Glaciers of the Karakoram Himalaya

The Karakoram Himalaya comprises the highest, most heavily glacierised watersheds of the upper Indus and Yarkand River basins. It is set within other vast mountain systems of High Asia that support more than 100,000 km² of perennial snow and ice cover. The chapter situates the Karakoram in this Central Asian context and outlines key aspects of the regional environment. A variety of other cold region or cryosphere features are of interest including seasonal snow covers and perennially frozen ground or permafrost. These affect much greater areas than the glaciers, as do periglacial conditions and snow avalanches. Rock glaciers are concentrated in some parts, periglacial systems closely related to glaciation in distribution, climate responses and genesis. Two main aspects of the Karakoram environment are reviewed: the region’s geology and geotectonic evolution, and its climate. The glaciers would not exist at all, or be so extensive, without the great elevations created by mountain-building forces. Significant aspects of their morphology and behaviour relate to rugged, steep landscapes developed through deep dissection by rivers and past glaciations. The Karakoram climate is influenced by three seasonally varying weather systems: the predominantly winter Westerlies, the summer monsoon and Inner Asian high-pressure systems. The glaciers are affected by each system, their relations to one another, by their strong year-to-year variations and how, in turn, they are influenced by the high mountain terrain. In the lower Karakoram valleys and surrounding high plateaux, dry conditions are found, including where most weather stations are located. However, the glacierised area is, in fact, largely humid. Measured snowfall and water yields from glacier basins challenge a long-held view of the Karakoram as part of what has been termed ‘the semiarid Himalaya’. Snowfalls in glacier source areas are in the range 1,000–2,000 mm water equivalent. Also, summer snowfall at high elevations will be shown to be a major factor in sustaining the glaciers, whereas winter precipitation dominates in weather station records. Estimates of the much-studied snowlines are shown to be problematic since they occur where freeze–thaw cycles, wind and avalanche redistribution of snow, are concentrated. The importance of Karakoram glaciers for the flows of the Yarkand River and main stem of the Indus is outlined.

The distribution of perennial snow and ice in the Karakoram Himalaya is examined and its area–altitude relations. The presence and extent of snow and ice are shown to depend upon, and be positively correlated with, interfluve heights. The elevations and extent of the highest altitude terrain are of decisive significance. The size, length and lowest reach of glaciers increase as elevation increases up to the highest watersheds. In the Central Karakoram, the ‘glaciation level’, or minimum elevation needed to generate a glacier, is found at about 5,250 m on north-facing slopes and 5,500 m on south-facing slopes. At the western margins, the averages are 4,600 m and 5,200 m, respectively. They rise eastwards by about 1,200 m and 900 m to the highest glaciation levels found in the eastern margins of the Karakoram. A main set of the 42 largest valley glaciers is introduced, with basin areas exceeding 130 km² and ice streams over 16 km in length. These have exceptional elevation ranges, five spanning more than 5,000 m and 34 more than 3,000 m. Their long profiles exhibit two main features. Most of the vertical descent is accomplished in less than 10 % of ice stream lengths, mainly in icefalls in the upper parts of the basins. However, their longest sections, in the middle and lower reaches, are of relatively gentle gradient. Some 85 % of main ice stream areas lie between 4,000 and 6,000 m, the critical elevation zone in terms of ice cover. This must be balanced against the extreme high elevations of their watersheds and also the exceptionally low termini of many Karakoram glaciers compared to most in the Greater Himalayan Region.

This chapter focuses on the glaciers themselves, ice morphology and distribution and relations to terrain in their basins. To identify the relative extent of zones with particular conditions, criteria are outlined to define terrain elements on and off the ice. Estimates are given of their share of basin areas for the 42 largest glaciers and provide a basis for the investigation of glacier mass balance and landforms in later chapters. The importance of verticality is further reinforced. Relative distributions of perennially frozen, snow-covered areas and seasonally snow-free and thawed surfaces are underscored. These prove to be distinct from conventional glacier accumulation and ablation zones in some important respects, mainly due to complications arising from the extent of off-glacier rock walls and of snow redistribution by wind and avalanches. Icefalls, supraglacial debris mantles and ice-margin deposits are of particular interest. Glacier ice itself is shown to comprise, on average, one-third of basin areas above glacier termini and about 45 % of perennial snow and ice areas. Rock walls make up more than 65 % of the latter and 60 % of whole basin areas. Considerable differences emerge between basins, as illustrated by comparing Baltoro, Biafo and Toltar–Baltar Glaciers. They support an argument for revisiting older classifications of Karakoram Glaciers into Mustagh, Turkestan and Alpine types, which emphasise differences in nourishment, especially the relative shares of avalanche-fed ice. A fourth ‘wind-fed’ class, widely present in the region, is added. With certain revisions, these classes help to identify key conditions in the Karakoram. The four types are also distinguished by the proportions of terrain elements in their basins, notably rock walls and the extent or absence of accumulation zones. Classifications by size, morphology, climatic and thermal regimes are also examined. They raise some unique questions and reinforce a sense of the distinctiveness and diversity of ice masses in the Karakoram.

The inputs to glacier mass balance and phenomena that influence them in the Karakoram are considered in this chapter. Snowfall in the source areas of the glaciers is the first concern. Observations at Biafo Glacier, an Alpine type, offer a rare, relatively detailed quantification of high-altitude snowfall in the accumulation zone. The data show that maximum precipitation in the Central Karakoram occurs in glacier source zones. Between 4,800 and 5,800 m snowfall exceeds 1,000 mm water equivalent annually, rising to as much as 2,500 mm in some years and locations. This is much more than at valley weather stations, and whereas the latter have a dominant winter season precipitation, an ‘all-year’ input regime is indicated for the glacier basins in the Mustagh Karakoram. Summer snowfall is partly from Arabian Sea moisture but the summer monsoon proves significant in most years. Establishing the presence and quantities of firn and firn limits is problematic. In Biafo’s accumulation zone, extensive, changeable surface and near-surface subzones are described in cold, dry and wet snow, summer horizons with dirt and ice layers, and complicated relations to the ablation zone below. Thick, clean snow layers from individual, mainly winter, storms are observed and also depth hoar. Such information, from the accessible high elevation areas of an Alpine-type glacier, is a basis to consider what happens on the much greater numbers of less accessible Mustagh- and Turkestan-type glaciers where the intervening role of avalanching is critical. Avalanches dominate snow inputs and continue to add mass far into what otherwise appear as ablation zones. Thus, a significant component of avalanche nourishment occurs below where firn lines and equilibrium line altitudes are conventionally placed. At higher elevations and around interfluves, the evidence of wind action is particularly important in feeding small and minor ice masses and as an influence on avalanching.

The negative or loss factors in mass balance are examined, mainly processes, controls and rates of ablation on glacier surfaces and some other contributors to water yields from glacier basins. Limited but actual ablation measurements exist for three of the large glaciers, Baltoro, Biafo and Batura. They are described and interpreted in relation to differing nourishment and debris covers. Ablation is largely due to solar radiation but regulated by surface conditions in ablation zones. The length of the ablation season is important, as are the extent of exposed ice and of debris covers. These vary strongly with elevation, leading to verticality effects. Quantities and duration of seasonal snow cover also increase with elevation. A further control is the ‘carapace’ of refrozen meltwater, snow and icy layers superimposed on ablation zone ice in the fall and spring shoulder seasons. It also migrates vertically. The interactions of elevation and seasonal snow cover, debris mantles and ice surface topography are keys to specific and net annual ablation losses. When snow and the icy carapace are removed, when debris covers are thin or absent and when weather conditions are favourable, ablation rates are high and similar over a broad range of elevations. Hence, ablation season length and differences in surface conditions are critical for specific and total losses. On lower ablation zone areas, heavy debris mantles protect ice against the longer ablation seasons and higher air temperatures. However, areas of clean-to-dusty or dirty ice, thin or scattered debris, are more extensive and much more critical for net and total ablation losses and water yields. Thinner debris covers, dust, dirt, cold-tolerant algae and so-called cryoconite fragments can drive ablation rates much higher than for clean ice, in places more than double. These are found predominantly in the mid- and upper ablation zones. They are more sensitive to weather and climatic variability than where there are heavy debris mantles. However, geomorphological events in glacier basins, notably rock falls and massive rock slope failures, can suddenly alter and increase the protective role of debris covers for some decades as they are transported down the glacier. They are recurrent phenomena in the Karakoram which, along with motion instabilities and glacier surges, may alter mass balance or glacier advance and retreat.

The two parts of mass balance, inputs and outputs, are now considered together. Budget estimates are provided for Biafo Glacier and from partial data for Batura, Baltoro and some others. Overall quantities reflect the greater moisture availability and mass inputs at high elevations in the Karakoram. From glacier basins of the Central Karakoram, the total annual water yields appear to be of the order of 700–900 mm; less than maximum snowfalls measured in glacier source zones but three or more times greater than precipitation reported from valley weather stations. The mass balance regime is characterised as an all-year accumulation and summer ablation type, a distinctive form of Sub- or Outer Tropics regime. The vast areas of rock wall in source zones and predominance of avalanche nourishment and the significant ‘inputs’ in conventional ablation zones create problems largely ignored in mass balance assessments anywhere. Wind-fed and wind-stripped areas high up and disconnected tributaries above the snowlines add to the unique problems of input estimation. Wind redistribution and avalanches generate complicated patterns not clearly separated into accumulation and ablation zones or according to linear change with elevation. Mass balance gradients depart widely from those described elsewhere and point to distinctive constraints. An S-shaped vertical profile arises even in Alpine-type Karakoram glaciers reflecting reduced losses under debris-covered ice low down and input concentration where maximum snowfall occurs and especially through avalanche deposition. Further complications arise in profiles of Turkestan-type glaciers, where all positive as well as negative components for main ice masses are located in the conventional ablation zones. For such reasons, the notions widely employed in mass balance work are compromised. Equilibrium line altitudes (ELAs), thresholds between positive and negative mass balance, mostly occur below where firn limits or climatic snowlines are expected. These are, however, among the distinguishing features of the region that arise from the regional climatic regime, the exceptional extent of high elevations and verticality relations.

The geomorphology of Karakoram glacier basins involves high mountain terrain and surface processes responding to exceptional gravitational forces. Relations between high elevations, steepness, great relief, strong seasonal variations and their vertical migration help to organise the landscape. There are distinct sets of landforms in different elevation bands and zones of transition between them. An approach is developed in terms of landsystems, highlighting assemblages of processes and forms in the various zones. The analysis moves broadly downwards, beginning in glacier source zones where perennial sub-zero or frigid conditions prevail. Rugged interfluves and rock walls are picked out by wind- or avalanche-generated snow forms. Primary erosion of rock is shown to be fierce and former notions of the survival of ancient erosion surfaces in accordant summits unlikely. Below is a second transitional zone where the most extensive glacier surfaces form. Main ice streams are initiated by the run-out and deposits of avalanches, by icefall activity and, to a more limited extent, in broad accumulation zones. On-ice ablation zone landscapes are treated as a third system responding especially to seasonal rhythms. The neglected shoulder seasons are underlined, when freeze–thaw and wind action modify glacier surfaces. In the better-known summer conditions, the varieties of debris-covered ice assume great importance, including medial and lateral moraines, cryoconite forms, facets, on-ice ponds and streams. An off-ice periglacial zone compliments the ablation zone. Here, seasonal conditions, especially freeze–thaw, are again key to the scope and timing of surface processes. Near and beyond glacier margins, there is permafrost with a seasonal active layer and a wide variety of talus forms, hummocky and patterned ground. Though small in relative area, the lateral and terminal margins of the glaciers, with complex sediment assemblages and valley-side troughs have received the fullest attention of researchers. They are treated as a further landsystem. Effects of the cold season advance of termini are noted. There are distinctive developments related to high debris loads from the glaciers and surrounding slopes. Most larger glaciers have substantial areas of pro-glacial outwash sediments, mainly controlled by highly concentrated meltwater yields in a few weeks or months of summer. These, and other processes affecting the glaciers or their margins although mainly operating beyond them, introduce compound or transglacial landforms. They are important links in a web of processes and developments that serve to configure landforms and cryosphere relations to climate and tectonics.

The Karakoram is one of a relatively few mountain regions with high numbers of surge-type glaciers, subject to sudden, short-lived accelerations that transfer large volumes of ice down glacier. Some 55 surges have been identified since the 1860s involving 46 glaciers. Various studies suggest many more are surge type, perhaps one-third of Karakoram valley glaciers. Existing observations are reviewed and how they compare with those in other, better-known regions. The glaciers involved are predominantly or wholly avalanche-fed; occur in the highest, steepest parts of the Karakoram; and have large elevation spans. In recent decades, many surge-type tributaries have also been identified; 20 out of 33 events since the 1960s and nearly half (22) of all surges are recognised. In the past, most would have been missed, their numbers underestimated. Surges have impacts out of proportion to their short duration, affecting glacier morphology, surface features, hydrology, erosion and deposition. They are explained by instabilities at the bed of the glacier. Timing and recurrence intervals are peculiar to each case and have little or no relation to climate change or fluctuations in adjacent glaciers. They create huge mass balance anomalies and compromise the use of glacier advances or retreats as climatic signals. They pose a range of hazards for nearby communities.

Countless small and many larger lakes exist in Karakoram glacier basins. The region has a long history of outburst floods from them. Impoundments may be on the glacier, beside, in front of and even beneath the ice. The actual dams may be formed of ice, moraines or a combination of the two. Smaller glacial lake outburst floods (GLOFs) seem to occur somewhere in every year, the commonest glacier hazard. More rarely, large lakes occur and threaten much greater damage. Most have involved ice dams of a single type, where a substantial tributary glacier advances across and impounds a main river valley. Over the last two centuries, more than 100 glaciers of over 10 km in length have interfered with upper Indus and Yarkand streams. Large reservoirs have only been definitely identified with 23 glaciers, but other evidence shows many more at some time in the past. Large Karakoram ice dams develop quickly and rarely last more than a few months. The most dangerous cases, in particular at Chong Khumdan and Kyagar Glaciers, have involved two or more major outburst floods in episodes lasting several years. These GLOF hazards differ from smaller ones in the Karakoram and those receiving attention recently in the rest of the Himalaya. Specifically they require glacier advances. In the past decade or so, some of the glaciers associated with large ice dams have advanced and caused, or threatened to cause, GLOFs.

This chapter looks at fluctuations in Karakoram glaciers, mainly in the last 200 years, their consequences and implications for future responses to climate change. Most of the evidence available concerns terminus changes. Improvements in satellite coverage and analytical techniques have increased the range and quality of information but results can raise as many problems as answers. Records are reviewed back to the mid-nineteenth century for some of the larger and more frequently visited glaciers, including Baltoro, Biafo, Batura, Chogo Lungma and Hispar. This information broadly confirms an ice cover decline since the Little Ice Age (LIA), although it does not seem to exceed 5 % of the greatest LIA extent. Large glaciers with high elevation watersheds appear less reduced than small and minor ice masses, although evidence from the latter is very limited. The timing of the greatest advances in the LIA varies by decades for different glaciers, in some cases by centuries. Between the 1920s and 1980s, most of the larger glaciers had a net retreat, but an almost chaotic situation emerges from the late LIA through the early twentieth century, and again since the mid-1990s. This is only partly accounted for by unsystematic fluctuations in surge-type glaciers. There has been no great loss of ice, and more than 40 high Karakoram glaciers have undergone advances of varying extent. At the latest time frame to 2010, few if any glaciers were at their most advanced positions of the past 150 years but also, no case was at its greatest reported retreat. Advances and retreats have been more or less out of phase even in neighbouring glaciers. Few of the large glaciers have retreated in proportion to the amount of thinning, and some have even advanced while appearing to thin. The recent picture differs from other parts of the Himalaya and common global trends. This may be due to the distinctive climatic regime or how it responds to global climate change as discussed in earlier chapters. Out-of-phase relations of terminus fluctuations may also follow from different styles of nourishment, thermal regimes and shifts causing movement instability. As yet unrecognised surge-type glaciers are another source of complexity. Confusion has also arisen in relating Quaternary glaciations to the state of present-day glaciers, notably debris covers, and ice margin and pro-glacial deposits near them. This is illustrated by former and emerging interpretations of the so-called Great Lateral Moraine, the nearly ubiquitous, relatively well-preserved and massive lateral margin deposits overlooking present ice levels. They create an impression of glacier decline that may be misleading. Originally viewed as equivalent to the ‘1850’ Little Ice Age moraines of the European Alps, the deposits turn out to be more diverse in age and origin. Post-glacial geomorphic developments along the Indus streams are at least as important as climate change and neoglaciation.

The following three main concerns are addressed: (1) mountain peoples closely involved with glaciers, (2) down-country relations in Pakistan, China, India and Afghanistan, especially for water resources and risks, and (3) trans-boundary questions. In the latter, glacierised areas are involved in resource and geostrategic agendas, cultural and religious questions between countries and, not least, recurring and ongoing armed conflicts. There has been a tendency, beginning in the earliest modern work, to detach glaciers and glacial science from human contexts. Much has been investigated in both areas but their interrelations largely ignored, in particular how scientific glacier knowledge might engage with local knowledge for mutual benefit and to help address mountain land concerns. This bears on problems of some urgency in adapting to glacier and other climate-related changes. Lately, there has been the so-called ‘melting Himalayas’ issue, possible water crises due to loss of ice through global climate warming. It is a justifiable concern if, to date, not as evident in Karakoram glaciers or clear in its implications as for some other High Asian mountains. A brief survey is given of how mountain communities relate to and view the glaciers, mainly their agricultural and pastoral activities, and how modernization is affecting them. A case study of the Hopar villages, Barpu and Bualtar Glaciers, illustrates something of the complexities and the risks from glacier hazards. Interest in the water resources of the Indus and Yarkand Basin is overwhelmingly about supplies for the larger populations, cities and industries of the lowlands. The two rivers, specifically their main stems, may well have the world’s highest ratios of glacier meltwaters to numbers of inhabitants dependent on them. How glaciers relate to other cryosphere elements in terms of water resources, notably to snowfall and permafrost, remains largely unknown. How they relate to rapid socio-economic changes and large-scale water resource projects is even murkier. Many developments are necessarily poorly informed by knowledge of the glaciers because monitoring and research are so limited. Their present state and potentially adverse impacts of regional temperatures and other changes are equally uncertain. Meanwhile, the glacial waters of the Indus flow through four countries and, in the Yarkand Basin, China has issues with the indigenous cultures of Tibet and Xinjiang Province. Future prospects seem likely to depend more on relations between Pakistan, India, Afghanistan and China than on glacier change. The Indus Waters Treaty is generally considered a model for resolving differences and avoiding conflict over trans-boundary water resources. Nevertheless, for decades the upper Indus Basin has been subject to armed intervention and recurring wars. Relations between the countries concerned and the Treaty are under considerable stress these days. Of special concern is how conflict has blocked and disrupted scientific work on the glaciers and continues to do so, not least the much-needed improvements in monitoring the cryosphere and responding to glacier change.