The Norse colony in Greenland was founded by Erik the Red in 985 AD and lasted, in its modest fashion, for about four hundred and fifty years. At its peak in the early thirteenth century the colony had two main settlements with perhaps three thousand inhabitants, a cathedral at Garðar with imported stained glass and a bronze bell, sixteen churches, twelve monasteries, an annual shipping route to Bergen in Norway, and a small export economy based on walrus ivory and polar bear hides. By 1450, every Norse Greenlander was dead or had left. The last bishop of Garðar died in 1378 in Norway, never having seen his diocese. The last ship known to have called at Greenland reached the Eastern Settlement in 1410. After that, silence. When Hans Egede, a Danish missionary, sailed to Greenland in 1721 to find the lost colonies and reintroduce them to Christianity, he found ruins and Inuit settlements. He found no Norse.

What killed Greenland’s Vikings was not, as the Norse sagas would later have it, dragons or political intrigue. It was, in significant part, the weather.

Beginning sometime in the early fourteenth century — the dating is genuinely uncertain and varies by region — the climate of the North Atlantic became progressively colder. Sea ice expanded southward. The summer growing season in southern Greenland shrank from about four months to about two. Grass for the sheep and cattle the Norse depended on failed to ripen. The annual ship from Bergen could no longer reliably navigate the ice-choked Davis Strait. The colony lost first its prosperity, then its supplies, then its population. The same process — the same drift toward shorter summers, longer winters, and pack ice in unaccustomed waters — was happening across the entire northern hemisphere.

The cold lasted, with multiple sub-peaks and partial recoveries, until the early nineteenth century. It is now called the Little Ice Age. It is the single largest climate event of the past two thousand years.

When and where and by how much

The dates and the magnitudes are still debated by paleoclimatologists. The conservative consensus, based on the synthesis of ice core, tree ring, lake sediment, and historical documentary records by the PAGES 2k Consortium and others, is roughly as follows.

The Little Ice Age was a multi-century interval of below-average temperatures across most of the northern hemisphere, with the deepest cooling in the North Atlantic basin and in the Alps. Average temperatures in northern Europe ran roughly 0.6 to 0.8 degrees Celsius below the modern (twentieth-century) mean. In specific subperiods and specific regions the deviation was larger — winters in seventeenth-century England were about 1.3 degrees colder than the modern average, and particular peak-cold winters (1683-84, 1709, 1740-41) were 2 to 3 degrees below modern norms.

Within the multi-century span there were three particular deep-cold subperiods:

  • The Sporer Minimum (roughly 1450-1540), correlated with reduced solar activity and the abandonment of Norse Greenland
  • The Maunder Minimum (1645-1715), correlated with the near-total disappearance of sunspots and the coldest winters in the documented English meteorological record
  • The Dalton Minimum (1790-1830), correlated with the Year Without a Summer (1816) caused by the Tambora eruption

Between these deep periods the climate fluctuated. There were warmer decades in the early sixteenth century and again in the late seventeenth. The Little Ice Age was not uniformly cold; it was a cold envelope with internal variation. Modern attempts to give it a single start or end date are doomed to approximation.

In tropical regions the cooling was smaller or absent. In the southern hemisphere it was probably smaller and the timing was not synchronized. The Little Ice Age was, in physical terms, primarily a northern hemisphere phenomenon.

What probably caused it

Modern climatology has identified three plausible drivers, which probably acted in combination.

Solar variability. The seventeenth-century English astronomer Edward Walter Maunder, working at the Royal Greenwich Observatory in the 1890s, noticed that the historical sunspot record for the period 1645-1715 showed almost no sunspots at all. Fewer than fifty sunspots were recorded across the entire seventy-year period — compared with the modern average of about a thousand per year. The reduced solar activity correlates closely with the coldest interval of the Little Ice Age. The mechanism is small — reduced solar radiation accounts for a forcing of about 0.1 to 0.3 watts per square meter, comparable to a 0.1 degree Celsius global cooling — but it is real and well-documented.

Volcanic forcing. Large equatorial volcanic eruptions inject sulfate aerosols into the stratosphere, which reflect incoming sunlight and produce a multi-year cooling effect. The 1257 eruption of Mount Samalas in Indonesia (only identified as the source of a global 1258 cooling event in 2013, by a French-Indonesian team led by Franck Lavigne) was probably the trigger for the initial descent into the Little Ice Age. Subsequent large eruptions — Kuwae in the South Pacific in the 1450s, Huaynaputina in Peru in 1600, Laki in Iceland in 1783, and Tambora in 1815 — each produced multi-year cold spikes that, on the already-cooling baseline, made bad situations worse.

Ocean circulation feedbacks. The North Atlantic thermohaline circulation, which carries warm tropical water to northern Europe and Scandinavia, weakened during the Little Ice Age. The mechanism is debated. The most plausible reconstruction is that volcanic and solar forcing initially cooled the North Atlantic enough to expand sea ice, that the expanded sea ice then locked in cold surface conditions for decades by suppressing the convection that drives the circulation, and that the system entered a partly self-sustaining cold state that required a major positive forcing (the late nineteenth-century industrial CO2 increase) to escape.

None of these causes acted in isolation, and the relative contribution of each is still being argued out in journals like Nature Climate Change and Reviews of Geophysics. The dominant modern view is that the Little Ice Age was driven by all three together — a combination of weak solar output, sporadic volcanic cooling, and ocean-circulation feedbacks — with each contributing something and none being the sole cause.

What the cold did

Famines. The most thoroughly documented were the Great Famine of 1315-1322 (which killed perhaps 10-15% of the population of northern Europe and is now seen as one of the opening events of the Little Ice Age), the European famine of 1709 (after the worst winter on record, in which the Thames, the Seine, and even parts of the Mediterranean coast froze solid), and the 1816 famine (Tambora-driven, killed an estimated 100,000 to 200,000 across Europe).

Glacier advances. The European Alps reached their Holocene maximum extent around 1850, with glaciers in the Mont Blanc and Bernese Oberland regions extending several kilometers beyond their modern positions. Villages in the high Alps — Argentière in France, Fiesch in Switzerland — were destroyed or forced to relocate as ice tongues pushed into farmland in the seventeenth and eighteenth centuries. The Catholic Church in the Alps recorded several formal exorcisms of advancing glaciers, with one in 1644 led by the Bishop of Geneva. The glaciers continued advancing.

Social unrest. Geoffrey Parker’s Global Crisis makes the case that the political upheavals of the seventeenth century across Europe and Asia — the Thirty Years’ War, the English Civil War, the collapse of the Ming Dynasty in China, the Time of Troubles in Russia — share, beneath their specific local causes, a common climate trigger in the Maunder Minimum cold spike. Famine and crop failure led to economic stress, which led to tax pressure, which led to peasant revolts, which led to dynastic crises. The thesis is contested but well-documented for individual regions.

Frost fairs. The Thames froze through central London at least eight times between 1608 and 1814, providing the occasion for fairs on the ice with stalls, oxen roasted on spits, and printing presses producing souvenirs. The Thames has not frozen at London since.

When it ended, and what came next

The Little Ice Age ended, in the technical climatological sense, when the global average temperature returned to its pre-1300 baseline. This happened sometime in the mid-to-late nineteenth century. Different regional records give different end dates, but most place the transition between 1850 and 1900.

The cause of the warming is now generally accepted: the rising atmospheric concentration of carbon dioxide and methane from industrial activity. The same factor that ended the Little Ice Age is the factor that continues to warm the climate. Modern global temperatures are now substantially warmer than the pre-Little-Ice-Age medieval baseline as well — about 1.2 degrees Celsius warmer than 1850, according to the latest IPCC reconstructions, and rising.

The Norse Greenlanders have not yet been replaced. The Eastern Settlement, abandoned around 1450, is now under modern Greenlandic ownership; the cathedral at Garðar is a sheep pasture with low ruins of the chancel walls. The bronze bell from the cathedral, lost for centuries, was recovered in 1926 from a Norwegian collection that had purchased it from an Icelandic trader who had purchased it from somewhere unknown in the early sixteenth century. It is now in the National Museum of Denmark.

The bell is too cracked to ring.