Long before anyone kept written records, the earth was busy keeping its own. Glaciers advanced, seas rose and fell, winds carried fine dust across entire continents, and rivers carved their way down valley after valley. None of this needed a witness. It left behind rocks, sediments, and landforms that geologists and archaeologists still read today to reconstruct the Pleistocene ice ages. Four types of evidence in particular, moraines, sea level changes, loess, and river terraces, form the backbone of how we know glaciers once reached far beyond where ice exists now.
Table of Contents
Moraines: the glacier’s calling card
When a glacier moves, it does not travel lightly. It drags, scrapes, and plucks rock from the land beneath and around it, grinding some of it into fine powder and leaving the rest as jagged, angular chunks. This mixture, unsorted and ranging from clay-sized particles to boulders, is called till. When a glacier eventually melts or retreats, it dumps this load in distinct patterns known as moraines, and their shape and position tell geologists exactly where the ice used to be.
Types of moraines
Moraines are classified by where they sit in relation to the glacier that made them:
Ground moraine forms an uneven blanket of till spread across the valley floor once the ice retreats.
Lateral moraine builds up along the sides of a glacier, made from rock debris that falls from the valley walls.
Medial moraine appears when two glaciers merge and their lateral moraines join, leaving a dark ridge running down the centre of the combined ice stream.
End or terminal moraine marks the farthest point a glacier reached before it began to retreat.
The formation of a medial moraine in particular is treated as strong proof that glaciers actually flow, since two separate ridges of debris could not merge into one line down the middle of the ice unless both masses were moving.
Other telltale signs
Moraines rarely appear alone. Glaciated valleys tend to show a U-shaped cross-section instead of the V-shape carved by rivers, along with smaller hanging valleys left where a tributary glacier’s floor sat higher than the main valley it fed into. Boulders that do not match the local bedrock, called erratics, are another strong clue. These rocks were picked up somewhere entirely different and carried long distances before being dropped as the ice melted, so finding one sitting on unrelated rock is a direct sign that ice once passed through.
When the sea went missing: sea level as a climate record
Ice sheets do not appear out of nowhere. The water that builds them has to come from somewhere, and that somewhere is the ocean. During glacial periods, enormous volumes of seawater got locked up as ice on land, and global sea level dropped as a result. At the peak of the last glaciation, sea levels are estimated to have been around 100 metres lower than they are today, enough to expose entire stretches of continental shelf as dry land.
Signs of a shrinking sea
Evidence of these lower sea level stands survives in a few recognisable forms. Submerged organic remains, such as pollen or the bones of land animals, are sometimes recovered from areas that are now well underwater, confirming that the spot was dry land not long ago. Deltas built by rivers at the old, lower coastline now sit submerged far offshore. Perhaps the clearest sign is anomalous faunal distribution, meaning related species of plants or animals turning up on opposite sides of a body of water that should have kept them apart. The Bering Land Bridge is the textbook case: as sea level fell, the shallow seafloor between Siberia and Alaska turned into passable land, and the connection between the two continents lasted for roughly 80 percent of the last million years, allowing species and eventually people to cross on foot.
Signs of a rising sea
The pattern reverses during interglacial phases. As ice sheets melt, the water returns to the oceans and sea level climbs, at times higher than today’s level. This leaves fossil-bearing marine sediments stranded well above the current shoreline, along with cliffs and wave-cut platforms shaped by waves that no longer reach that high. Reading a coastline’s layered history of submerged and elevated features lets researchers reconstruct a fairly detailed timeline of glacial and interglacial swings, even in places with no other datable material.
Loess: dust with a story
Not every clue left by an ice age is a boulder or a beach. Some of it is dust. Loess is a fine, wind-blown sediment, coarser than clay but finer than sand, that settles in a periglacial zone, the cold, largely treeless region bordering an ice sheet or glacier. Glacial grinding produces huge quantities of this fine rock flour, and strong, dry winds sweeping off the ice pick it up and deposit it across nearby lowlands, sometimes building layers many metres thick over repeated glacial cycles.
Because loess needs a nearby glacier to supply the raw material, its presence in a region is treated as indirect but reliable evidence that ice once sat close by, even if no moraine survives at the same spot. Europe’s Danube valley holds one of the thickest and best-studied loess belts on the continent, largely because the river carried glacially ground silt down from the Alps and surrounding mountain ranges during successive glacial periods.
Closer to home, the Kashmir valley preserves a similar record in its Karewa formations, thick lake and river sediments capped by loess. These deposits, locally known as Kaerwas, form the terrace-like plateaus that dominate much of the valley’s landscape today. Researchers studying this sequence have found that the loess layers alternate with buried soils, recording repeated swings between cold, dry glacial phases and warmer, wetter interglacial ones. Each buried soil marks a pause in dust accumulation when the climate turned mild enough for vegetation to take hold, only for wind-blown loess to bury it again as the next cold phase set in.
River terraces: steps left by a restless river
A river rarely sticks to one level for long. Over thousands of years, as climate shifts, a stream may cut down into its old floodplain and settle at a lower elevation, leaving the earlier surface stranded above the new channel as a flat bench along the valley side. These abandoned surfaces, called river terraces, form a kind of staircase up the valley wall, and each step corresponds to a former floodplain.
What drives a river to change levels so dramatically comes down to the balance between aggradation, when a river deposits more sediment than it removes and its bed rises, and incision, when it cuts down instead. Glacial and interglacial periods tend to push this balance in opposite directions. During a glacial phase, increased sediment supply and reduced water flow often cause a valley floor to fill in and rise. As the climate warms and glaciers melt, higher water discharge triggers renewed down-cutting, stranding the older, higher surface as a terrace.
This connection between climate and terrace formation has a long history in South Asian archaeology. Working in the Sohan valley of the Potwar region in the 1930s, Helmut de Terra and T.T. Paterson proposed one of the earliest models linking fluvial terrace sequences to a fourfold cycle of glacial and interglacial periods, an idea that shaped decades of Palaeolithic research across the Indian subcontinent. While later scholars have questioned the details of their original stratigraphy, the broader principle still holds: terrace sequences preserve an indirect record of past climate swings, and the stone tools and fossils buried within specific terrace layers can often be placed in a rough chronological order using this framework.
What do you think?
What do you think? Of the four types of evidence covered here, moraines, sea level shifts, loess, and river terraces, which do you think would be hardest to date accurately, and why? And if you were surveying a valley for signs of a past ice age but found no moraines at all, which of the remaining clues would you look for first?
References
- https://www.britannica.com/science/moraine
- https://www.antarcticglaciers.org/glacial-geology/glacial-landforms/glacial-depositional-landforms/moraine-types/
- https://education.nationalgeographic.org/resource/continental-shelf/
- https://www.nps.gov/articles/aps-v12-i2-c8.htm
- https://sciencedirect.com/science/article/abs/pii/S0012825215300076
- https://www.frontiersin.org/journals/earth-science/articles/10.3389/feart.2020.00113/full
- https://ebooks.inflibnet.ac.in/antp13/chapter/archaeological-river-terraces-stratigraphy-sediments/
Leave a Reply