Multiple origins of large hummock deposits in Alai Valley, Northern Pamir: Implications for palaeoclimate reconstructions

Abstract

This paper interprets the origin of several of the largest hummocky landform assemblages in the Alai Valley, Northern Pamir – a formerly glaciated intermontane depression. The vast hummocky deposits in the Koman-Suu and Achik-Tash river catchments are found to be of two contrasting modes of formation: glacial hummocks deposited during Koman and Lenin Glaciers withdraw and avalanche hummocks produced during catastrophic Koman and Lenin rock avalanches. The origins of the deposits we assessed through remote and field-based geomorphological mapping, as well as sedimentological investigations, which included clast analysis and the identification of micro-scale agglomerates indicative of rock avalanche emplacement. Both the Koman and Lenin rock avalanches were large, catastrophic events (with run-outs of 34 and 24 km, respectively, and a volume over 1 × 10⁹ m³ each) that occurred subsequent to glacier withdrawal from the area. The complex conditions on the moment of the rock avalanche emplacement promoted unusual deposits geomorphology and extensive run-outs. The landslide landforms formed over the pre-existing glacial hummocks and fluvial deposits, and are geomorphologically and sedimentologically distinct from the larger glacial hummocks. The reconstruction of this sequence of events has implications for how hummock dating should be interpreted. This research illustrates large scale catastrophic landsliding in the glacial environment, and adds to the ongoing debate about the misidentification of rock avalanche deposits as of glacial origin, and their relevance to palaeoclimatological and palaeoseismological reconstructions.

Introduction

During the last few decades, advances in dating techniques, such as Terrestrial Cosmogenic Radionuclide (TCR), have facilitated detailed interpretations of the chronology of geomorphological events, including the use of glacial deposits to infer past climates and glacier dynamics (Kuhle, 1997, Jomelli et al., 2009, Ivy-Ochs et al., 2009, Schaefer et al., 2009, Kaplan et al., 2010, Putnam et al., 2010, Putnam et al., 2012). These glacial deposits often can prove to be an important means for inferring palaeoclimate (Schaefer et al., 2009, Putnam et al., 2010). Crucially, however, these palaeoclimate chronologies are only valid if such deposits can be demonstrated to have formed because of the glacier responses to climate changes. Therefore, the correct identification of the deposit origin is required: eliminating the possibility of a rock avalanche deposit (Hewitt, 1999, Deline, 2009, Deline and Kirkbride, 2009, Kariya et al., 2011) or a deposit from moraines formed by supraglacial rock avalanches (Tovar et al., 2008, Shulmeister et al., 2009, Shulmeister et al., 2010, Reznichenko et al., 2012a, Reznichenko et al., 2015).

The term ‘hummocky terrain’ is applied to a wide range of these moraine and also to rock avalanche deposits, which present striking morphological similarity despite their very different origins. In this study, hummock deposit is defined as undulating relief with alternating convex and concave topography (Zech et al., 2005). Glacial hummocky deposits are composed of chaotic assemblages of hillocks and depressions of variable size and shape when viewed from the ground. These are typically interpreted as the product of widespread supraglacial deposition from in-situ ice stagnation resulting from a rapid climatic amelioration (Sissons, 1974, Sissons, 1979), and subsequently are often used as palaeoclimate indicators. They are commonly regarded to form either from meltdown during gradual glacier stagnation, from englacial thrusting in response to compression near the terminus as a consequence either of a) change in basal thermal regime, b) flow against a reverse slope or obstacle (Sharp, 1985, Hambrey et al., 1997, Benn and Evans, 2010, Iturrizaga, 2012), or c) from subglacial meltwater erosion (Munro-Stasiuk and Sjogren, 2006). Hummocky deposits can also result from rapid glacial changes such as surging following by advanced of the tongue to a lower elevation than would normally be supported by climate (Evans and Rea, 1999, Hewitt, 1999) or ice collapse involving displacement of the glacier body (e.g., Huascaran, Peru, 1970; Evans et al., 2009).

Rock avalanches (RA, defined as large slope failures > 10⁶ m³) onto glaciers have been shown to trigger glacier responses independent of concurrent climate, often accompanied by moraine formation or adjustment (Orombelli and Porter, 1988, Hewitt, 1999, Hewitt, 2009, Larsen et al., 2005, Hewitt et al., 2008, Tovar et al., 2008, Deline, 2009, Deline, 2011, Reznichenko et al., 2010, Kariya et al., 2011, Reznichenko et al., 2011, Reznichenko et al., 2012a, Reznichenko et al., 2012b, Reznichenko et al., 2015). They regularly occur in high mountain regions covering large areas with deposits of irregular topography, often with a boulder-rich carapace on the surface (McSaveney and Davies, 2007). These high magnitude catastrophic landslides result in topography similar to glacial hummock fields, which was first described as a “toma-landscape” by Penck and Brückner (1901) (cited in Iturrizaga, 2012), and subsequently around the world (Prager et al., 2006, Dufresne and Davies, 2009, Paguican et al., 2014), including very large volcanic debris avalanches (e.g. Mount Shasta, U.S.A.; Mount St. Helens, U.S.A.), debris flows (e.g. Mount Taranaki Lahars, New Zealand), and large rock avalanches (e.g. Flims RA, Switzerland; Blackhawk RA, USA; Koman RA, also known as Komansu, Kyrgyzstan). The morphology and resulting ground pattern of these hummocks varies from transversely (e.g. Blackhawk RA, U.S.A; Shreve, 1966) or radially aligned ridges, ripples or hummocks (e.g. The Hillocks RA, New Zealand; McColl and Davies, 2011), to almost uniformly distributed, tightly packed hummocks (e.g. Koman RA, Kyrgyzstan; Robinson et al., 2014). In the absence of eye-witness reports, this variety of morphology of the RAs deposits have traditionally been mapped as moraines rather than of mass movement origin (Whitehouse and Griffiths, 1983, Eisbacher and Clague, 1984, Orombelli and Porter, 1988, Evans et al., 1994, Hewitt, 1999, Sanhueza-Pino et al., 2011, Yuan et al., 2013).

The Tien Shan and Pamir are amongst the most tectonically active regions of central Asia, with some of the largest glaciated valleys in the world. Numerous RA deposits have recently been identified in these regions (Owen, 1991, Strom and Korup, 2006, Owen et al., 2008, Sanhueza-Pino et al., 2011, Yuan et al., 2013) including some that were initially considered to be of glacial origin (Zabirov, 1955, Kuhle, 1997, Hewitt, 1999, Strecker et al., 2003). The Alai Valley is the largest intermontane depression in the upper reaches of the Amudarja river basin, which separates the Alai Range of the Southern Tien Shan from the strongly glaciated Zaalai (Transalai) Range of the Northern Pamir (Fig. 1). This region also exhibits extensive hummocky deposits, which have traditionally been attributed to a glacial origin (Zabirov, 1955, Arrowsmith and Strecker, 1999); these hummock features are amongst the largest valley deposits in the world. Catastrophic rock avalanches, such as the Koman RA sourced from the Koman Glacier back wall (Kurdiukov, 1950, Kurdiukov, 1964, Strom, 2014, Reznichenko and Davies, 2014), may be more common than previously thought, carrying risk to local populations and tourists (Robinson et al., 2014). A differentiation in genesis of these deposits and reconstruction of their formation is essential before regional palaeoenvironmental, palaeoseismological or geohazard reconstructions are undertaken. In this study, we aim to determine the genesis of the extensive hummocky deposits in the Koman and Lenin glacial catchments and reconstruct the nature of their emplacement.