T34B-01 16:00h
Cooling and Denudation History of the Marsyandi Drainage in the Central Nepalese Himalaya From Apatite Fission Track and (U-Th)/He Analyses
Apatite fission track and (U-Th)/He analyses document the cooling history and the links between climate and tectonic processes of the Marsyandi river catchment in the central Nepal Himalaya. The Marsyandi river originates in the Tibetan Plateau and traverses the High Himalaya to the east of Annapurna, crossing the Tibetan detachment and the Main Central Thrust. We have obtained $>$80 apatite fission track (AFT) and (U-Th)/He ages from samples collected along the Marsyandi drainage, and along ridges on either side of the river. The youthfulness of this orogen is pushing the limits of both techniques. AFT ages range from $>$3 to 0$\pm$0.2 Ma, and He ages from 0.7 to 0.3 Ma, with AFT and He ages at river level of 0.4 and 0.3 Ma, respectively. Samples were collected along ridge lines that ascend away from the river. Although ages in general increase with increasing elevation away from the valley bottom, striking geographic variation in the total age range occurs: several ridges have maximum ages of 1 to 0.8 Ma at 2 km above the valley bottom, whereas only one ridge crest (in the southwest of the field area) displays ages as old as $\sim$ 2 Ma. We suspect that there is a structural boundary between this ridge and others to the north and east. With these data, we calculate cooling and infer denudation rates over the last 2 Ma. Cooling and denudation rates are uniformly high ($\geq$1 km/m.y.) from 2 to 0.8 Ma, but they appear to have accelerated at 0.8 Ma to 3 km/m.y. Denudation rates from 0.3 Ma until the present are calculated by assuming (appropriately) high FT and He closure temperatures of 140 and 90$\deg$C, and a geothermal gradient of 50 to 100$\deg$C/km, indicating rates of 3 to 8 km/m.y. The thermochronologic data are being used in conjunction with geomorphic and climate data to evaluate the mechanisms controlling the physiography of one of the world's major mountain belts.
T34B-02 INVITED 16:15h
Evidence for Repeated Megafloods Down the Tsangpo River Gorge, Southeastern Tibet
Lacustrine terraces record the extent of at least four glacially dammed lakes immediately upstream of the Tsangpo River gorge at the eastern syntaxis of the Himalaya. Field work in 2002 and subsequent GIS analyses constrained the extent of two of the lakes, a younger 240-m-deep lake and an older, 680-m-deep lake. Radiocarbon dating of wood and charcoal yielded ages of 8860$\pm$40 and 9870$\pm$50 $^{14}$C yr B.P. for the higher set of lake terraces, and 1220$\pm$40 and 1660$\pm$40 $^{14}$C yr B.P. for sediments from the lower terraces. Field work in 2004 revealed evidence for both a younger shallower lake and an older deeper lake. The 680-m-deep paleolake discovered in 2002 covered almost 2850 km$^{2}$ and contained an estimated 835 km$^{3}$ of water; the 240-m-deep paleolake contained an estimated 81 km$^{3}$ of water. These two dated paleolakes correlate with the timing of glacial advances due to monsoon strengthening indicated by glacial advances at Chomolongma (Mount Everest). In addition, preliminary $^{10}$Be dating of ridge-crest boulders from a sequence of moraines about 50 km NE of the Tsangpo gorge indicates a series of at least three glacial advances, the youngest of which appears to be the same age as the 680-m-deep paleolake. Two earlier Pleistocene glacial advances were larger and may correlate with the deeper lake discovered in 2004. Catastrophic failure of the glacial dams that impounded the two dated paleolakes would have released outburst floods down the gorge of the Tsangpo River with estimated peak discharges of up to 1 to 5 x106 m$^{3}$ s$^{-1}$. The erosive potential represented by the unit stream power calculated for the head of the gorge during such a catastrophic lake breakout indicates that post-glacial megafloods down the Tsangpo River were likely among the most erosive events in recent Earth history. Our evidence for previously unrecognized glacially dammed lakes at Namche Barwa show that monsoon-driven valley glacier advances dammed even the largest Himalayan rivers, and repeatedly created unstable glacier-dammed lakes that generated floods likely unparalleled in the recent history of the Himalaya. Hence, such immense outburst floods may well have played an important role not only in carving the deepest valley on Earth, but also, more generally, in the development of the spectacular topography across the Himalaya and other high, glaciated ranges.
T34B-03 16:30h
Erosion rates and processes across two trans Himalayan transects in central Nepal
Quantifying erosion rates across mountain ranges is difficult. Constraining long and short term erosion rates for the Himalaya is especially important as we attempt to quantify the degree of coupling among climate, tectonics, and surficial erosion in the quintessential collisional orogen. Conclusions drawn from recent studies in adjacent drainage basins across the higher Himalayan Range in central Nepal are inconsistent in their interpretations of the coupling between climate and long-term exhumation inferred from low-temperature thermochronometers. These differing interpretations underscore the need for additional erosion rate data representing timescales approaching the decadal timescales of our precipitation records. Here we present new results from an extensive study using in situ produced cosmogenic nuclides in both of these adjacent drainages: the Marsyandi and Burhi Gandaki Rivers. Average erosion rates determined from nuclide concentrations in detrital sands from first and second order catchments show similar patterns for both transects. Specifically, erosion rates increase from roughly 1-2 mm/yr where monsoon precipitation is under 2 m, to peak rates of about 2 and 7 mm/yr for the Marsyandi and Burhi Gandaki, respectively, corresponding to the region of highest precipitation. Both basins show a decline in average erosion rates and a corresponding decline in precipitation across the "rain shadow" of the main Himalayan crest. We suggest that the observed pattern of erosion rates may be largely controlled by climate, but is modulated by the locus of active structures in each drainage. Measurements of nuclide concentrations from ridge crest samples for the Marsyandi transect show little change in erosion rates with the same range of precipitation, suggesting a process-based discontinuity between ridge crest and side slopes. These ridge crest rates, averaging under 0.5 mm/yr, show a similar pattern to, but are significantly lower than, the spatially constant apatite fission track ages predicting long-term exhumation rates of about 2-5 mm/yr. Overall, these data support work correlating decadal records of climate with long term inferences of exhumation and erosion.
T34B-04 16:45h
Thermal and Mechanical Modeling of Bedrock Cooling Ages in the Nepalese Himalaya
We introduce a new method to use detrital mineral cooling ages in conjunction with a digital elevation model (DEM) to test numerical models of collisional orogens. We apply this methodology to the Marsyandi valley in the central Nepalese Himalaya, where we use a 2D kinematic-and-thermal model to predict variations in bedrock cooling ages along an averaged structural and topographic transect within the Himalaya. The model is based upon a simple ramp-and-flat style decollement, representing the Main Himalayan Thrust and is constrained by the INDEPTH transect, surface geology, seismicity, and geomorphology. The range is assumed to be in a topographic steady state, such that all rock advected to the surface is eroded. The 2D kinematic-and-thermal model in the plane of advection is extrapolated laterally (parallel to the range) to calculate the 3D distribution of cooling ages predicted for actual Himalayan topography. The modeled detrital cooling-age signal results from convolving the distribution of cooling ages within a catchment with the rate at which individual points are eroding and with the distribution of the mineral used as the thermochronometer. Predicted distributions of cooling ages are compared with detrital $^{40}$Ar/$^{39}$Ar muscovite data to assess varying tectonic scenarios. Model results, assuming that the Main Boundary Thrust represents the surface expression of the Main Himalayan Thrust, illustrate that the distribution of detrital cooling ages is sensitive to how the observed 20 mm/yr of Indo-Tibetan convergence is partitioned: the best-fit model assigns 5 mm/yr to overthrusting by the Asian plate and 15 mm/yr to underthrusting by the Indian plate. A trade off exists, however, between ramp geometry and the best-fit rate. A model using the approximate present position of the Main Central Thrust to represent the surface expression of the Main Himalayan Thrust provides a better fit to the observed data and supports evidence for recent activity on the Main Central Thrust.
T34B-05 17:00h
Holocene Monsoonal Dynamics and Fluvial Terrace Formation in the NW Himalaya
Along the lower Sutlej Valley in the NW Indian Himalaya, cut and fill terraces reflect Holocene climatic variations that match known episodes of variable precipitation. We dated a flight of six fluvial terrace surfaces in conglomeratic valley fill using cosmogenic radionuclides (26Al, 10Be) to reveal the link between Holocene Indian Summer Monsoon (ISM) oscillations and fluvial processes. The onset of heavy precipitation and farther penetration of moisture into the orogen during the Holocene ISM optimum at approx. 10ka and coeval amplified sediment flux at higher elevations choked the wide, low-gradient valleys with debris along the southern Himalayan front. Sedimentation of the highest fluvial valley fill, 120m above the present-day river, occurred almost instantaneously at 9.7ka, when vast amounts of sediment from erosion off vegetation-free hillslopes and formerly lodged glacigenic deposits were deposited. Between 8.3 and 2.5ka, well preserved terraces sculpted into these deposits at successively lower elevations reflect short-term, transient episodes in river profile development. Incision and channel abandonment between ISM onset and today coincides with centennial-long, drier intervals in the oscillating monsoonal system. Contrary to common concepts linking fluvial incision of terrace systems to increased precipitation and runoff, we demonstrate that less moisture during centennial oscillations of the Indian monsoonal system results in lower sediment supply, allowing flux-undersaturated rivers to incise episodically.
T34B-06 INVITED 17:15h
Dynamics of Monsoonal Forcing, Tectonics and Erosion Processes Along the Southern Himalaya, Sutlej Valley, NW India
Under present-day conditions the interplay between topography and Indian Summer Monsoon (ISM) circulation controls precipitation, sediment transport, and river discharge along the Southern Himalayan Mountain Front (SHF). Precipitation measurements derived from passive microwave analysis (SSM/I) show that monsoonal vortexes draw moisture from the Bay of Bengal, move northwestward and result in heavy rainfall when colliding with the mountain front. The end of the monsoonal conveyer belt is near the Sutlej Valley of NW India and corresponds to a climatic transition sensitive to changes in ISM strength. For example, during the 2002 and previous abnormal monsoon years, violent rainstorms penetrated far into otherwise arid regions in the NW Himalaya at elevations > 3000 m, while rainfall amounts in lower-elevation sectors were not significantly increased. Farther moisture transport into the orogen also enhanced fluvial sediment flux and emphasizes the importance of these short-lived events in the modern ISM variability for geomorphic processes. On centennial time scales the Sutlej region also shows a close link between monsoon oscillation, sediment flux and relief production. The onset of the strengthened ISM in early Holocene time was accompanied by deposition of massive conglomerate deposits in low-gradient sectors of the Sutlej. Sculpted into this 120-m-thick valley fill were six fluvial terraces at successively lower elevations, cut between 9 and 2.5 ka (Bookhagen et al., this session). Channel abandonment and terrace formation correlate with reduced moisture during centennial oscillations of the Indian monsoonal system, which have resulted in lower sediment supply, allowing flux-undersaturated rivers to incise episodically and evacuate material to foreland areas. At the level of millennial ISM oscillations clusters of exceptionally large bedrock landslides were generated in the high-elevation sectors of the orogen during protracted intensified monsoon phases at about 27 and 6 ka. These episodes also correlate with significantly increased sediment-flux rates and are inferred to correspond to higher precipitation, lateral undercutting of rivers, and higher pore pressures, related to moisture penetration far into the arid parts of the orogen. Over much longer time spans, involving hundred thousands to millions of years, late Pliocene to Quaternary apatite fission track cooling ages reveal a coincidence between rapid erosion and exhumation focused in an up to 70-km-wide sector of the SHF (Thiede et al., this session). This sector of the orogenic wedge currently receives more than 80% of precipitation related to moisture-bearing winds impinging on the SHF. This concentrated loss of material is compensated by an area of pervasive brittle deformation, suggesting that protracted climatically controlled surface processes may localize deformation in the internal part of the Himalaya.
T34B-07 INVITED 17:30h
Self-organized balance between rapid erosion and uplift in the eastern Himalayan syntaxis.
Spectacular knickpoints on all the major rivers crossing the Himalaya are enigmatic because they are expected to be among the most dynamic elements of the landscape and yet appear essentially stationary. According to a simple but robust model of fluvial incision, knickpoints would retreat $\sim$100 km in 1 million years where fluvial incision of bedrock is rapid (up to $\sim$10 mm/a), in the absence of localized tectonic uplift. Knickpoint retreat slows markedly with modest tectonically induced antiformal uplift. With increasing, but still modest tectonic rock uplift (up to 2 mm/a) as well as crustal rebound due to erosional unloading, knickpoint retreat stops. If antiformal rock uplift locally outpaces fluvial incision, the knickpoint will tend to shift downstream and sediments will accumulate upstream of the uplift. To illustrate that knickpoints are sensitive indicators of the balance between erosion and crustal uplift, we present diverse data from the Yarlung Tsangpo/Brahmaputra where the river plummets $\sim$2000 m over the largest knickpoint in the range. In this region, which includes the Namche Barwa-Gyala Peri massif as well as the Yarlung Tsangpo gorge, exceptionally young zircon fission-track dates (0.2 Ma), biotite $^{40}$Ar/$^{39}$Ar ages (1.0 Ma), [(U-Th)/He] cooling ages (0.3 Ma) for zircons, and U-Pb ages of anatexis from bedrock samples define distinct regions where extremely rapid exhumation has been sustained for at least 3 Ma. Because these independently determined regions of rapid exhumation closely coincide spatially with one another, as well as with the region where calculated current fluvial incision rates reach their peak, the locus of rapid erosion has not migrated significantly relative to the underlying crust. As the rapid exhumation is clearly associated with the knickpoint, sustaining rapid erosion in this region requires the knickpoint to be essentially stationary. Our simple model 1) suggests that rock uplift must be within 1.5% of the incision rate for the knickpoint retreat not to exceed 20 km, consistent with the geochronological data, and 2) shows that localized uplift naturally accounts for the sediment ponding and alluvial reach upstream of the knickpoint that is characteristic of the Yarlung Tsangpo and other major rivers traversing the range. We suggest that major knickpoints on other large rivers in the Himalaya are also essentially stationary, and hence, reflect a near-perfect balance between erosion and rock uplift. This balance is not coincidental; it must arise from direct mechanisms, involving strong feedbacks that bring erosion and rock uplift locally into mutual alignment. This erosion/uplift balance may be of broad interest, because where it occurs, insight into the spatial variation in the rates of tectonic, structural, and metamorphic development could be obtained readily as they should mimic the spatial variation in rates of erosion, which can be estimated for broad areas with relative ease using remotely sensed data. Rapid erosion and uplift tend to be highly localized, hence even in steady state landscapes rates of erosion and uplift tend to vary spatially, and fluvial evacuation of crustal material can be highly localized.
T34B-08 17:45h
Lateral Advection of Bedrock and Steady-State Convexo-concave Stream Longitudinal Profiles: Implications for the History of the South Tibetan Fault System
Recent geological studies and geodynamic models suggest that rock comprising the Higher Himalaya is extruded from beneath the Tibetan Plateau along the South Tibetan fault system and the Main Himalayan and Main Central Thrusts. If true, longitudinal profiles of Transhimalayan rivers should reflect the velocity field of extrusion. Longitudinal profiles of the Transhimalayan rivers are convexo-concave as they pass from the Tibetan Plateau through the Higher Himalaya to the Lesser Himalaya and Sub-Himalaya. Convex reaches, on the plateau and in the Higher Himalaya, are as long as 50 km. Does this convexo-concavity develop as a steady-state consequence of extrusion? Other possible explanations include transient knickzone migration due to changes in climate/base-level or spatial variations in lithology and sediment size. Using the CHILD coupled surface process-tectonics model, we demonstrate conditions under which steady-state convexo-concave profiles may occur on a plateau margin undergoing lateral crustal extrusion. Whereas stream-power law models of detachment-limited streams in equilibrium with vertical rock uplift and uniform climate and lithology produce typical concave stream profiles, the inclusion of lateral bedrock advection can produce steady-state convexo-concave profiles between an upper-bounding low-angle normal fault and lower-bounding thrust fault. The convexities in about half of the Transhimalayan rivers occur above the South Tibetan fault system, suggesting that a simple extrusion model alone does not satisfactorily explain the profiles of Transhimalayan rivers if they are in equilibrium with bedrock flux. Results are consistent with GPS measurements showing the southward displacement of Tibet relative to India and no appreciable modern activity on the South Tibetan fault system. An increase in precipitation in southern Tibet during the late Cenozoic associated with the Asian monsoon or a decrease in the rate of lateral extrusion are more likely causes of convexities north of the South Tibetan fault system.