James Croll (1821–1890) was a gifted scientist whose revolutionary theories had a profound impact on our understanding of the Earth's climate, ice ages and glaciation. While his contributions are recognised by an increasing number of modern-day scientists, in the public domain his legacy has been all but forgotten. Popularising Croll's story brings its own challenges: we know more about the science than we do about the man, and his theories do not lend themselves to quick and easy explanation. While Croll's scientific theories, presented alone, risk being viewed as complex and difficult to digest, his lifelong struggle against adversity is a compelling story with the potential for widespread public appeal. In recent years, interested individuals and institutions have begun to increase public awareness of Croll through talks and lectures, exhibitions, theatrical events and articles in print and online. There are many more possibilities that are worth investigating, in order to inspire and engage people locally, nationally and even internationally. Croll's story is fascinating from many points of view, and is open to interpretation by people of different ages and backgrounds. Perhaps even the gaps in our knowledge can be turned to advantage, allowing for imagination, creativity and expression.

James Croll was a pioneer in studies of the impact of the slowly changing orbital dynamics of the Earth on climate change. His book Climate and Time in their Geological Relations (1875) was far ahead of its time in seeking correlations between climate change, the occurrence of ice ages and perturbations to the Earth's orbit about the Sun. The astronomical cycles he discovered are now called ‘Milankovitch Cycles’ after the Serbian scientist whose research was first published in the Handbuch der Klimatologie in 1930. The celestial mechanical and astronomical background to Croll's research is the focus of this essay. The development of the understanding of the impact of perturbations of the elliptical planetary orbits by other bodies in the solar system paralleled new mathematical techniques, many of which were developed in association with celestial mechanical problems. The central contributions of many of the major mathematicians of the late 18th and 19th Centuries, including Euler, Lagrange, Laplace and Le Verrier, are highlighted. Although Croll's contributions faded from view for several generations, his pioneering insights have now been demonstrated to have been basically correct.

This paper examines the astronomical theory of ice ages of James Croll (1821–1890), its influence on contemporaries John Tyndall, Charles Lyell, and Charles Darwin, and the subsequent development of climate change science, giving special attention to the work of Svante Arrhenius, Nils Ekholm, and G. S. Callendar (for the carbon dioxide theory), and Milutin Milanković (for the astronomical theory). Croll's insight that the orbital elements triggered feedbacks leading to complex changes – in seasonality, ocean currents, ice sheets, radiative forcing, plant and animal life, and climate in general – placed his theory of the Glacial Epoch at the nexus of astronomy, terrestrial physics, and geology. He referred to climate change as the most important problem in terrestrial physics, and the one which will ultimately prove the most far reaching in its consequences. He was an autodidact deeply involved in philosophy and an early proponent of what came to be called ‘cosmic physics’ – later known as ‘Earth-system science.’ Croll opened up new dimensions of the ‘climate controversy’ that continue today in the interplay of geological and human influences on climate.

Global glaciations have varied in size and magnitude since the Early–Middle Pleistocene transition (~773 ka), despite the apparent regular and high-amplitude 100 ka pacing of glacial–interglacial cycles recorded in marine isotope records. The evidence on land indicates that patterns of glaciation varied dramatically between different glacial–interglacial cycles. For example, Marine Isotope Stages (MIS) 8, 10, and 14 are all noticeably absent from many terrestrial glacial records in North America and Europe. However, globally, the patterns are more complicated, with major glaciations recorded in MIS 8 in Asia and in parts of the Southern Hemisphere, such as Patagonia, for example. This spatial variability in glaciation between glacial–interglacial cycles is likely to be driven by ice volume changes in the West Antarctic Ice Sheet and associated interhemispheric connections through ocean–atmosphere circulatory changes. The weak global glacial imprint in some glacial–interglacial cycles is related to the pattern of global ice buildup. This is caused by feedback mechanisms within glacial systems themselves that partly result from long-term orbital changes driven by eccentricity.

The magnitude of late Wisconsinan (post-35,000 yr B.P.) ice-sheet growth in the Northern Hemisphere is not well known. Ice volume at ∼35,000 yr B.P. may have been as little as 20% or as much as 70% of the volume present at the last glacial maximum (LGM). A conservative evaluation of glacial–geologic, sea level, and benthic δ18O data indicates that ice volume at ∼35,000 yr B.P. was approximately 50% of that extant at the LGM (∼20,000 yr B.P.); that is, it doubled in about 15,000 yr. On the basis of literature for the North Atlantic and a sea-surface temperature (SST) data compilation, it appears that this rapid growth may have been forced by low-to-mid-latitude SST warming in both the Atlantic and Pacific Oceans, with attendant increased moisture transport to high latitudes. The SST ice-sheet growth notion also explains the apparent synchroneity of late Wisconsinan mountain glaciation in both hemispheres.

We applied magnetostratigraphy and mammal biostratigraphy to date climate-sensitive pollen cycles and lithostratigraphic units of the Pliocene–Pleistocene Leffe sedimentary succession from the Southern Alps, Italy. The Leffe section was correlated to additional sections (Casnigo, Fornaci di Ranica, and Pianengo) to construct a stratigraphic network along a common fluviatile system (the Serio River) sourced in the Southern Alps and flowing southward into the Po River Basin. We obtained a coherent scenario of climate variability for the last ∼ 2 Myr. At Leffe, lacustrine deposition commenced during the Olduvai Normal Subchron (1.94–1.78 Ma) and lasted up to a chronologic level compatible with Marine Isotope Stage (MIS) 22 (0.87 Ma). Pollen analysis revealed that climate varied cyclically from warm-temperate to cool during this time interval, but never as cold as during glacial intervals. At around MIS 22, climate cooled globally. Gravels, attributed to high-energy braided river systems fed locally by alluvial fans, prograded from the Serio River catchment area over the Leffe Basin and toward the Po Plain in response to a generalized event of vegetation withdrawal and enhanced physical erosion. At this time, Alpine valley glaciers reached their first maximum southward expansion with glacier fronts located at only ∼ 5 km upstream from Leffe.