Unravelling the Pleistocene glacial history of the Pamir mountains, Central Asia

https://doi.org/10.1016/j.quascirev.2021.106857Get rights and content

Highlights

  • Large Middle Pleistocene (≥220 ky) glaciers shaped the major west Pamir valleys.

  • During the penultimate glacial cycle, major valleys were largely ice free.

  • Paraglacial destabilization led to enormous (≥0.05 km3) debris flows.

  • Warming and/or aridisation at the end of MIS-5 led to regional glacial melting.

  • We propose a new agorithm to detemine a moraine age from a boulder age distribution.

Abstract

Several hundred thousand year old moraines preserved in the semi-arid environment of High Mountain Asia attest to Middle Pleistocene glaciations, but the regional correlation of glacial stages and the spatial extent of the glacial advances remain poorly constrained. We examined glacial landforms and Quaternary sediments in the Bartang valley, northwestern Pamir, a region with no previous quantitative glacial chronology. Using cosmogenic 10Be exposure ages, we dated glacially polished bedrock, moraines, and mass wasting deposits. Our data show that the northwestern Pamir was heavily glaciated in the Middle Pleistocene (≥220 ky) with large valley glaciers occupying some of the major valleys in the western Pamir. During the penultimate glacial cycle (191–130 ky) these valleys may have been largely ice free. Catastrophic mega debris flows with volumes ≥0.05 km3 occurred after the ice retreat and reflect paraglacial destabilization of glacial sediments. The age of the best-dated mega debris flow (81 ± 4 ky) is similar to moraine ages ∼70–80 ky documented throughout the Pamir, demonstrating that remobilized sediments may provide valuable age constraints on glacial histories. In order to facilitate regional comparison of glacial chronologies, we developed a Gaussian separation algorithm, which determines a moraine age from a distribution of boulder exposure ages based on the assumption that post-depositional processes prevail over inheritance, and that the oldest boulder ages best represent the timing of moraine formation. We compiled moraine boulder exposure ages from the Pamir and adjacent regions and provide a summary of Middle and early Late Pleistocene glacial cycles of western High-Mountain Asia.

Introduction

High Mountain Asia contains the highest concentration of glaciers outside of the polar regions (Yao et al., 2012). The monsoon-influenced eastern and southern margins of High Mountain Asia mostly record glacial advances ≤30 ky, older glacial landforms are preserved in the semi-arid orogenic interior (Amidon et al., 2013; Owen and Dortch, 2014). Quantifying the age and extent of old glacial stages is challenging because in degraded moraines, cosmogenic exposure ages of boulders underestimate the age of moraine formation (e.g. Heyman et al., 2011). Moreover, glacial sediments may be similar to mass wasting deposits, and distinguishing landslides and rock avalanches from moraines is difficult if the diagnostic morphologies are poorly preserved (Hewitt, 1999; Benn and Owen, 2002). Middle and Late Pleistocene glacial advances in High Mountain Asia have been related to global climate oscillations, to changes in the strengths of regional climate systems, and to topographic controls (e.g., Derbyshire, 1996; Benn and Owen, 1998; Amidon et al., 2013; Dortch et al., 2013). However, the interpretation and regional correlation of glacial stages depend on accurate determination of moraine ages, which is often compromised by highly dispersed boulder ages.

A long history of Pleistocene glacial advances has been documented in the Pamir mountains at the western end of High Mountain Asia (Fig. 1A). The western Pamir is a high-relief mountain range with peak altitudes ∼7000 m above sea level (asl) and deeply incised valleys (∼1500–2500 masl). Glacier snowlines (equilibrium-line altitude, ELA, from the Randolph Glacier Inventory, RGI Consortium, 2017) range from <4000 masl west of the NW Pamir massif to >5000 masl east of the massif (Fig. 1B). The ELA variation reflects moisture availability: the Pamir receives precipitation from the mid-latitude westerlies resulting in higher annual precipitation and wetter winters in the western Pamir compared to the east (Fig. 1C). An orogenic plateau (median elevation 4350 masl) exists east of ∼73°E. The peaks of the “Chinese” Pamir east of the plateau are part of the uplifted footwall of the Muji–Tashkurgan graben system and include the Kongur Shan and Muztagh Ata ice caps (Fig. 1A-B). Pleistocene glacial records have been studied in the Chinese Pamir (Seong et al., 2009; Owen et al., 2012; Hedrick et al., 2017), in the central Pamir (Zech et al., 2005; Abramowski et al., 2006; Röhringer et al., 2012) and in the SW Pamir (Stübner et al., 2017, Fig. 1A). Despite the abundance of moraine boulder ages, the spatial extent and temporal correlation of glacial advances across the Pamir remain elusive. Here, we provide new age constraints on glacial landforms and mass wasting deposits in the NW Pamir, a region with no previous quantitative glacial chronology. We document extensive Middle Pleistocene valley glaciation consistent with similar results from the SW Pamir (Stübner et al., 2017) and with Middle Pleistocene glacial sediments in the Chinese Pamir (Seong et al., 2009; Owen et al., 2012). To facilitate the regional comparison of Pleistocene glacial chronologies, we developed an algorithm that determines probable moraine ages from dispersed boulder exposure ages based on the assumption that post-depositional processes prevail over cosmogenic nuclide inheritance, and that the oldest boulder ages best represent moraine formation.

Section snippets

Field mapping and 10Be dating

We mapped Quaternary deposits and landforms in the upper Bartang valley, NW Pamir (Fig. 1, Fig. 2A). Quaternary deposits were characterized in the field and by the analysis of Google Earth images and 12-m TanDEM-X WorldDEM (Wessel et al., 2018) digital elevation data. We outlined deposits based on their morphology, characteristic colours, and weathering patterns. Special emphasis was put on evidence for erosion of the surface of the deposits such as the development of desert pavement, hillslope

Landforms of the upper Bartang valley

The Bartang river valley is one of the narrowest and most deeply incised valleys of the western Pamir. The topographic relief calculated over a 10-km radius is 3–4 km, valley cross sections are V-shaped, and the valley floor is typically not much wider than the river channel (50–100 m). East of 72°15′E, the valley widens to form the ∼5 × 20 km2 upper Bartang basin (Fig. 2A). The basin is connected to the glaciers of the NW Pamir massif via the 75 km long and ∼1 km wide Tanymas–Gudara valley, a

Quaternary evolution of the upper Bartang basin

The bedrock morphology in the upper Bartang basin was shaped by glacial erosion (axial ridge, hillocks, bedrock flats) and subsequent fluvial re-equilibration (river gorges, Bartang river terraces). The axial ridge, the glacial striations, and the knobby bedrock, which we interpret as rock drumlins, indicate a W to SW direction of ice flow. The Bartang glacier was probably sourced in the NW Pamir massif from where it flowed via the Tanymas and Gudara valleys (Fig. 3). The accumulation area may

Conclusions

We documented glacial erosion surfaces and glacial, proglacial and paraglacial sediments in the upper Bartang valley, NW Pamir. Using cosmogenic 10Be exposure ages of boulders and glacially polished bedrock, we constrained the Middle and Late Pleistocene glacial history and landscape evolution:

  • In the Middle Pleistocene, the Tanymas–Gudara–Bartang glacier extended over >100 km from the NW Pamir massif to the upper Bartang basin. The ‘Bartang’ glacial stage ended at ∼220 ky. We propose that the

Author roles

Konstanze Stübner: Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Software; Validation; Visualization; Writing – original draft. Bodo Bookhagen: Methodology; Software; Validation; Data curation; Writing – review & editing. Silke Merchel: Investigation; Validation; Writing – review & editing. Johannes Lachner: Investigation; Validation; Writing – review & editing, Mustafo Gadoev: Investigation; Writing – review & editing

Data availability

Datasets related to this article can be found at https://github.com/kstueb/MoraineAge_GaussianMixture.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

This study was funded by the German Science Foundation grants STU 525/2–1 and STU 525/2-2. AMS measurements at the VERA facility (Uni. Vienna) were supported by the RADIATE project under the Grant Agreement 824096319 from the EU Research and Innovation programme HORIZON 2020 via the Transnational Access grant 19001739-ST. TanDEM WorldDEM data were made accessible by the German Aerospace Center (DLR) through grant DEM_GEOL1296. We are thankful to Jakob Heyman and an anonymous reviewer for

References (43)

  • J.M. Dortch et al.

    Quaternary glaciation in the Nubra and Shyok valley confluence, northernmost Ladakh, India

    Quat. Res.

    (2010)
  • J.M. Dortch et al.

    Timing and climatic drivers for glaciation across semi-arid western Himalayan–Tibetan orogen

    Quat. Sci. Rev.

    (2013)
  • E. Grin et al.

    10Be surface-exposure age dating of the Last Glacial Maximum in the northern Pamir (Tajikistan)

    Quat. Geochronol.

    (2016)
  • K.A. Hedrick et al.

    Towards defining the transition in style and timing of Quaternary glaciation between the monsoon-influenced Greater Himalaya and the semi-arid Transhimalaya of Northern India

    Quat. Int.

    (2011)
  • K.A. Hedrick et al.

    Quaternary history and landscape evolution of a high-altitude intermountain basin at the western end of the Himalayan-Tibetan orogen, Waqia Valley, Chinese Pamir

    Geomorphology

    (2017)
  • K. Hewitt

    Quaternary moraines vs catastrophic rock avalanches in the Karakoram Himalaya, northern Pakistan

    Quat. Res.

    (1999)
  • J. Heyman et al.

    Too young or too old: evaluating cosmogenic exposure dating based on an analysis of compiled boulder exposure ages

    Earth Planet Sci. Lett.

    (2011)
  • G. Korschinek et al.

    A new value for the half-life of 10Be by heavy-ion elastic recoil detection and liquid scintillation counting

    Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. Atoms

    (2010)
  • S.M. Marrero et al.

    Cosmogenic nuclide systematics and the CRONUScalc program

    Quat. Geochronol.

    (2016)
  • S. Merchel et al.

    Attempts to understand potential deficiencies in chemical procedures for AMS: cleaning and dissolving quartz for 10Be and 26Al analysis

    Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. Atoms

    (2019)
  • L.A. Owen et al.

    Nature and timing of Quaternary glaciation in the Himalayan–Tibetan orogen

    Quat. Sci. Rev.

    (2014)
  • Cited by (7)

    • A tool for the ages: The Probabilistic Cosmogenic Age Analysis Tool (P-CAAT)

      2022, Quaternary Geochronology
      Citation Excerpt :

      These limitations have been addressed over the decade long development period of P-CAAT, based on preliminary work by Dortch et al. (2013) and Murari et al. (2014). Recently, Stübner et al. (2021) developed a Python implementation (henceforth referred to as S-PDE) inspired by Dortch et al. (2013) PDE approach; this method is included for comparison. Irrespective of the choice of statistical approach for outlier identification, there is a clear need to standardize uncertainty reporting.

    View all citing articles on Scopus
    View full text