Earth's surface: land, water, ice and erosion: Ice: glaciers, ice ages and the Himalayan retreat
All courses / Understanding Science / Module 19: Earth's surface: land, water, ice and erosion Part 4 of 6Ice: glaciers, ice ages and the Himalayan retreat
Ice is the slowest agent that shapes the land, and one of the most forceful. This part explains how glaciers move and carve, how an ice age was proved and timed, and what is happening to the ice of the Himalaya.
4.1 A glacier is flowing ice, and it carves valleys that no stream could
A glacier begins as snow that outlasts the summer. Year on year the layers thicken, and the weight above packs the buried snow into grainy firn and then into dense, interlocking ice. When the pile is thick enough, the ice deforms and creeps downhill under gravity, and also slides on a thin film of meltwater at the bed. This is a glacier. Valley glaciers flow down mountain valleys, and ice sheets cover whole regions, as in Greenland and Antarctica. About two thirds of the freshwater on Earth is held as ice, and the Antarctic ice sheet alone would raise sea level by tens of metres if it melted.
A glacier has an upper zone where it gains mass and a lower zone where it loses mass by melting. Its end sits where the two balance. If snowfall grows or summers cool, the end advances. If not, it retreats. Retreat does not mean the ice flows uphill.
Moving ice is a powerful tool because it carries rock. Boulders frozen into the base scrape and polish the bed, leaving parallel scratches, called striations. The glacier plucks blocks from the rock. High bowls cut at the head of a valley are cirques, and where cirques meet, sharp ridges (arêtes) and pointed peaks (horns) remain. A river valley is V-shaped, but a glacier leaves a U shape with a wide, flat floor and steep sides, and tributary valleys left hanging. If the sea floods a glacial valley it becomes a fjord.
What a glacier drops is just as telling. Its deposit, till, is unsorted, clay and boulders together, because ice does not sort by size as water does. Ridges of till at its sides and snout are moraines. Boulders carried far from their source and left stranded are erratics. Meltwater lays sorted sand and gravel in ridges (eskers).
- Hewitt, Lyons, Suchocki and Yeh, Conceptual Integrated Science, 2nd edition (Pearson, 2013) · ch. 24, 'Earth's Surface: Land and Water'
- Hewitt, Lyons, Suchocki and Yeh, Conceptual Integrated Science, 2nd edition (Pearson, 2013) · ch. 25, 'Surface Processes'
- US Geological Survey, Water Science School, How much water is there on Earth? · web
4.2 Louis Agassiz argued in 1837 that ice once covered Europe, using boulders that water could not have moved
In the early nineteenth century many geologists explained scattered boulders and gravel across northern Europe as the work of Noah's flood or of floating icebergs. Swiss naturalists thought otherwise. Ignaz Venetz, Jean de Charpentier and Karl Schimper noticed that mountain glaciers carry boulders and drop them in moraines, and that Alpine valleys far below the present ice contained the same marks.
Louis Agassiz, a young professor at Neuchatel, took the idea further. In an address to the Swiss Society of Natural Sciences in 1837 he argued that a vast cap of ice had once covered much of Europe. His case rested on four kinds of evidence. Erratics of Alpine granite lay on the Jura Mountains, far above the valley between, where no flood could have lifted them. Rock surfaces beneath the debris were polished and scratched in parallel lines in the direction of old ice flow. Ridges of unsorted rubble matched the moraines at the ends of modern glaciers. A flood sorts material by size and does not scratch bedrock in a straight line. Ice dumps everything together and scores the rock.
The idea was resisted at first, then confirmed by fieldwork in Britain, North America and elsewhere. Today we know of a series of cold periods over the last 2.6 million years, with the most recent peak, the last glacial maximum, about 20,000 years ago. Ice sheets then covered much of Canada and northern Europe, and sea level stood about 120 metres lower than now.
- Hewitt, Lyons, Suchocki and Yeh, Conceptual Integrated Science, 2nd edition (Pearson, 2013) · ch. 25, 'Surface Processes'
- Agassiz's 1837 Neuchatel address, summary by the Debating Glacial Theory site · web
- NASA Goddard Institute for Space Studies, research brief by Vivien Gornitz (2007), Sea level rise, after the ice melted and today · web
4.3 The ice ages keep time with changes in Earth's orbit, and mud on the sea floor recorded it
The ice ages came and went in a regular beat, which needs a clock. The clock is Earth's orbit. The Serbian scientist Milutin Milankovitch worked out in the early twentieth century that three slow cycles change how sunlight is spread over the planet.
The shape of Earth's orbit stretches and relaxes between near-circular and slightly oval over about 100,000 years: this is eccentricity. The tilt of the axis, now about 23.4 degrees, swings between 22.1 and 24.5 degrees over about 41,000 years. More tilt makes sharper seasons. And Earth's axis wobbles like a spinning top, so that the season in which Earth is closest to the Sun shifts. That precession has an average cycle of about 23,000 years. None of these changes the total sunlight much. They change where and when it falls. Ice sheets grow when northern summers are cool, because snow that survives the summer builds up year after year.
The test came from the ocean floor. In quiet parts of the deep sea, mud settles at a few centimetres every thousand years, and carries tiny shells of foraminifera. The shells hold oxygen in two forms. Ice sheets take up the lighter form from sea water, so in a cold period the ocean, and the shells, hold relatively more of the heavy form. A drill core therefore gives a record of ice volume through time. In 1976 James Hays, John Imbrie and Nicholas Shackleton analysed two such cores from the southern Indian Ocean, covering 450,000 years, and found rhythms at about 100,000, 42,000 and 23,000 years, matching the orbital cycles. Orbit is the pacemaker, and feedbacks, particularly carbon dioxide and the reflection of sunlight by ice, amplify it. Module 20 treats those feedbacks.
| Quantity | Value | Working |
|---|---|---|
| Length of record needed | 450,000 years | Span studied in 1976 |
| Accumulation rate | More than 3 cm per thousand years | Hays and colleagues; take 3 cm |
| Core length | About 13.5 m (calculated) | 450 x 3 cm = 1,350 cm |
| Eccentricity cycles in the record | About 4.5 (calculated) | 450,000 / 100,000 |
| Tilt cycles in the record | About 11 (calculated) | 450,000 / 41,000 |
| Precession cycles in the record | About 20 (calculated) | 450,000 / 23,000 |
A core of this length fits in a research ship's drilling gear. A few cm of mud for each thousand years is slow, and that is why the record is long.
Source: NASA, Milankovitch (Orbital) Cycles and Their Role in Earth's Climate, https://science.nasa.gov/earth/climate-change/milankovitch-orbital-cycles-and-their-role-in-earths-climate/ (eccentricity about 100,000 years; obliquity 22.1 to 24.5 degrees over about 41,000 years; precession about 23,000 years on average; Hays and colleagues' deep-sea cores in Science, 1976, matched the cycles to climate change over 450,000 years); Hays, Imbrie and Shackleton, Variations in the Earth's Orbit: Pacemaker of the Ice Ages, Science 194, no. 4270 (10 December 1976), pp. 1121 to 1132, https://www2.atmos.umd.edu/~nigam/AOSC617/Hays.Imbrie.Shackelton.Orbital.theory.ice.age.Science.1976.pdf (cores RC11-120 and E49-18 from the southern Indian Ocean; the past 450,000 years; spectral peaks at 23,000, 42,000 and about 100,000 years; accumulation rates above 3 cm per 1,000 years)- Hewitt, Lyons, Suchocki and Yeh, Conceptual Integrated Science, 2nd edition (Pearson, 2013) · ch. 25, 'Surface Processes'
- NASA, Milankovitch (Orbital) Cycles and Their Role in Earth's Climate · web
- Hays, Imbrie and Shackleton, Variations in the Earth's Orbit: Pacemaker of the Ice Ages, Science 194, no. 4270 (10 December 1976), pp. 1121 to 1132 · web
4.4 The glaciers of the Hindu Kush Himalaya are losing ice faster, and a billion people depend on the water
The Himalaya and its neighbouring ranges hold more ice than any place outside the poles, which is why it is called the third pole. Its glaciers feed the rivers of the Indus, the Ganga and the Brahmaputra. In summer, meltwater adds to the monsoon, and in the dry months before the monsoon it makes up a larger share of river flow.
The International Centre for Integrated Mountain Development (ICIMOD) summarised the evidence in 2023. Glaciers in the Hindu Kush Himalaya lost ice 65 per cent faster in 2011 to 2020 than in the previous decade. On the current track of emissions, they could lose up to 80 per cent of their present volume by the end of the century. About 240 million people live in the mountains, and a further 1.65 billion live downstream.
The near-term effect is odd. As glaciers melt faster, river flow in the dry season first rises, and then, once the ice is mostly gone, it falls. Glacier lakes also grow behind the moraines left at the ends of retreating ice. A moraine dam can fail, releasing a glacial lake outburst flood, which is one hazard documented in Sikkim in October 2023.
A glacier is a balance of snow gained and ice lost. Warming raises the altitude of the rain-snow line, so more precipitation falls as rain, and it lengthens the melt season. Both shrink the zone of gain. The glacier then must retreat until its smaller area balances.
- Hewitt, Lyons, Suchocki and Yeh, Conceptual Integrated Science, 2nd edition (Pearson, 2013) · ch. 24, 'Earth's Surface: Land and Water'
- ICIMOD, Water, ice, society, and ecosystems in the Hindu Kush Himalaya, press release (Kathmandu, 20 June 2023) · web
See it in the table
H₂O WaterO2 OxygenCO2 Carbon dioxideCheck your understanding
U-shaped valleys, striations and erratics are evidence of
Moving ice scours a wide U-shaped valley, scratches bedrock and carries boulders far away. Water forms V-shaped valleys and sorts its load.Agassiz's 1837 argument differed from the flood explanation because it
He compared fields of boulders and scratched bedrock with what living glaciers leave. Floods do not scratch rock in parallel lines or heap unsorted rubble.The three Milankovitch cycles have periods of about
Eccentricity, tilt and precession give those periods, and deep-sea cores showed the same rhythms in ice volume.