In early February 1814, an elephant walked across the surface of the Thames near Blackfriars Bridge in London. The stunt was performed during the frost fair, when temperatures were so cold that for four days the top layers of the river froze solid. Londoners promptly held a festival, complete with what we might now call pop-up shops and a lot of unlicensed alcohol. Nobody could have known it at the time, but this was the last of the Thames frost fairs. They had taken place every few decades, at wildly irregular intervals, for several centuries. One of the most celebrated fairs took place during the Great Frost of 1683-84 and saw the birth of Chipperfield’s Circus. But the river in central London has not frozen over since 1814. The frost fairs are perhaps the most emblematic consequences of the “little ice age”, a period of chilly weather that lasted for several centuries. But while Londoners partied on the ice, other communities faced crop failures and other threats. The story of the little ice age is one of societies forced to adapt to changing conditions or perish. It’s also a long-standing mystery. Why did the climate cool and why did it stay that way for centuries? Thanks to decades of studies, we are finally closing in on an explanation. The emerging story involves volcanoes, the oceans, possibly the sun and possibly also genocide. A question of degree Like most things in science, the little ice age was discovered slowly and piecemeal. “This all came about because there were lots of documentary records from around Europe, stating that there were some really cold winters,” says climatologist Paola Moffa-Sánchez of Durham University. These were reflected in records of grain prices, which rose because of crop failures and ships’ logs saying Greenland was surrounded by sea ice and unreachable. The term “little ice age” was coined by a Dutch-born geologist named François Matthes, who in a 1939 report noted that glaciers in the Sierra Nevada in California had regrown at some point within the last few millennia. The term stuck, but it took decades to narrow down the timeframe. This was done by British climatologist Hubert Lamb, who went on to found the Climatic Research Unit at the University of East Anglia. In a study published in 1965, Lamb used European temperature records going back many centuries to identify a “notably warm climate… around AD1000–1200” that was “followed by a decline of temperature levels till between 1500 and 1700 the coldest phase since the last ice age occurred”. This chilly period was “undoubtedly upsetting for the human economies of those times (and perhaps of any time)” [his italics]. Since then, climatologists have tried to specify the duration and extent of the little ice age, but this has proved difficult. Most of the records showing the cooling are from Europe and records from elsewhere do not always show it. “It’s not a global phenomenon, in that it wasn’t cool everywhere,” says Alexander Koch at Simon Fraser University in Burnaby, British Columbia, Canada. In some places, such as China, the period was marked by a wetter but not cooler climate. What’s more, “the cooling was not continuous”, says Dagomar Degroot of Georgetown University in Washington DC. “It came in waves… that reached different places at different times at different magnitudes.” For this reason, the start and end of the little ice age are a little vague. “The classical definition is between 1400 and 1850,” says Moffa-Sánchez. However, “some people say 1300 to 1850”, pushing the start date back significantly. There is broad agreement that the period 1400-1800 is within the little ice age, she says, but outside that it’s fuzzy. Either way, the 1814 frost fair was near the end of the period. Furthermore, the little ice age was not technically an ice age. These are periods in which Earth has permanent ice at both poles and we have been in one for more than 2.5m years. Within that period, the ice sheets have repeatedly waxed and waned. The most recent ice advance, which we colloquially call the ice age but is really a glacial period, occurred from about 115,000 to  11,700 years ago. Since then, we have been in a relatively warm period called the Holocene. In that context, the little ice age is minor. “We’re talking about several 10ths of a degree C,” says Degroot. In contrast, the coldest portion of the last glacial period is thought to have been about 6C cooler than the past 4,000 years. Still, Degroot says, even this small change mattered. “On a regional or local scale, the anomalies could be really quite extreme,” he says, and documentary evidence shows that “at least some people were aware that they were living in an anomalous kind of climate”. Mystery cooling What was going on? The story isn’t entirely settled, but researchers are increasingly confident about the initial trigger: volcanoes. “You have these eruptions that are happening in clusters,” says Degroot. A 2015 study used data from ice cores to identify 25 major eruptions from the past 2,500 years. Between 1200 and 1400, there were huge eruptions of the Samalas volcano in Indonesia, Quilotoa in Ecuador and El Chichón in Mexico. Big eruptions blast sulphate aerosols into the stratosphere, the layer of the atmosphere above the weather. These aerosols reflect some of the sun’s rays back into space, cooling the Earth. In recent times, the 1991 eruption of Mount Pinatubo in the Phillippines triggered up to 0.5C of surface cooling. Usually, an eruption only cools the climate for a few years. “But if you have big eruptions in clusters, that can set off positive feedbacks in the climate system,” says Degroot. For example, several years of cooler temperatures cause sea ice to expand. Sea ice is white, so compared with dark blue water it reflects more radiation back into space. “That can prolong and exacerbate the cooling,” he adds. There can also be knock-on effects in the ocean, says Moffa-Sánchez. If the winds change, huge rafts of sea ice can be carried south from Greenland into the Labrador Sea. There, they interfere with the huge Atlantic currents that carry warm water from the tropics towards Europe. This is a complex story, in which the slight initial cooling from the volcanoes triggers changes in the Earth system that lead to more and lasting cooling. But it seems to hold true. A 2018 modelling study found that it was not possible to explain the little ice age without invoking volcanic eruptions, although that did not mean other factors were not also at work. A fading star Another possible factor is the sun. The amount of energy it pumps out varies ever so slightly, most famously over the 11-year solar cycle during which activity varies from a maximum to a minimum and back again. The effects on Earth are so small they’re hard to detect, but the sun sometimes has more of an impact. Several times in the past 1,000 years, our star has entered a “grand minimum”, in which it spends several decades being less active. The most recent was the Dalton Minimum between 1790 and 1820. This was preceded by the Maunder Minimum of 1645-1715. Before that, there is thought to have been the Spörer Minimum between about 1460 and 1550 and the Wolf Minimum around 1280-1350. Such grand minima can cool the planet, though by no more than 0.3C and probably less. Grand minima probably played a role in some of the chilliest episodes, says Moffa-Sánchez. In her studies of the period, “cold centuries coincided with these really famous solar minima.” She has found evidence that grand minima affect wind patterns, with knock-on effects for ocean currents and heat distribution. However, it seems unlikely that grand minima alone caused the little ice age. The timings don’t fit and, in any case, the climate impacts of grand minima are much smaller than those of massive eruptions. It may be that a solar maximum in the late 1300s also played a role. A study published in December 2021 found that this shifted wind patterns, interfering with the crucial Atlantic warm water current. The key point is – this isn’t an either-or debate, says Moffa-Sánchez. “It’s a likely combination of all of them.” Numerous causes helps explain why the climate shifts came in waves, rather than there being one abrupt change to a lasting cooler state. “It was not that you had this little ice age period where it was always cold,” says Moffa-Sánchez. “You just had several cold centuries throughout this four-century-long period.” However, one big puzzle remains. The coldest period of the little ice age was around 1610 and it doesn’t coincide with a grand minimum. Nor was there a particularly big eruption: Peru’s Huaynaputina went off in 1600, but while the blast was large it wasn’t exceptional. Instead, the suggestion is that this cold spell was caused by humanity – in a truly horrible way. The great dying In 1492, Christopher Columbus reached the Americas. Over the following decades, Europeans began colonising them. In the process, they fought with Indigenous Americans, often killing them. But even more lethally, they brought diseases. One of the worst was smallpox, which killed millions. As well as being a genocide and a tragedy, this may have had an impact on the climate. Many Indigenous Americans were farmers who had cleared forests for their crops and when they died the trees grew back, drawing carbon dioxide out of the air and cooling the planet. This scenario was first outlined by climatologist William Ruddiman in 2003, as part of his “early Anthropocene” hypothesis that humans have been affecting Earth’s climate for millennia, albeit less than we are today. The idea that mass deaths among Indigenous Americans led to climate cooling has received tentative support from modelling studies. Still, it has been controversial because there are so many uncertainties about the key numbers. However, in 2019 Koch and his colleagues published an updated analysis. They went through the argument step by step and tried to quantify everything, from the number of people who died to the extent of reforestation. They concluded that the European arrival led to 56m deaths by 1600. This dreadful toll meant trees grew again on 56m hectares of land, removing 27.4bn bn kilograms of carbon dioxide from the air. “It’s a really interesting theory,” says Degroot. However, he remains sceptical because we don’t know how land use was changing in other parts of the world, especially Africa. Resilience While the question of whether human society contributed to the little ice agee remains up for discussion, what’s clear is that the little ice age affected human society. For example, Norse settlers made their home in Greenland in 985 and stayed for centuries, only abandoning it in the early 1400s. It has often been suggested that the little ice age played a role in that. However, a study published in March looked at sediments from the region and found no sign of cooling, but it did find a drying trend, which would have meant less grass to feed livestock. Such stories may be true, but it’s crucial to remember that people were not passive victims, says Degroot. “You don’t just see people who are completely at the whim of changing climatic conditions, you see people changing adaptively.” Degroot points to the Arctic, where there was often remarkable activity despite the increase in sea ice. Between 1611 and 1619, European whalers operated off the shores of Svalbard, because there was a shortage of vegetable oil and whale oil was a substitute. For the Dutch Republic, the period between 1560 and 1720 was something of a golden age, despite or even because of the cold winters. It did not depend on home-grown agriculture, so local crop failures were less of a problem, says Degroot. Instead, the republic’s economy was driven by merchant ships, whose operators devised ingenious ways to cope with the cold – and the Dutch thrived while their neighbours struggled. “It so rarely is what you’d expect,” says Degroot. Studying how past societies responded to climatic shifts such as the little ice age will help us all, as the climate crisis becomes ever more intense. “Hopefully, we can learn from them, figure out what they did right, what they did wrong.” We had better learn fast, because the little ice age was just a taster. Back then, the average global temperature cooled by a fraction of a degree, but we’ve already warmed it by 1.1C, and are set to blow past 1.5C in the next few decades.