In 1976, it was demonstrated that tiny wobbles in Earth's orbit led to the great ice-age cycles of the past few million years. This finding had wide implications for climate science and the details remain hotly debated today.
Forty years ago, Hays, Imbrie and Shackleton1 published one of the most influential papers in the study of Earth's past climate. The authors showed that variations in Earth's orbital path around the Sun were the pacemaker of the great ice ages in the Quaternary period (the current geological period, which began about 2.6 million years ago). Their study is a beautiful example of how three exceptional scientists worked together to robustly test a controversial theory, thereby allowing us to understand the fundamental causes of the ice ages and to peer into the future to see whether humanity has disrupted these natural cycles.
Milutin Milanković was a brilliant Serbian mathematician and climatologist who in 1941 postulated that wobbles in Earth's orbit changed the distribution of solar energy on the planet's surface, pushing Earth in or out of an ice age2 — a process called orbital forcing. He suggested that insolation (the amount of incoming solar radiation) at a latitude of 65 °N, just south of the Arctic Circle, was critical. At this latitude, insolation can vary by 25% (from 430 to 560 watts per square metre)2. Milanković argued that, when insolation was reduced during the summer, some of the ice in this region could survive. Each year, the ice would build up to eventually produce an ice sheet.
In 1976, three gifted scientists joined forces to test Milanković's theory, using long-term climate records obtained by analysing marine sediments. James Hays founded and led the international CLIMAP project, which used fossil assemblages to estimate past sea surface temperatures and brought the dream team together. Nick Shackleton was a master of stratigraphy and provided oxygen-isotope records, which showed past global ice volumes. John Imbrie was the expert in time-series analysis, particularly spectral analysis, which showed that the climate records matched calculations of high-latitude insolation. Remarkably, the authors discovered that these records contained the same temporal cycles — known as Milankovitch cycles — as three parameters that describe Earth's orbit: eccentricity, obliquity and precession (Fig. 1).
Eccentricity describes the shape of Earth's orbit around the Sun, and varies from nearly circlular to elliptical, in part because of Jupiter's gravity. Obliquity is the tilt of Earth's axis of rotation with respect to the plane of its orbit, and directly affects the intensity of the seasons. Finally, precession is the most complicated type of variation because it alters the distance between Earth and the Sun during each season, and has two components — Earth's rotational axis precesses (rotates) owing to tidal forces exerted by the Sun and the Moon on the solid Earth, and Earth's orbital path itself precesses around the Sun.
By demonstrating the clear link between orbital forcing and Earth's past climate, Hays, Imbrie and Shackleton legitimized what was to become one of the most powerful tools in stratigraphy. For example, reliable age models have been constructed for climate records covering at least the past 5 million years by tuning the orbital parameters to the ice-age cycles3. Such age models can be applied to any long-term palaeoclimate record, allowing marine and land records to be compared.
In addition, the various effects of the three orbital parameters have been used to study orbital forcing at different latitudes. Obliquity has a strong influence at high latitudes, whereas precession has a significant impact on seasonality in the tropics — precession has been linked to the rise and fall of the African rift-valley lakes, and even to our own evolution4. Evidence for the orbital forcing of climate has now been found as far back as 1.4 billion years ago, in the Proterozoic eon5.
Hays, Imbrie and Shackleton clearly set out the limitations of their study and presented the scientific community with a range of challenges, many of which remain today. In particular, the authors recognized that variations in the orbital parameters did not cause the ice-age cycles, but rather paced them. Any given combination of parameters can be associated with many different climates — for example, Earth's orbital configuration today is similar to to that of 18,000 years ago, when a 3-kilometre-thick ice sheet covered North America. Feedback mechanisms take the small changes in insolation that are driven by the orbital parameters and push Earth into or out of an ice age. Therefore, the next step was to understand the relative importance of the feedbacks involving the ice sheets, oceans and atmosphere; this led to the discovery that greenhouse gases had a pivotal role in controlling past climate.
The authors' work also recognized what is known as the 100,000-year problem. Before 1 million years ago, ice ages occurred roughly every 41,000 years owing to variations in Earth's obliquity6. This makes climatological sense, because Earth's axial tilt directly controls how warm or cold the summers are in the Northern Hemisphere. But the past 8 ice-age cycles had a longer period of 100,000 years7, which is similar to the period associated with eccentricity. In terms of forcing, eccentricity is by far the weakest of the three orbital parameters. Therefore, if eccentricity were responsible for such 100,000-year cycles, there would need to be a complicated 'nonlinear' amplification effect by Earth's climate system.
However, the similarity between the two periods turned out to be an artefact of spectral analysis6 — although the previous 8 ice-age cycles lasted for about 100,000 years on average, they ranged in length from 80,000 to 120,000 years. With the realization that eccentricity is not the major driving force, a debate has emerged as to whether precession or obliquity controlled the timing of the most recent ice-age cycles. Some researchers argue that the deglaciations occurred every four or five precessional cycles6,7, others that it was every second or third obliquity cycle8, and some argue that it was a combination of the two9. The debate started 40 years ago and still rages today.
The authors' work also provided a tool with which to investigate the future of Milankovitch cycles. It has been suggested that small increases in greenhouse gases due to the expansion of agriculture that started 8,000 years ago10 have, in fact, delayed the next ice age11. Moreover, if greenhouse-gas emissions continue to grow, the next ice age might be postponed for at least half a million years12. Understanding orbital forcing is therefore relevant to contemporary debates about the Anthropocene — a proposed geological unit that is defined by human activity. If we have merely delayed the next ice age, we will still be in the Quaternary period, and the Anthropocene can be defined as an epoch, a subdivision of time below a period. But if we have stopped the ice ages, we will have entered the Anthropocene period13, marking a much larger change in Earth's climate system. Therefore, unravelling the causes of the great ice ages is pivotal to understanding both the past and our future.