 20 km upstream from the gorge yielded depths to bedrock of 510 (no. 4) and 230 m (no. 5), respectively.

Hillslopes bordering the valley floor are generally steep, with uniform slope angles of ~30°, comparable to hillslopes downstream of the Tsangpo Gorge (Fig. 2B) and suggestive of threshold hillslopes that are at their critical angle of stability (14). Simple projection of hillslopes into the subsurface predicts a depth to bedrock of ~1000 m near the Yarlung Tsangpo-Nyang River confluence (Fig. 2A), where the valley floor is widest. Based on the assumption that hillslopes below and above the valley fill are similar, we reconstructed the depth to bedrock (Figs. 1C and 2A) using an artificial neural network approach (26). Close correspondence of estimated and observed depths in drill cores provide us with confidence in the reconstruction, which confirms our initial notion of a steadily decreasing elevation of the former valley floor toward the Yarlung Tsangpo-Nyang River confluence. Downstream of the confluence, the valley floor remains ~4 km wide and deeply filled with sediments (Fig. 2A) until the river leaves the Indus-Yarlung Tsangpo Suture Zone and abruptly narrows to <2.5 km. Because no obvious spill-over toward other valleys exists near the deepest reach of the valley fill, we argue that the former Yarlung Tsangpo was following its present-day course and that uplift of the Namche Barwa and Gyala Peri massifs resulted in steepening of the river where it crosses the gorge and concurrent backfilling of upstream reaches (fig. S4).

Valley-fill sediments from the three sites closest to the Tsangpo Gorge (no. 3 to no. 5) (Figs. 1C and 2A) consist of clastic, mainly fluvial deposits that are dominated by pebbly gravel and sand (figs. S1 to S3). The cores from sites no. 3 and no. 4 both have coarse grain sizes in the lower half, including boulders up to 50 cm in diameter, and fine grain sizes in the upper half, including silt and clay that probably stems from lake periods (Fig. 2A and table S1). This fining upward sequence indicates a decrease in stream competency, consistent with lowered river gradients during backfilling upstream of the Tsangpo Gorge. The lack of upward fining in drill core no. 5, which is closer to the gorge and located ~120 m lower, may be related to erosion of its upper part during uplift and incision of the gorge (fig. S4). We collected three samples from near the base of drill core no. 3 for cosmogenic nuclide burial dating with in situ–produced 10Be and 26Al (26). Our samples yield consistent results that overlap within uncertainties (table S2) and indicate that deposition at this site initiated between ~2 and 2.5 million years ago (Ma) (figs. S5 and S6). Because site no. 3 is ~150 km upstream of the gorge and deposition may not have started immediately, it is likely that uplift and steepening of the gorge was initiated somewhat earlier.

Steepening of the Tsangpo Gorge requires either an increase in rock uplift rate or a decrease in erosional efficiency. Young crystallization and mineral cooling ages from bedrock within the gorge are consistent with an increase in rock uplift rates after ~4 Ma (Fig. 2C) (1113). If rock uplift rates were constant and all surface uplift was due to changes in specific stream power [product of gravity, water density, channel slope, and discharge per unit width (12)], discharge would have to have decreased by more than a factor of four as channel slopes increased in the gorge (Fig. 2A). No observations exist to support discharge reductions of this magnitude, either due to climate change or drainage capture. River-blocking glacial dams (25) can also be excluded as causes of lowered erosional efficiency because the uplifted river reach was below ~2 km elevation, far from any glacial influence. Furthermore, the presence of >200 m of sediments beneath the moraines considered responsible for glacial damming (24) is incompatible with the reduced bedrock incision proposed previously (25).

Our results clearly show that ~2.5 Ma the Yarlung Tsangpo was able to erode back into the Tibetan Plateau and develop a nearly graded profile (Fig. 2A), which is consistent with provenance data from Miocene to Quaternary sediments in the Himalayan foreland of the Tsangpo Gorge (23). The reported increase of Eurasian-plate detritus in deposits of the Brahmaputra River between 7 and 3 Ma (22) could reflect the headward incision of the Yarlung Tsangpo into Tibet until uplift of the Namche Barwa and Gyala Peri massifs focused erosion to Indian-plate rocks within the Tsangpo Gorge (12). This scenario is consistent with an independently estimated onset of accelerated incision of the adjoining Parlung River at 4 to 9 Ma, based on detrital cooling ages (27), which most likely occurred as a result of capture by the Yarlung Tsangpo-Brahmaputra.

Striking similarities between the Indus and Tsangpo gorges have been noted previously (911) and suggest that the tectonic framework of the Himalayan syntaxes sets the stage for the exceptionally high rates of rock uplift (11, 28). Even if positive feedbacks between erosion and uplift nowadays help to maintain these gorges in their current location (9, 10), our results suggest that rapid incision within the Tsangpo Gorge is the result rather than the cause of rock uplift. A similar evolution could have taken place in the Indus Gorge, where mineral cooling and metamorphic ages (9, 10, 14) suggest equally high erosion rates, with some evidence for order-of-magnitude acceleration in exhumation by ~1.7 Ma (15). Notably, ponding of the Indus River upstream of the gorge and deposition of ~1200 m of fluvial and glacial deposits (29) apparently started before ~2.6 Ma (30) and might also reflect an increase in rock uplift in the western Himalayan syntaxis. Because new data are warranted, we leave it to future studies to test the suggested similarity and investigate the cause for synchronous initiation of rapid rock uplift of both Himalayan syntaxes.

Supplementary Materials

Materials and Methods

Supplementary Text

Figs. S1 to S7

Tables S1 to S3

References (3150)

References and Notes

  1. Materials and methods are available in the supplementary materials on Science Online.
  2. Acknowledgments: This work was supported by the National Natural Science Foundation of China (41372211 and 41172179) and the State Key Laboratory for Earthquake Dynamics (Project LED2013A07). D.S. was supported by the Alexander von Humboldt Foundation. Data presented here are archived with the supplementary materials. We thank four anonymous reviewers whose comments improved this paper and B. Hallet for insightful discussions.
View Abstract

Stay Connected to Science

Navigate This Article