[Download Kenneth Hewitt’s full report here]
Glaciers and their immediate environs present many dangers for humans, such as crevasses and glacier mills into which one might fall, heavily crevassed ice falls, snow and ice avalanches from the side walls and, along the flanks, dumping of great boulders, ponding and floods from melt water. For these reasons, there are hardly ever permanent settlements on or right beside the ice. These are hazards mainly to mountaineers, hunters, travellers and military expeditions. The more serious dangers arise from processes in the glacial environment that may extend their impacts beyond existing glacial areas. The more serious tend to involve ponding of water that leads to glacial outburst floods, or releases that generate debris flows.
Heavily debris-covered ice, Panmah Glacier Central Karakoram, around 4,000 metres above sea level. Note that even the heaviest debris on active ice is rarely more than 2 metres thick. The relief of mounds and cones is almost entirely ice cored and the debris is constantly shifting around. (Photograph/Kenneth Hewitt, June 2009)
The risk of glacier lake outburst floods has received particular attention in other parts of the Himalaya, notably Bhutan, Nepal and Tibet. In Nepal, some 25 glacial lake outburst floods have been recorded since the 1930s, with especially destructive events in 1985 and 1991. Bhutan also has a number of dangerous lakes, one of which burst with disastrous consequences in 1994. Reports suggest all of these lakes and the triggers for outburst floods are related to climate warming and glacier retreat. There is also a history of such outburst floods from Karakoram glaciers. However, the problem here is also very different from that recently reported elsewhere in the Himalayas. In particular, the most serious threats involve, specifically, much larger impoundments by short-lived, unstable ice dams. Crucially, all recorded examples have been associated with advancing glaciers.
In fact, the Karakoram presents two rather different groups of outburst floods. The most frequent are relatively local events. Collectively, they threaten dozens if not hundreds of small settlements in the higher valleys and examples occur in most years. They involve a wide variety of dam compositions, forms and outburst types, including ice, moraine, and mixed-barriers. Conversion of outburst floods into debris flows is quite common, usually the more severe risk. For the upper Indus, these are the only types of damaging outburst floods reported in the past several decades. Moreover, they occur whether glaciers are advancing, retreating and relatively stable. Conversely, the larger Karakoram dams involve impoundment of a main river valley by a relatively large tributary glacier. Most important, in the present context, these dams only form from a vigorous forward push of the ice.
A series of ice margin lakes along Nobonde Sobande arm of Panmah Glacier, central Karakoram seen from Drenmang (4,500 metres above sea level). Some are behind old lateral moraines, others ponded against the edges of active ice. The glacier is about two kilometres wide here and 10 kilometres of the main ice stream are visible. (Photograph/Kenneth Hewitt 1994)
More than 60 glaciers of intermediate-to-large size (10 kilometres to 65 kilometres in length) have a history of advancing into and interfering with tributaries of the upper Indus and Yarkand rivers. Not all are known to have created actual dams, but at least 30 have done so and involved outburst floods of exceptional size and destructiveness. However, while there have been several large dams recently on the Shaksgam, on the Indus the last major ice dam was in 1933. “Major” refers to outburst floods that were large enough to register hundreds of kilometres downstream at the river gauge at Attock, where the river leaves the mountains.
The most urgent questions today involve some Karakoram valleys whose glaciers created ice dams and catastrophic outburst floods in the past and that are advancing right now. Will they impound the rivers again? Three locations require special attention; the Shaksgam, upper Shyok and Shimshal valleys.
The Shaksgam is a tributary of the upper Yarkand. According to satellite imagery, five glaciers that have formed ice dams in the past are advancing at present. One of them, the Kyagar, has created several recent dams. An outburst from the one in 1999 caused severe damages along the lower Yarkand River in Kashgar district. In the summer of 2009, Kyagar again impounded the river and a 3.5 kilometre-long lake was formed. Fortunately it drained slowly but was close to dimensions that have led to disastrous floods in the past. There were great difficulties in obtaining satellite coverage and scientists were unable to visit the site and monitor the lake so as to predict its behaviour. This raises serious issues about what would have happened if a large outburst had occurred, and what will happen in future cases. It seems a new impoundment will form at Kyagar in 2010, and the four other glaciers are across or entering the river and may impound it.
The terminus of Yazghil Glacier, north-west Karakoram, where it enters the Shimshal River. This is one of several Karakoram glaciers on the upper Indus and upper Yarkand Rivers that have caused ice dams and glacier outburst floods in the past, and are presently advancing across the rivers. (Photograph/Kenneth Hewitt July, 1998)
On the Indus, three glaciers in Shimshal and three on the upper Shyok, that have formed ice dams in the past, began advancing about a decade ago. They have not yet reached positions where a dam could form, but could do so quite soon. Historically, the most dangerous have been the Chong Khumdan and Kitchik Khumdan on the Shyok. In 2009, satellite imagery revealed a sudden and large increase in thickness of the Chong Khumdan, and advance of its terminus into the river. Between 1926 and 1932, this glacier formed a series of large ice dams. At least four outburst floods were reported that caused appreciable rises in the river 1,100 kilometres away at Attock.
The 1929 event was the largest on record, and did great damage throughout the mountains and to the Indus Plains. The lake reached over 15 kilometres in length but drained in less than 24 hours. The Kitchik Khumdan also formed large ice dams in the nineteenth century, and its terminus is back in the river and has advanced across the river which passes beneath the ice. However, 2009 satellite imagery suggests it is beginning to waste back again. Conversely, its immediate neighbour the Aqtash Glacier which has also formed dams in the past advanced across the river in 2008 and 2009 and seems to be advancing very rapidly.
These glaciers highlight problems of security and the legacies of conflicts that exist in many parts of High Asia. They are in a militarised zone disputed by China, India and Pakistan. Apparently the Khumdan glaciers fall under the control of Chinese forces, but the dangers from the outburst floods are primarily in Indian and, especially, Pakistan-controlled areas. Given existing tensions, including the India-Pakistan “war” on the Siachen Glacier nearby, it is unclear how necessary studies, monitoring and warning systems can be set up.
Other hazardous phenomena
The focus here has been on glaciers, but it needs emphasising there is a range of cold climate or cryosphere phenomena that may become hazardous through climate change. Communities, infrastructure and related activities confront changes in snowfall, snow-on-the-ground and permafrost, specifically ground ice. They will also be affected by changes in distribution and intensities of freeze-thaw, the quantities and timing of surface and ground waters and their quality (water temperatures, turbidity and dissolved matter, for instance).
The accumulation zone of Biafo Glacier near Hispar Pass (5,150 metres), showing the development of cornices along ridge lines due to wind action, avalanched steep walls and heavy build up of snow on gentler slopes. (Photograph/Kenneth Hewitt, June 1999)
The entire mountain area is covered by seasonal snowfall, varying in duration and depth with elevation. Its melting provides about half of stream flows in an average year. Permafrost – perennially frozen ground – at intermediate altitudes is much more extensive than glaciers and includes hundreds of ice-cored rock glaciers. Freeze-thaw cycles affect even larger areas, as do erosion and deposition forms created by snow avalanches. All of these are affected by climate change. Their responses interact physically, and in ways that modify the scope or significance of glacier-related risks.
High elevation conditions on Karakoram glaciers: rockwalls, ice falls, avalanches of Broad Peak (8,050 metres), part of the watershed of Baltoro Glacier. (Photograph/Kenneth Hewitt, 2005)
Retreating glaciers and warming permafrost are associated with destabilised slopes. They can lead directly to landslides, or reduce the strength thresholds for, and the likelihood or size of, slope failures due to earthquakes or storms, which trigger most of the more destructive landslide events. For example, a dangerous landslide occurred on January 4, which blocked the Hunza River in the central Karakoram and probably involved destabilisation by changing moisture and temperature conditions in the slopes. The lake has already grown to 5.5 kilometres in length, forcing the evacuation of thousands of residents. Moreover, the lake behind a similar landslide dam in 1858, immediately upstream of the present one, lasted seven months then burst with catastrophic effects all the way to the Indus plains. Meanwhile, slopes exposed by reduced ice or snow cover may dry out and become useless. Conversely, some may also become vegetated and economically useful for timber, firewood, for pastoralists and even for cultivation.
The more immediate glacier hazards and response needs in the region involve communities and activities in the high mountains. Only the Andean highlands rival inner Asia in the numbers and diversity of settlements close to and at direct risk from glacier change. However, for the broader national and international contexts, the major issues raised concern water resources and their reliability.
Some caution is needed here. A commonplace of recent reports is to say that the lives and livelihoods of in excess of 1.5 billion people are critically dependent upon the glaciers in the headwaters of the largest Asian rivers. This is a misleading generalisation. Yes, such are the numbers of people living in river basins with tributaries coming from glacierised mountains. However, in most cases the glaciers are a tiny part of the river flows, notably in the most heavily populated areas of China, India and the south-east Asian mainland.
Snowfall affects much vaster areas than the glacier cover, and is more critical. For the vast majority of these populations, rainfall and ground waters are far more important than snowfall. Glacier change can have impacts on these other parts of the hydrological cycle or may compound changes in them, but the processes are mostly indirect and too poorly known to make such generalisations. Whether and how far there are significant risks for most of these populations, even from the “disappearing” glaciers’ scenario, is far from certain.
The Indus and Yarkand basins do involve large populations directly, or potentially, dependent on the glaciers. Even here, however, there have been exaggerated or misleading claims. Yes, glacier melt waters comprise more than one-third of the flow of the main stem of the Indus, snow and ice together providing over two-thirds. It has the largest ratio of melt water to population of any river, anywhere in the world. At the moment, however, nearly all the glacial melt water goes to the sea. It happens to coincide with the heavy monsoonal rains, making flooding the greater problem, and Pakistan lacks the capacity to store much or any of the melt waters at that time.
More exactly, the key roles of glacier melt waters have little to do with the total size of the ice cover, total melt water yields, or trends. Rather they turn upon demand in just a few weeks of the year and, in rare, extreme cases when the winter rains or monsoon are very weak, poorly timed, or fail. Even for Pakistan, the main dangers for the country as a whole are, therefore, potential rather than actual, and not so much in relation to glacier change as to planned and possible water resource developments. These seem to be being undertaken with inadequate understanding and assessment of how climate and glacier fluctuations will affect them.
This will become increasingly acute for all countries of the region and raise important transboundary concerns. There are the huge commitments being made now, to hydroelectric power, irrigation, urbanisation and other developments for which water from snow and ice will become increasingly crucial. More than 100 existing dams depend partly on glacial melt waters. Several hundred more, and some of great size, are under construction or planned for China, India, Pakistan, Nepal and Bhutan.
Given the present state of monitoring and scientific understanding, it is hard to believe any of these have adequate or accurate assessments of climate- and glacier-change impacts. For the Karakoram it is of singular concern to determine whether, as global warming continues, there will be a return to glacier retreat as some believe, or if the factors responsible for the present advances will intensify. Either way, there are serious implications for how communities in Pakistan, China and India, especially, are affected and need to respond.
The importance of climate change is not in doubt, but research and policies should be based on actual evidence. Where unavailable, that should be acknowledged, not – as has happened with glacier change in the Karakoram – simply replaced by supposition based on developments or models from elsewhere. Much of what is being said fails to recognise the patchiness of past research in space and time, and a nearly-total absence of glacier monitoring at elevations where the most critical ice and climate changes occur.
The limited evidence surely reflects, in part, the sheer scale, diversity and logistical difficulties of scientific work in much of the region. Now, as more resources become available to investigate these problems, it is important to identify what sorts of information are needed, where and how they can be best obtained. Science and information systems and regional cooperation need to address the complexity and diversity of the greater Himalayan region. Some practical suggestions being promoted by new programmes include the following:
* To set up improved monitoring systems that combine remotely sensed and automatic station measurements with ground control related to basic glaciological and hydrological research;
* To expand comprehensive, multi-disciplinary research that addresses environmental and cultural complexities in the region;
* To pursue regional cooperation in data sharing, risk and resource assessments; and
* To actively involve local communities in the mountains, so that their ecological knowledge and practical concerns inform understanding and help to shape appropriate development.
Kenneth Hewitt is professor emeritus in geography and environmental studies and research associate at the Cold Regions Research Centre at Wilfrid Laurier University in Ontario, Canada.
Homepage image by Kenneth Hewitt shows icefalls descending to the main glacier at Kaberi-Kondus Glacier, east-central Karakoram in 1998