Climate and the steppe empires

Rouran epoch Göktürk epoch Uyghur epoch Kirghiz epoch

Rouran First Göktürk Tang rule Second Göktürk Uyghur Kirghiz Huns Western Göktürk Türgesh Karluk Khazar

Rouran 400–554
First Göktürk 552–630
Tang rule 630–682
Second Göktürk 682–744
Uyghur 744–840
Kirghiz 840–924
Huns 370–469
Western Göktürk 552–659
Türgesh 699–766
Karluk 766–800
Khazar 650–969

The climate of the steppe between the Xiongnu and the Uyghurs can now be measured, and it is no longer known only from the sources. Larches on the timberline of the Altai, pines of northern Fennoscandia, sulphate layers in the ice of Greenland and Antarctica, and the mud of a lake at the western edge of the Khangai each keep a record, year by year or decade by decade. Since about 2015 those records have been dated well enough to be set beside the Chinese annals and the inscriptions. Historians and the palaeoclimatologists themselves have argued from these records about the Göktürks and their neighbours. Every proposition below is a position, recorded with the name of whoever holds it and the grounds on which it rests. The chronology lists the dates under the heading of environment.

Palaeoclimatic record

The sixth century opened with a volcanic double event. A first eruption, somewhere in the northern hemisphere, in 535 or early 536, threw sulphate and ash into the stratosphere. Observers in Rome and Constantinople saw the sun dimmed from 24 March 536 for as long as eighteen months. The summer of 536 was the coldest single year of at least the past two thousand in every tree-ring reconstruction of the northern hemisphere. A second eruption, tropical, in 539 or 540, loaded the atmosphere with up to a tenth more aerosol than Tambora in 1815. Summer temperatures in Europe fell again in 541, and the cold held until almost 550. A third, smaller eruption followed in 547. Carbon isotopes in Fennoscandian pines show the light itself reduced in 536 and again from 541 to 544. They also show the cooling of 574, which followed an eruption dated to the same year in the ice. The decade 536 to 545 was the coldest of the last two thousand years. In the ice-core record, the eruptions that opened it ranked only as the eighteenth and fifth strongest of the era. The modellers conclude that what made them exceptional was not the size of either but their coming four years apart.

In the Altai the cold lasted longer than in Europe. Büntgen and his colleagues worked from thirty-four tree-ring chronologies of the Russian Altai and the Austrian Alps. They define a Late Antique Little Ice Age from 536 to about 660. Ocean and sea-ice feedbacks and a solar minimum sustained it. It was more severe than the Little Ice Age of the second millennium. The 540s were the coldest decade of the Altai reconstruction, 3.2 degrees below the mean of 1961 to 1990. Of the twenty coldest central Asian summers of the era, thirteen fell in the sixth century after 536. Of the twenty coldest decades, thirteen fell in the sixth and seventh centuries. The Justinianic plague of 541 to 543 broke out across the eastern Roman empire. It came after the food shortages that followed the first cold.

Further records matter here. The revised ice-core chronology finds large sulphate signals in 574, 626 and 682. Among the sixteen coldest summers between 500 BC and 1000 AD, the four coldest are 43 BC, 536, 543 and 627. Each came shortly after a large eruption. The same study prints the historical accounts of a dimmed sky in 626 and 627 beside those of 536 and 537. On the monsoon margin of north-western China, a tree-ring reconstruction of rainfall over 6,700 years finds a long dry phase from 346 to 763. Within that phase the trend turned in 544. Dry extremes were most frequent around 660, and conditions shifted to wetter in 760.

The lake gives the longest view. A sediment core from Lake Telmen, at the western edge of the Khangai in Zavkhan, records humid conditions from 1200 BC to about 700 AD and drier ones from 700 to 1300. The humid phases coincide with solar minima, when lower temperatures meant less evaporation and more grass. A shoreline terrace 4.8 metres above the present lake, dated between 600 and 700, marks a high stand shortly before the drying. The authors explain it by the solar minimum around 700 or, within the dating error, by the Late Antique Little Ice Age itself. For the centuries after 900, a tree-ring record of the moisture balance of central Mongolia adds the droughts of the Medieval Climate Anomaly and the drought of the 1180s. It also adds the fifteen wet years of 1211 to 1225, which coincide with the rise of Chinggis Khan. A reconstruction of drought over 2,060 years marks a severe drought beginning in 804. The tree rings of the Orkhon country date the most severe drought of the Uyghur period to the thirteen years from 805.^[1]

Sixth century and the Göktürks

Büntgen and his colleagues see the Late Antique Little Ice Age as an environmental factor in the movements out of the Asian steppe. Shorter growing seasons and the nutritional deficiencies of people and livestock, they write, probably initiated large-scale pastoral movements toward China. In the sixth century, conflicts among nomadic groups and the regimes of northern China culminated in Mongolia in the replacement of the Rouran by the Göktürks. They see in this evidence of major upheavals in central Asia as early as the 550s, spreading through the 560s and 580s. The Avars arrived north of the Black Sea about 550. Struck and his colleagues, working from the lake, put the same years differently. The Rouran khaganate about 400 and the first Göktürk khaganate about 550 both formed during favourable grassland conditions in central Mongolia. That run ended with the onset of the Late Antique Little Ice Age around 600. Very harsh and long winters seem to have caused high livestock mortality, an increase in warfare, famines and cultural reorganisation. Failing grassland biomass, they suggest, may have undercut the economic and social power base of the first khaganate and contributed to its disintegration in about 603. A third study, citing Ganiev and Kukarskih, matches the six years of Bumin’s war with the extreme years of 536 to 545. It matches the first years of the civil war, 581 to 583, with extreme weather again. Turkic Atlas gives those interpretations under the War of Independence and the civil war. Under the war of 588, Turkic Atlas gives the speculation that drought drove Bagha to the cities of the Oxus.

The case for treating the climate as a cause, and not only as a setting, is that the sequence is too close to be coincidence. The coldest decade of the Altai record fell in the 540s. Bumin’s war followed in the six years after it. The extreme years of 581 to 583 coincide with the civil war that opened in 581. The sulphate signal of 626, the dimmed sky of 626 and 627 and the fourth-coldest summer of fifteen hundred years in 627 precede the collapse of the eastern khaganate in 630. A nomad state lives off grass and herds. A state that loses its herds loses its army within a season. The Chinese account of 630 names hunger, natural disasters and a tax on an exhausted people. This is how a failed grazing year appears in an administrative record.^[2] The palaeoclimatologists ask for caution about the causal nexus. That caution bears on how the case is argued, not on whether the cold belongs among the causes.

The two studies do not agree on the date at which the cold began to matter for Mongolia. Büntgen’s Altai record has the coldest decade in the 540s, before the Göktürk revolt. The lake has the good years running to about 600, so that the first khaganate rose in favourable conditions and broke in bad ones. Both are consistent with the record as each author measures it. The difference is one of proxy and place: summer temperature on the timberline of the Russian Altai against the moisture balance of a lake six hundred kilometres to the east. The disagreement concerns which years were worst, not whether the cold mattered. On that question the two records agree.

Eruption of 626 and the fall of 630

Fei, Zhou and Hou argued in 2007 that a volcanic eruption around 626, and the cooling that followed, stand behind the collapse of the eastern Göktürk khaganate in 630. Di Cosmo, Oppenheimer and Büntgen returned to the question in 2017 under the title of an interplay of environmental and socio-political factors. Büntgen summarises the sequence this way: Ilig raided and invaded northern China in the 620s. After 626, however, he experienced internal wars and Tang counterattacks, which led to the defeat and collapse of 630. The ice-core chronology of 2015 gives the argument its footing: a large sulphate signal in 626, a dimmed sky recorded in 626 and 627, and the summer of 627 among the four coldest of fifteen hundred years. The Chinese account of the fall, as the article on the khaganate gives it, names natural disasters, hunger, a tax on an exhausted people and the flight of the tribes. The measured account and the written account describe the same years from opposite sides.

Uyghurs and the drought

Di Cosmo and his colleagues worked with the tree-ring data of the Orkhon country. They identified favourable humid conditions at the emergence of the Uyghur empire and a persistent drought from 783 to 850, and they rethought the history of the empire around them. The drought reconstruction of 2018 marks the onset as a severe drought beginning in 804. The Xin Tangshu records that 839 brought famine, pestilence and heavy snow to the Uyghurs and killed many of their sheep and horses. It records that the Kirghiz came the next year. The Uyghur article tells the fall.^[3] After 850 the lake and the tree rings agree on the droughts of the Medieval Climate Anomaly. They agree on them until the wet years of the early thirteenth century, which lie beyond the period of the khaganates.

Limits of the evidence

The authors of the Altai reconstruction write that any hypothesis of a causal nexus between the sixth-century thermal shock and the plagues, empires, migrations and upheavals that followed requires caution. What their record offers, they write, is a fit with the transformative events of the age, not a proof of them. The same rule holds here. The measurements are findings: the eruptions and their years, the cold decade of 536 to 545, the summer of 627, the drought of 804 and after. The links between them and the wars of the Göktürks are positions, held by named authors, and they stand in the chronology as positions. A khan who lost his herds to a winter did not thereby lose his khaganate. The Chinese histories name the winters. They also name the Sui agents, the Tang cavalry and the tribes that changed sides.

References

^ Cosmo 2018, p. 448
^ Jiu Tangshu (Old Book of Tang), juan 194
^ Xin Tangshu (New Book of Tang), juan 217

Sources

Xin Tangshu (New Book of Tang), juan 217 (after Jiu Tangshu)
Jiu Tangshu (Old Book of Tang), juan 194

Literature

Ulf Büntgen; Vladimir S. Myglan; Fredrik Charpentier Ljungqvist; Michael McCormick; Nicola Di Cosmo et al., Cooling and societal change during the Late Antique Little Ice Age from 536 to around 660 AD (2016)
Michael Sigl; Mai Winstrup; Joseph R. McConnell; Ulf Büntgen et al., Timing and climate forcing of volcanic eruptions for the past 2,500 years (2015)
Matthew Toohey; Kirstin Krüger; Michael Sigl; Frode Stordal; Henrik Svensen, Climatic and societal impacts of a volcanic double event at the dawn of the Middle Ages (2016)
Samuli Helama; Laura Arppe; Joonas Uusitalo et al., Volcanic dust veils from sixth century tree-ring isotopes linked to reduced irradiance, primary production and human health (2018)
Julian Struck; Marcel Bliedtner; Paul Strobel; William Taylor et al., Central Mongolian lake sediments reveal new insights on climate change and equestrian empires in the Eastern Steppes (2022)
Bao Yang; Chun Qin; Achim Bräuning; Timothy J. Osborn; Valerie Trouet et al., Long-term decrease in Asian monsoon rainfall and abrupt climate change events over the past 6,700 years (2021)
Amy E. Hessl; Kevin J. Anchukaitis; Casey Jelsee; Benjamin Cook; Oyunsanaa Byambasuren; Caroline Leland; Baatarbileg Nachin; Neil Pederson; Hanqin Tian; Laia Andreu Hayles, Past and future drought in Mongolia (2018)
Neil Pederson; Amy E. Hessl; Nachin Baatarbileg; Kevin J. Anchukaitis; Nicola Di Cosmo, Pluvials, droughts, the Mongol Empire, and modern Mongolia (2014)
Jie Fei; Jie Zhou; Yongjian Hou, Circa A.D. 626 volcanic eruption, climatic cooling, and the collapse of the Eastern Turkic Empire (2007)
Nicola Di Cosmo; Clive Oppenheimer; Ulf Büntgen, Interplay of environmental and socio-political factors in the downfall of the Eastern Türk Empire in 630 CE (2017)
Nicola Di Cosmo; Amy Hessl; Caroline Leland; Oyunsanaa Byambasuren; Hanqin Tian; Baatarbileg Nachin; Neil Pederson; Laia Andreu-Hayles; Edward R. Cook, Environmental Stress and Steppe Nomads: Rethinking the History of the Uyghur Empire (744–840) with Paleoclimate Data (2018)
Rustam Ganiev; Vladimir Kukarskih, Climate Extremes and the Eastern Turkic Empire in Central Asia (2018)
Emre-E. Yavuz, The Gokturk Empire: 2nd Edition (2025)
Colin Mackerras, The Uighur Empire According to the T'ang Dynastic Histories (1972)

What links here18

Persons: Ilig Khan
States: First Göktürk Khaganate · Rouran Khaganate · Second Göktürk Khaganate · Uyghur Khaganate · Western Wei · Xueyantuo Khaganate (Syr Tardush)
Conflicts: Göktürk–Persian war, 588–589 · Tang conquest of the eastern Khaganate, 626–630 · The Turkic War of Independence · Uyghur–Kirghiz War
Battles & Events: The great snow of 627
Places: Altai · Orkhon valley
Peoples: Avars · Eurasian Nomads · History of the Turkic peoples
Institutions and Titles: Livestock Losses in the Annals

Cite this article

Yavuz, Emre-E. “Climate and the steppe empires.” Turkic Atlas, revised 8 November 2026. https://turkicatlas.org/historiography/climate-and-the-steppe-empires/

Last revised: 8 November 2026