When Did the Great Acceleration Start? Saskatchewan Might Hold the Answer

Scroll this

This is the eighth post in a series about the Great Acceleration as a framework and reconnaissance for Canadian environmental history. The posts in this series are cross-posted with Active History.

When did the Great Acceleration start? Saskatchewan might hold the answer. Between the 1890s and the 1930s, the settler population exploded, and these newcomers broke 20 million acres of prairie grassland into wheat farms. The transformation released vast quantities of CO2 held in the soil and was inseparable from the genocidal dislocation of Indigenous people from their land.1 Saskatchewan’s agricultural transformation coincided with settlement on the Great Plains in the United States, the pampas in Argentina, the cocoa boom in the Gold Coast (modern Ghana), the sugar boom in Fiji and the Philippines, and rubber booms in Brazil, Ceylon and the Congo Free State. This global “golden age” of resource-led development transformed ecosystems across the globe and contributed to the early stages of anthropogenic climate change.2 It all took place decades before the conventional 1950 start of the Great Acceleration.3

Twenty-four charts showing the rapid increases in human activity across a number of indicators, including land use, water use, population, atmospheric concentration of carbon dioxide, and so on.
Figure 1. The Great Acceleration dashboard. Twenty-four global trends — twelve socio-economic, twelve Earth System — plotted from 1750 to 2010, with the post-1950 take-off that gives the Great Acceleration its name (Steffen et al. 2015). Image: via Courtney White, “What Is Earth For?,” The Grass Canoe/Resilience.org, August 19, 2019.

Will Steffen used simple data visualizations to make a compelling case the Great Acceleration started in the mid-twentieth century.4 A large number of socio-economic and earth systems trends appeared to reach an inflection point in the decade after the end of the Second World War. There were regional examples of rapid change in the preceding centuries, such as the Industrial Revolution in Great Britain, but the mid-twentieth century saw this acceleration spread to the rest of the globe.

The problem with this method is that we can create very similar charts with a different x axis that suggest the Great Acceleration was well underway by the end of the Second World War.5 Global population grew slowly for millennia, then jumped from 1 billion in 1800 to almost 2.5 billion in 1950 (Our World in Data). This process continues in the decades that followed, but it is hard to say the inflection point comes in 1950 instead of 1850. It is also hard not to link most of the other changes in Steffen’s charts to the accelerating population growth that dates back centuries. We can also look at CO2 emissions and built up areas (the x axes to range from 1750 to 1950) and find clear inflection points around 1850; agricultural area and population follow a similar trend through this period and temperate forest in England and the United States reach their low points and start to rebound before 1950. We could continue to collect datasets and find a dramatic increase in intercontinental shipping or global kilometers of railway tracks to create a dashboard that starts to resemble Steffen’s but identifies 1850 as the inflection point.

Graphs showing CO2 emissions, CO2 emissions from coal, agricultural area, built-up area, and population from 1850, all of which go upwards drastically after 1850; last graph shows forest cover in the United States, China, and England, all of which chart downwards until a small rebound upwards in the U.S. and England.
Figure 2. Six global indicators of human activity, 1750–1950. CO₂ emissions (land use in green, fossil + cement in grey); CO₂ from coal; agricultural land; built-up area; population; and forest cover for the United States, China, and England. Every panel inflects well before the Great Acceleration’s conventional 1950 start (Steffen et al. 2015); the shaded band marks 1850–1950, the “inflection century.” Through 1900, land-use change supplied roughly two-thirds of all human CO₂ emissions. Data: Global Carbon Budget v15 (Friedlingstein et al. 2025) and HYDE 3.5 (Klein Goldewijk et al. 2017), via Our World in Data; pre-1850 land-use CO₂ is the author’s HYDE-calibrated extension; US forest cover from Williams (1989) and USDA Forest Service (2014), China from He, Yang & Wang (2024), England from Our World in Data after Rackham (1986).

Defenders of the mid-century date have a real point: 1950 is when human influence on earth systems became measurable across a wide range of indicators. But that is a claim about legibility, not the start of the Great Acceleration, and the dashboard is a little guilty of choosing the data that fit it. Steffen’s earth-systems panel charts the loss of tropical forest, a curve that necessarily looks like a twentieth-century phenomenon, because as we saw above, in 1900 the temperate forests of Europe and eastern North America were already near their low point. Picking tropical forest doesn’t tell us when deforestation accelerated; it tells us which forests were still standing once the dataset begins. Run the same panel on temperate forest clearance or the decline of prairies ecosystems and the curve peaks earlier. The indicators on the dashboard are the ones still in motion after the war; the violent transformations that came before are simply not plotted.

We can focus on global warming to flesh out the issues with marking 1950 as the start of the Great Acceleration. G.S. Callendar had already recorded the early signs of global warming by 1938. Using long-running weather station records, he found world temperatures had risen about 0.005°C per year over the preceding half-century, and he attributed much of the rise to fossil-fuel carbon dioxide.6 Modern reconstructions confirm his estimate of roughly 0.3°C of warming was remarkably accurate.7 Anthropogenic climate change was underway well before 1950, even though it took until the 1980s for scientists to confidently distinguish the warming from natural climatic variation.

Line graph showing rise in global mean surface temperature, showing Callendar's instrumental confirmation of CO2-based warming in 1938, and Steffen et. al.'s 1950 date of the Great Acceleration's conventional "start" date.
Figure 3. Global mean surface temperature, 1850–2024, with Callendar’s 1938 measurement marked. By 1938, when Callendar reported roughly 0.3 °C of warming from 147 station records (Callendar 1938), the smoothed series already shows +0.33 °C since the 1850s — an estimate reconfirmed seventy-five years later (Hawkins and Jones 2013). The Great Acceleration’s conventional clock starts twelve years after Callendar’s paper. Data: Berkeley Earth Land + Ocean anomaly relative to 1951–1980 (Rohde and Hausfather 2020); shaded band marks 1850–1950 as in Figure 2.

Until roughly 1950, the largest source of human carbon emissions was not fossil fuel but land-use change: the clearing of forests and the plowing-up of grasslands. Fossil fuels only overtook land conversion as the dominant source around the middle of the twentieth century. So the famous take-off in the CO2 curve at 1950 is not the moment humans began to force the carbon cycle. CO2 emissions are cumulative and all the CO2 released by agriculture and forestry combined with the accelerating combustion of fossil fuels.

Line graph showing annual global CO2 emissions from land change use, showing a peak at 1959 and slow downward trend since.
Figure 4. Annual global CO₂ emissions from land-use change, 1750–2024. Land-use emissions roughly tripled across the shaded 1850–1950 band and peaked at 8.7 Gt CO₂ per year in 1959 — nine years after the Great Acceleration’s conventional clock starts (Steffen et al. 2015). By the time the framework dates “the start,” this flux was not accelerating; it was peaking. Data: Global Carbon Budget v15 bookkeeping reconstruction (Friedlingstein et al. 2025; Hansis, Davis & Pongratz 2015), via Our World in Data; before the dotted line at 1850, the author’s extension from HYDE 3.5 agricultural area (Klein Goldewijk et al. 2017).

The Global Carbon Budget’s reconstruction confirms that land-use change, not fossil fuel, was the major source of global CO2 emissions before the mid-20th century, when fossil fuels became the primary driver. The land-use data do not match the exponential growth curves in Steffen’s charts. Instead we see a doubling of emissions in the century after 1750, roughly a tripling in the century between 1850 and 1950, and a final postwar surge that peaked in 1959 before beginning a long uneven decline.8

Chart showing national shares of global annual land use CO2 emissions from 1850 to 2024.
Figure 5. National shares of global annual land-use CO₂ emissions, 1850–2024. Fifteen focal countries as fractions of the global flux; the residual is “Rest of world.” Through the shaded 1850–1950 window the United States, Russia, China, and India supply over half the signal; after 1950 the flux flips to the tropical-forest frontier of Brazil, Indonesia, and DR Congo. The prairie/steppe land rush — including the Saskatchewan and Great Plains wheat boom — was the defining land-use carbon signal of its era. Data: Global Carbon Budget v15 by country, via Our World in Data; sources as in Figure 4.

Focusing on Canada, there were two accelerations in CO2 emissions from land use: the Prarie boom circa 1896-1931 and World War 2 through to 1960. The remainder of this post will focus on Canada’s Great Acceleration recorded in the first boom in land-use CO2 emmissions.

Line graph showing Canadian land-use CO2 emissions, showing a sharp upsurge from 1896 to 1930, followed by another sharp upsurge after World War II, followed by steady decline after 1959.
Figure 6. Annual Canadian CO₂ emissions from land-use change, 1850–2024. The shaded band marks 1896–1931, the Saskatchewan wheat-boom window, during which annual emissions doubled from 184 Mt CO₂ (1900) to 376 Mt (1919). The all-time peak — 470 Mt — falls in 1959, late in the postwar expansion that pushed the wheat economy north into the boreal margins. The abrupt steps at 1899–1900 and 1949–1950 likely reflect discontinuities in the bookkeeping model’s input datasets rather than real one-year events. Data: Global Carbon Budget v15, via Our World in Data; sources as in Figure 4.

John Sandlos has already made the case that Canada experienced a great acceleration in mining starting in the last decades of the 19th century and coinciding with electrification and urbanization.9 I want to build on his case and add that Canada was a part of a global land rush that transformed temperate grasslands and tropical rain forests into wheat farms, sheep stations, rubber plantations and cocoa farms in the later nineteenth and early twentieth century. This was not a new process and we can track agricultural colonization across the Holocene. What makes this period different was the speed, scale and global synchronicity of the transformation.

The carbon record traces that synchronicity. The Global Carbon Budget’s reconstruction of land-use emissions shows the American clearances peaking around 1880 (by which point the United States alone was releasing more than a billion and a half tonnes of CO2 a year from land conversion) and then, as that boom subsided, a second wave igniting almost at once around the temperate grasslands of the southern hemisphere and the northern plains. Canada, the Argentine pampas and the Australian rangelands all surge between 1900 and the 1920s; Argentina’s land-use emissions jump roughly eightfold across those decades (data from the same Global Carbon Budget country data behind Figure 5). Saskachewan’s boom was the Canadian component of a global process: same plow, the same handful of decades, on four continents at once.

Line graphs showing Saskatchewan's field-crop area, wheat production, livestock production, railway track in service, population, and reserve land surrenders, all of which experience rapid upsurges from 1900 onward (with plateaus occurring at distinct time periods between each graph).
Figure 7. Saskatchewan, 1880–1950: six indicators of a pre-1950 Great Acceleration. All six series — field crops, wheat, livestock, railways, population, and reserve land surrendered by First Nations — inflect and plateau within the shaded 1896–1931 window. The settler S-curve and the dispossession S-curve are the same curve viewed from two sides. Data: annual agricultural series from the author’s research dataset; census population from HGIS Canada; Historical Canadian Railroads GIS; surrenders from Martin-McGuire (1998).

Starting in the last few years of the nineteenth century, hundreds of thousands of settlers arrived in Saskatchewan by railway and began transforming the prairie ecosystem into a blanket of wheat. Charting the kilometers of railway track, acres of field crops, population and livestock all confirm an acceleration in Saskatchewan. The wheat production chart records both the acceleration and then the collapsing yields during the drought years in the 1930s. And every acre broken was a carbon release. Prairie grassland holds most of its carbon underground, in soils built over millennia, and the plow that turned it into wheat fields oxidized that store into the atmosphere. In the Global Carbon Budget’s reconstruction, Canada’s land-use emissions more than doubled between 1890 and 1920, the window when the prairie was broken. Saskatchewan’s transformation was not only a regional ecological event but a contribution to the very global carbon signal Steffen’s dashboard dates to 1950, released decades before the dashboard begins to read it.

This acceleration also depended on clearing the land of its people. Through the 1880s and 1890s, as starvation and disease brought mass death to Plains peoples confined to reserves, the Canadian government responded not with relief but by negotiating large land surrenders. Some 400,000 acres of reserve land were given up in the aftermath of this genocidal Starvation Policy.10 The prairie was broken for wheat only after it had been emptied of the people who had lived on it.

Saskatchewan and the Canadian prairies were a continuation of a process that started in the United States, making it difficult to pin down a key decade in this earlier stage in the Great Acceleration. The American experience was a little more linear as the booms moved from state to state in the decades after the end of the Civil War. The Canadian and American processes share a peak in the early 1930s before the Dust Bowl forced an end to the constant growth in cropland.

Line graphs showing Great Plains cropland, population, railway construction, and livestock, all of which experience rapid upward growth from the 1880s to 1930s, when cropland and livestock decline. Population continues to grow.
Figure 8. The U.S. Great Plains, 1865–1950: four indicators of a continental Great Acceleration. Cropland, population, railway miles, and livestock across the twelve Great Plains states defined by Cunfer’s On the Great Plains, with the same 1896–1931 band as Figure 7. The Great Plains story runs roughly thirty years ahead of Saskatchewan’s — the steepest growth in every indicator predates the Saskatchewan window — and both regions peak together in the dust-bowl decade. The Saskatchewan wheat boom was a late entrant to a continental land rush whose first wave had already crested on the U.S. plains. Data: county-level agricultural census compilation underlying Cunfer (2005); population from Cao and Richardson (2023); railway miles from Railroads and the Making of Modern America, University of Nebraska-Lincoln.

My goal here isn’t to upend the Great Acceleration framework. I’m a fan of Steffen’s work and I find it useful in teaching and thinking about Canadian history. But I do not think global history fits into such a neat periodization. We can leave the battle over golden spikes that do or do not mark the start of the Anthropocene to the geologists. Historians focus on the continuity, the ruptures and how they relate. Mineral extraction and settler colonial transformations of the prairies in Canada were a part of global transformations that accelerated in the late nineteenth and early twentieth centuries. These processes snowballed into further economic development, industrialization, population growth and mass urbanization in postwar Canada. I’m not sure we need to settle on a definition as long as we contextualize the rapid change in the postwar period with the histories of other periods of rapid change in Canadian history and how they fit into larger global processes.

Notes

1. Robert Alexander Innes, “Historians and Indigenous Genocide in Saskatchewan,” Shekon Neechie, June 21, 2018, https://shekonneechie.ca/2018/06/21/historians-and-indigenous-genocide-in-saskatchewan/.

2. The phrase is Edward B. Barbier’s: Scarcity and Frontiers: How Economies Have Developed Through Natural Resource Exploitation (Cambridge University Press, 2011), 368.

3. J. R. McNeill and Peter Engelke, The Great Acceleration: An Environmental History of the Anthropocene since 1945 (Belknap Press of Harvard University Press, 2016).

4. Will Steffen, Wendy Broadgate, Lisa Deutsch, Owen Gaffney, and Cornelia Ludwig, “The Trajectory of the Anthropocene: The Great Acceleration,” The Anthropocene Review 2, no. 1 (2015): 81–98, https://doi.org/10.1177/2053019614564785.

5. Other scholars have questioned the statistical validity of Steffen’s charts: Ron W. Nielsen, “Mathematical Analysis of Anthropogenic Signatures: The Great Deceleration,” arXiv preprint arXiv:1803.06935 (2018), https://arxiv.org/pdf/1803.06935. Nielsen’s analysis of the Great Acceleration dataset finds no abrupt acceleration around 1950: the curves follow smooth growth trajectories reaching back centuries.

6. G. S. Callendar, “The Artificial Production of Carbon Dioxide and Its Influence on Temperature,” Quarterly Journal of the Royal Meteorological Society 64, no. 275 (1938): 223–240, https://doi.org/10.1002/qj.49706427503.

7. Ed Hawkins and Phil D. Jones, “On Increasing Global Temperatures: 75 Years after Callendar,” Quarterly Journal of the Royal Meteorological Society 139, no. 677 (2013): 1961–1963, https://doi.org/10.1002/qj.2178.

8. Louise Chini, George Hurtt, Ritvik Sahajpal, Steve Frolking, Kees Klein Goldewijk, Stephen Sitch, Raphael Ganzenmüller, Lei Ma, Lesley Ott, Julia Pongratz, and Benjamin Poulter, “Land-Use Harmonization Datasets for Annual Global Carbon Budgets,” Earth System Science Data 13, no. 8 (2021): 4175–4189, https://doi.org/10.5194/essd-13-4175-2021.

9. John Sandlos, “Mining Data and Canada’s Great Acceleration,” NiCHE: Network in Canadian History & Environment, May 20, 2026, https://niche-canada.org/2026/05/20/mining-data-and-canadas-great-acceleration/.

10. On the starvation policy and its death toll, see James Daschuk, Clearing the Plains: Disease, Politics of Starvation, and the Loss of Aboriginal Life (University of Regina Press, 2013). The figure of roughly 400,000 acres surrendered is from Peggy Martin-McGuire, First Nation Land Surrenders on the Prairies, 1896–1911 (Indian Claims Commission, 1998).

Feature image: Steam traction engines breaking sod, Western Development Museum Collection.

NiCHE encourages comments and constructive discussion of our articles. We reserve the right to delete comments that fail to meet our guidelines including comments under aliases, or that contain spam, harassment, or attacks on an individual.