Quick Take: A Brief History of Climate Change

In historical eras, warmer times have been a boon for human civilization

(This Quick Take is free. About two-thirds of my posts, those of the History of Mankind series, are for paying subscribers only. Don’t hesitate to comment and let me know what you think I got wrong, or right or whatever: the chance to get that kind of feedback from a larger audience precisely is one of the main reasons why most of my Quick Takes are free.)

Climate change has been the single most contentious issue in my lifetime. Political discussions are friendly rounds of golf compared with the average argument about climate change.

There are many specialized places when one can go have average arguments about climate change and learn about current measurements, climate models and projections; and yet there is surprising little discussion out there about the impact of climate change across history. That ends now: strap in for a bumpy ride with dramatic temperature swings.

Let’s start by having a look at this map below. It shows Lake Chad in the Sahel region just south of the Sahara Desert, and the network of rivers feeding it:

It’s a lot of rivers for a fairly dry region, isn’t it?

It is because that map doesn’t represent the current Lake Chad, but the one that existed (very roughly) around 8,000-6,000 BC, soon after the start of the Holocene, the historical period in which we currently live.

The Holocene is an interesting geological period, and it followed one of the most mysterious periods in the Earth’s geological history: the Younger Dryas¹, an extremely cold spell that lasted for about a millennium, driving many human populations to extinction or close to it, from around 10,800 BC.

There are many competing theories on what exactly triggered the Younger Dryas, and I have my favorite². That’s not important³. What’s important is that the Younger Dryas was huge, and I can’t believe why it’s not discussed more often by news media given the contemporary obsession with the climate: it’s even possible that agriculture arose as an evolutionary response to the colder conditions under the Younger Dryas, as hunter-gatherers found ways to turn much more vegetarian, and find and nurture places where attractive crops grew, or else starve to death.

Regardless, the Younger Dryas ended with a bang, just as it had started. In 9600 BC, global temperatures rose by seven degrees in less than a decade, marking the end of the period and the start of the Holocene. Yes: in less than a decade. Remember that the next time you listen to some TV pundit speaking about “unprecedented climate change” because something went up or down by 0.1%.

The explosion in human population was immediate and, according to some estimates, nearly three quarters of Earth’s land was inhabited within a century, and therefore shaped by humans, including more than 95% of temperate and 90% of tropical woodlands⁴.

Food production in the Fertile Crescent soared: the new crop varieties, accustomed to drier weather and shorter growing seasons, now grew riper and more nutritious due to the suddenly warmer, wetter climate⁵. Agriculture started in full, and the first truly sedentary settlements appeared: while, in the past, no area could sustain a large permanent population of hunter-gatherers, agriculture offered the possibility of maintaining a much larger density of people on a smaller footprint. Civilization, for better or worse, was now a thing.

Let’s go back to Lake Chad and Africa. The Sahara Desert, like all other deserts on the planet, had greatly expanded during the Younger Dryas. This is common-sensical: whatever caused the cooling period made the Earth’s polar ice caps expand — to levels well above those of the 21st century — which meant that there was less water available in the oceans, being frozen in place in the caps, and it was harder to evaporate because it was colder. Thus, solar irradiation led to less evaporation, less water in the atmosphere and lower rainfall levels: thus, the Sahara grew.

This is not something particular or unique that happened in the Younger Dryas. It’s the normal Earth cycle, and it was only extremely evident during the Younger Dryas, because everything about that period was so radical and involved such rapid changes.

Take the Nile River. For most of the prehistory, Egypt’s most famous (and now sole) river didn’t run through Egypt: its course was well to the west and drained in modern Libya, for various, not very well understood reasons, flowing into the Gulf of Sidra, and helping to make Libya a wetter and more hospitable place than it's now.

Still lacking the Nile, Egypt was inhabited because the warmer climate right at the start of the Holocene meant more rainfall: in places like Wadi Kubbaniya, dry as shit these days, hunter gatherers hunted fish in seasonal ponds or small lakes, or hunted small fauna in the surroundings. And Lake Chad was that huge thing you can see in the map at the top of the post. Now, if you scroll back up and look closely at southwest corner of the lake, you can see a faint green shape: that’s the current extension of the entire Lake Chad, right now⁶.

This period during which the Sahara was an appealing steppe full of water-holes, small rivers and animals to hunt is called the African Humid Period. Again, there are multiple estimates on when it actually ended. What we know for certain is that there’s nothing specifically African about it: the whole planet was undergoing a humid period, that only ended as the Earth slowly cooled again, leading to a drier climate.

I could go on and on with examples, so I’ll just stick to some.

Lakes ballooned in size across the world for a few thousand years after 9600 BC. It wasn’t just Lake Chad: the gigantic salt basins of Makgadikgadi in Namibia, Dasht-e Kavir in Iran, Rub' al-Khali in Saudi Arabia, White Sands in New Mexico and Salar de Uyuni in Bolivia, among many others, are remains from huge water reservoirs that once overflowed with life and sustained verdant ecosystems, and are now essentially dead zones; Hamun Lake, now seasonal and mostly a wetland area fed by the Helmand river in modern Afghanistan, was also massive before the climate cooled. Water levels at Lake Turkana between Ethiopia and Kenya were around 100 meters higher in 7500 BC than they were in 2000 AD.

Arabia is an interesting case. If one doesn’t understand that Arabia was wetter for most of the Holocene, one doesn’t understand how it’s possible that wave after wave after wave of Arabians have been leaving the peninsula for thousands of years, conquering lands near and far. Because, if Arabia had always been almost entirely desertic, as it is now, it would have never sustained a sizeable population.

There’s much evidence for Arabia’s wetter history. Around 5000 BC, tribesmen living in the then-dry steppe of central Arabia, now rocky desert, built massive mounds of rocks with likely religious functions, called “mustatil.” One clue to their purpose is that the head walls of many mustatils have a small chamber or niche that seems to have been used for sacrificial animal offerings. Excavations have found the horns and bones of wild and domesticated animals, including sheep and gazelles, but mostly cattle, kept when the region was much wetter and greener than the arid landscape to be seen millennia later.

The cooling and drying of Arabia was, like elsewhere, a gradual process, but it appears that, by 500 BC or so, the Arabian deserts were much as they are now and the same can be said of others like the Gobi and Sahara deserts. They all briefly became greener again later, because the history of the climate didn’t stop there.

The best evidence available indicates that the global climate kept cooling until what we call the Roman Climatic Optimum (sometimes also referred to as the “Beysehir Occupation Phase,” after an influential 1998 paper) started in around 250 BC, leading to warmer trends across the globe, until the 5th or 6th centuries AD⁷.

A massive shift in the nature of the Roman countryside was evident in the era, leading to an explosion in agricultural output. Tree rings from the Italian Peninsula in the late 3rd century BC indicate a time of mild conditions there at the time of Hannibal's crossing, which possibly made it easier for some elephants to survive that famous march – something that would have been outright impossible with mid-20^th century Alpine temperatures.

Again, this was a global trend, even if it’s named after the Romans because they provided the best available evidence, particularly after their empire grew to span much of the most developed land of Eurasia and Africa. Additional evidence for this trend comes from all over the world⁸.

Just exactly when the Roman Climatic Optimum ended is a contentious issue. it’s possible that there was significant global cooling as early as the 4th or 5th century. What is clear is that the 6th century was an absolute disaster for mankind, especially after a cooling wave started exactly in 535-536, according to multiple sources across the world. Droughts and harvest losses were reported in Europe, Peru (the Moche culture that emerged during the first half of the millennium along the coast to the north of the earlier Chavín culture), Iran’s Sasanian state and China.

“For the sun gave forth its light without brightness, like the moon, during the whole year,” wrote the historian Procopius about 536. John of Ephesus (507-588), a Syriac bishop and historian, made similar remarks, noting in his “Ecclesiastical History” that “everyone declared that the sun would never recover its full light again.”

This was almost certainly related to a volcanic eruption, due to the evidence of sulfate deposits in ice cores; the eruption may have taken place in Iceland, North America, Indonesia or elsewhere. As a result, temperatures fell as much as two degrees below normal in Europe and many other places. Another, perhaps related, eruption in 539/540 led to another volcanic winter that may have been less severe, and other eruptions may have followed until at least 547⁹.

The term “Late Antique Little Ice Age” is used by specialists to describe a cooling period that may have lasted until the 8th century, possibly as a result of this string of eruptions¹⁰. Its consequences were certainly severe, particularly across the northern hemisphere and very much so around the Mediterranean.

Snow fell across Scandinavia even in summer, trees were stunted and died in India and China, crops failed across the planet; wine cultivation ended in Britain for over a millennium — although this was perhaps for the best, depending on your opinion on English wine.

Happy days returned around 750, given widespread evidence of rising global temperatures from roughly that year¹¹, although the process was far from steady: there was again a worsening, cooling climate from the decades before 900, with higher aridity conditions evident in Central America from around 800 and Peru from at least around 880. As usual, such cooling snaps triggered droughts across the globe as atmospheric humidity became ice in northern latitudes¹².

Also from about 800, the Mendenhall and Exit Glaciers in southern Alaska regained ground for some time, engulfing trees that were again disgorged in the early 21^st century, when the ends of both glaciers melted back. Byzantine chroniclers reported that the winter of 927-928 was terrible, with four months of frost and many killed by cold and starvation, particularly in the Anatolian highlands, where people were forced “to pick the grains of barley from the dung of horses and asses and eat them,” as a very fastidious chronicler wrote.

As a separate chronicler put it, after seven successive years of crop failure in Armenia in the 950s, some people went mad and attacked each other senselessly. The 969 Fatimid conquest of Egypt was facilitated by the catastrophic failure of the annual Nile floods, leaving many dead or starving and most unable to support the status quo.

Perhaps it was seeking to escape this cold snap that Viking settlers migrated across the Atlantic, although the fact that they colonized Iceland between 870 and 930, as well as Greenland’s west coast from 986 and Newfoundland soon thereafter, would indicate conditions not as bad as those in the (nearby) mainland Americas – or that the climate took a turn for the better from mid-century, having reached a brief cold peak sometime before or around the Fatimid move on Egypt, so it was already warm enough to make Greenland (and Iceland) bearable by the 980s.

This idea of quick improvement is reinforced by the fact that, in Europe and nearby regions, the effects of the cooling period may have been more subdued than in other continents, which can only partly be explained by technological advance: it was around the year 1000 that the horse collar arrived in Europe from China, which helped to increase farming output but almost certainly is not the only reason for the kind of spike in yields seen over the 11^th century and later.

Thus, climate historians speak of a Medieval Warming Period that stretched from 900 to 1300, even though the evidence for this warmer climate is particularly concentrated in the final two centuries: for example, wine grapes – grown in Germany in the 21^st century up to elevations of about 560 meters – extended all the way to 780 meters from about the year 1100, which would imply temperature warming of up to a degree and a half¹³. There’s evidence of similar developments in neighboring Poland¹⁴.

At the same time, the first Russian state expanded under beneficial conditions that helped boost production across the steppe and elsewhere in the globe particularly from the years just before the turn of the millennium. Like Germany and Poland, Eastern Siberia was a degree and a half warmer than in the early 21^st century during the period¹⁵, at a time when mean sea surface temperatures were also warmer¹⁶ and ice buildup at the Antarctic Ross Sea was significantly smaller¹⁷.

To summarize, on a global level it appears safest to state that the climate turned cooler at the end of the Roman Warm Period, possibly some time before 400, and reached a peak of coldness, possibly because of low solar irradiation connected to volcano activity, for well over a century after 535; this period led, for example, to the complete desiccation of the Negev between the Sinai and Palestine, exemplified in the abandonment of the town of Nessana, south of Gaza, sometime in the late 7^th century.

This cooling phase ended around 750, although cooling stretches like those reported by Mayan and Byzantine sources were still common. These stretches lasted until around 950-960 in some regions, and had a clear impact in Maya lands and northern China, leading to a degree of desertification that made it harder for the Tang Dynasty to survive. At that point, the Medieval Warming Period took over, almost unimpeded, until roughly 1300¹⁸. This slowly gave way to a longer cooling phase, with a temporary cooling spike during the Maunder Minimum¹⁹, a span of particularly low solar irradiation between 1460 and 1550 perhaps caused by the effects of a large eruption of Peru’s Misti Volcano.

Still, this cooling phase was still to peak later, during the Maunder Minimum starting in the early 17^th century, sometimes described as the Little Ice Age²⁰. Smaller cooling spikes followed throughout the 18^th century until a warming phase started in the mid-19^th century, only interrupted briefly in the mid-20^th century.

And here we are, after all those centuries worth of accumulated evidence, and much of mainstream media is still consistently tying global warming with droughts.

1. The period is named after the alpine-tundra wildflower Dryas octopetala, as its leaves are occasionally abundant in the Late Glacial, often minerogenic-rich, like the lake sediments of Scandinavian lakes. The Younger Dryas was the most recent and longest of several interruptions to the gradual warming of the Earth's climate since the severe Last Glacial Maximum. The trend stopped and reversed from about 3500 BC, when a new period of cooling (with fits and starts) kicked off, that should have led to the next glacial era in a few thousand years. That, of course, is now very much in doubt due to renewed global warming observed over the last two centuries, at least partly because of human activity.

2. I’m supportive of theory that the Younger Dryas was triggered by a change in course for the Lake Agassiz's basin, in modern Canada. This lake had formed in modern Manitoba and Saskatchewan around 30,000 BC, with meltwater from the three-kilometer-thick Laurentide Ice Shield of the Wisconsin Glaciation and occupied much of modern central-western Canada. For millennia, it fed rivers that flowed southwest into the Pacific, while creating a dam that prevented meltwaters from making their way to Hudson Bay; for some reason, possibly connected to the melting of large ice-walls. There's a long historiographical discussion on the exact changes suffered by the lake. Broecker at al (1989) argued that a change in the routing of the melt waters from the Laurentide ice sheet was the cause of the Younger Dryas. The drainage area then switched east fairly abruptly and huge amounts of extremely cold water came into contact with the Arctic Ocean, with the consequence that the Gulf Stream flowing north from the Caribbean was disrupted and even stopped completely. The sudden flow was estimated at an initial rate of more than 800 Olympic swimming pools per second, according to “Catastrophic Drainage From the Northwestern Outlet of Glacial Lake Agassiz During the Younger Dryas,” a 2021 paper by S. L. Norris et al in “Geophysical Research Letters.” It lasted for between six and nine months. The flash-floods connected to the sudden discharge of Agassiz water may have left a legacy in the multiple human traditions of ancient floods caused by upset gods. Some of these traditions may also refer to the final draining of Agassiz in around 8,000 BC, which caused another, shorter, cold weather snap, combined with floods in parts of the Northern Hemisphere.

3. There's a cool debate (pun intended) about the actual trigger for the Younger Dryas. The “Younger Dryas impact hypothesis” posits that a meteorite, which likely hit Earth on Abu Hureyra, Syria, was the main cause for the global cooling period; in “Evidence of Cosmic Impact at Abu Hureyra, Syria at the Younger Dryas Onset (~12.8 ka): High-temperature melting at >2200 °C,” a 2020 paper by Andrew M. T. Moore et al published in Nature, the authors note that the Younger Dryas Boundary layer contains enough evidence to support the hypothesis that the impact was significant enough to create the colder conditions. In my opinion, the delayed, and highly irregular effects across various regions of the planet is better explained by the slow drainage of Lake Agassiz and its impact on the Gulf Stream.

4. One such estimate is in “People have shaped most of terrestrial nature for at least 12,000 years,” a 2021 paper published in PNAS by Erle C. Ellis et al. A similar conclusion is reached in “Ancient herders enriched and restructured African grasslands,” a 2018 paper in Nature by Fiona Marshall et al.

5. For a primer on the difficulties involved in estimating climate patterns since the Younger Dryas, see Cartapanis, O., Jonkers, L., Moffa-Sanchez, P. et al. “Complex spatio-temporal structure of the Holocene Thermal Maximum.” Nature Commun 13, 5662 (2022). As the authors write: “Climate reconstructions suggest an early-middle Holocene Thermal Maximum (HTM) followed by gradual cooling, whereas climate models indicate continuous warming. This discrepancy either implies seasonal biases in proxy-based climate reconstructions, or that the climate model sensitivity to forcings and feedbacks needs to be reevaluated.”

6. At times, Lake Chad reached a maximum extent of some 400,000 square kilometers in surface area, larger than the modern Caspian Sea, with a surface level about 30 meters higher than its twentieth-century average, becoming a draw for migrants from both the south and the north, who came together to form the southernmost family of the Semitic Afro-Asiatic languages, that of Chadic languages still widespread in parts of Nigeria, Niger, Chad, the Central African Republic and north Cameroon. The Caspian Sea itself grew enormously larger after 9600 BC: the warming climate melted the northern glaciers and the permafrost, releasing their combined meltwater in a torrential surge that flowed south, so the Caspian ballooned into a vast interior sea designated the Khvalynian Sea. For two thousand years the northern shoreline stood near Saratov on the middle Volga and Orenburg on the Ural River, restricting east-west movement south of the Ural Mountains. The Khvalynian Sea separated the already noticeably different late-glacial forager cultures that prospered east and west of the Ural Mountains. Around 9000 BC, the water finally rose high enough to overflow catastrophically through a southwestern outlet, the Manych Depression north of the North Caucasus Mountains; some believe that a violent flood poured into the Black Sea, contributing to its filling.

7. Multiple papers support this view. In “Demographic dynamics between 5500 and 3500 calBP (3550–1550 BCE) in selected study regions of Central Europe and the role of regional climate influences,” (PLOS One, 25.10.2023) Ralph Grossmann et al note that a consistently cooling climate in the areas studied (in central/southern Germany, Austria and the Czech Republic) led to population decreases among those who stayed, as well as a spike in social inequality as the most powerful grabbed scarce resources. In fact, it's likely that a particularly cold snap occurred around the turn of the 13th-12th centuries BC, leading to particularly low rainfall in Africa and unusually weak flooding in Egypt, which contributed to troubles there. In “Severe multi-year drought coincident with Hittite collapse around 1198–1196 BC” (Nature, 8.2.2023) Stuart W. Manning et al examined the width tree rings and stable isotope records in central Anatolia and concluded that a severely dry period spanned between 1198 BC and 1196 BC, just as the Hittite state was hit hardest by the Sea Peoples.

8. This is reported, for example, by McDermott et al in 2001 (”Centennial-scale Holocene climate variability revealed by a high-resolution speleothem:18O record from SW Ireland,” Science 294: 1328-1331), Olafsdottir, R. and Gudmundsson, H.J. in 2002 (Holocene land degradation and climatic change in northeastern Iceland, The Holocene 12: 159-167), Niggemann, S. et al in 2003 (“A paleoclimate record of the last 17,600 years in stalagmites from the B7 cave, Sauerland, Germany,” in Quaternary Science Reviews 22: 555-567) and Desprat, S. et al in 2003 (“Revealing climatic variability of the last three millennia in northwestern Iberia using pollen influx data,” Earth and Planetary Science Letters 213: 63-78). In “Persistent warm Mediterranean surface waters during the Roman period” (Nature, 26.6.2020) G. Margaritelli et al write that “the Roman (was) the warmest period of the last 2,000 years, about 2 °C warmer than average values for the late centuries for the Sicily and Western Mediterranean regions. After the Roman Period a general cooling trend developed in the region with several minor oscillations.” Importantly, this period, while characterized by a warm global climate overall, contained significant fluctuations: in “Climate change, society, and pandemic disease in Roman Italy between 200 BCE and 600 CE” (Science Advances, 26.1.2024), Karin F. Zonneveld et al document phases of instability and cooling from around 100 AD onward but more notably after 130 AD. As they add: “Pronounced cold phases between 160 to 180 AD, 245 to 275 CE, and after 530 CE associate with pandemic disease, suggesting that climate stress interacted with social and biological variables.”

9. The Ilopango volcano in El Salvador is the top suspect for both eruptions, particularly the second. Its eruptions were so brutal that the caldera left behind, now Lake Ilopango, is 72 square kilometers in size. See “Radiocarbon and geologic evidence reveal Ilopango volcano as source of the colossal ‘mystery’ eruption of 539/40 CE,” by Robert A. Dull et al, Quaternary Science Reviews, Volume 222, 15 October 2019. This eruption likely was bigger than the 1815 Mount Tambora eruption that led to the so-called year without a summer. Also see Helama et al., Op. Cit. Other plausible candidates have been proposed for eruptions in 536, 540 and later in 547, in locations like Papua New Guinea and Krakatau. The smaller, nearby Volcan de San Salvador, just 30 kilometers to the west, erupted in around 650, and several times more over next few centuries, well into the 2^nd millennium.

10. There’s of course, much debate in this subject. See, for example, “Cooling and societal change during the Late Antique Little Ice Age from 536 to around 660 AD,” by Ulf Büntge et al, Nature, 8.2.2016.

11. See, for example, “Historical population changes of Adélie penguins in the Ross Sea region, Antarctica, and its climatic forcings,” by Zhangqin Zheng et al (Quaternary Science Reviews, 15.10.2023). The authors found that the penguin population on Inexpressible Island, Antarctica, peaked between 750 and 1350 AD, possibly due to habitat expansion in a warmer climate.

12. Cit. “Putting the rise of the Inca Empire within a climatic and land management context,” A. Chepstow-Lusty et al, Climate of the Past, 5, 375–388

13. See “Geologic Evidence of Recurring Climate Cycles and Their Implications for the Cause of Global Climate Changes – the Past is the Key to the Future,” by Don J. Easterbrook, in Evidence-Based Climate Science, 2011.

14. See “The climate in Poland (central Europe) in the first half of the last millennium, revisited,” by Rajmund Przybylak (Climate of the Past, Vol. 19, issue 11, 2023): “All the analysed quantitative reconstructions suggest that the Medieval Warming Period in Poland was comparable to or warmer than the mean temperature in the period 1951–2000.”

15. See the 2021 Springer paper “Reconstruction of Environmental Conditions in the Eastern Part of Primorsky Krai (Russian Far East) in the Late Holocene,” by L. B. Nazarova et al. Also, “Комплексная реконструкция температуры российской Арктики за последние два тысячелетия,” by B.V. Klimenko et al (Научные исследования в Арктике, 2013).

16. See the 2021 Nature paper “Modern and sub-fossil corals suggest reduced temperature variability in the eastern pole of the Indian Ocean Dipole during the medieval climate anomaly,” by Sri Yudawati Cahyarini et al. Also, the 2021 Cambridge UP paper “Sea surface temperature seasonality in the northern South China Sea during the middle Holocene derived from high resolution Sr/Ca ratios of Tridacna shells,” by Pengchao Zhou et al.

17. See the 2021 paper “6,000-Year Reconstruction of Modified Circumpolar Deep Water Intrusion and Its Effects on Sea Ice and Penguin in the Ross Sea” by Q. B. Xu et al, in Advancing Earth and Space Science. The paper concludes that sea ice extent for Antarctica’s Ross Sea was more extensive in the early 21^st century (and temperatures cooler) than nearly any time in 6,000 years, and that penguin numbers decline with cooling/increased sea ice.

18. The eruption of North Korea’s Paektu Volcano in 946 may have briefly interrupted the process, as its cooling effects were felt in Japan and perhaps elsewhere.

19. Or Sporer Minimum.

20. There are multiple caveats and regional adjustments to add to this picture, of course. Le Roy Ladurie, in “Histoire et Climat” (Annales, Economies, Sociétés, Civilisations 14, 1959), notes evidence in Scandinavia that the Little Ice Age may have been on the wane by 1460, and a whole century of better (warmer) climate may have preceded the turn for the worse in the mid-16^th century. Then again, evidence for the southern hemisphere is much less abundant.