Milankovitch Theory and climate

Among the longest astrophysical and astronomical cycles that might influence climate (and even among all forcing mechanisms external to the climatic system itself), only those involving variations in the elements of the Earth's orbit have been found to be significantly related to the long‐term climatic data deduced from the geological record. The aim of the astronomical theory of paleoclimates, a particular version of which being due to Milankovitch, is to study this relationship between insolation and climate at the global scale. It comprises four different parts: the orbital elements, the insolation, the climate model, and the geological data. In the nineteenth century, Croll and Pilgrim stressed the importance of severe winters as a cause of ice ages. Later, mainly during the first half of the twentieth century, Köppen, Spitaler, and Milankovitch regarded mild winters and cool summers as favoring glaciation. After Köppen and Wegener related the Milankovitch new radiation curve to Penck and Brückner's subdivision of the Quaternary, there was a long‐lasting debate on whether or not such changes in the insolation can explain the Quaternary glacial‐interglacial cycles. In the 1970s, with the improvements in dating, in acquiring, and in interpreting the geological data, with the advent of computers, and with the development of astronomical and climate models, the Milankovitch theory revived. Over the last 5 years it overcame most of the geological, astronomical, and climatological difficulties. The accuracy of the long‐term variations of the astronomical elements and of the insolation values and the stability of their spectra have been analyzed by comparing seven different astronomical solutions and four different time spans (0–0.8 million years before present (Myr B.P.), 0.8–1.6 Myr B.P., 1.6–2.4 Myr B.P., and 2.4–3.2 Myr B.P.). For accuracy in the time domain, improvements are necessary for periods earlier than 2 Myr B.P. As for the stability of the frequencies, the fundamental periods (around 40, 23, and 19 kyr) do not deteriorate with time over the last 5 Myr, but their relative importance for each insolation and each astronomical parameter is a function of the period considered. Spectral analysis of paleoclimatic records has provided substantial evidence that, at least near the obliquity and precession frequencies, a considerable fraction of the climatic variance is driven in some way by insolation changes forced by changes in the Earth's orbit. Not only are the fundamental astronomical and climatic frequencies alike, but also the climatic series are phase‐locked and strongly coherent with orbital variations. Provided that monthly insolation (i.e., a detailed seasonal cycle) is considered for the different latitudes, their long‐term deviations can be as large as 13% of the long‐term average, and sometimes considerable changes between extreme values can occur in less than 10,000 years. Models of different categories of complexity, from conceptual ones to three‐dimensional atmospheric general circulation models and two‐dimensional time‐dependent models of the whole climate system, have now been astronomically forced in order to test the physical reality of the astronomical theory. The output of most recent modeling efforts compares favorably with data of the past 400,000 years. Accordingly, the model predictions for the next 100,000 years are used as a basis for forecasting how climate would evolve when forced by orbital variations in the absence of anthropogenic disturbance. The long‐term cooling trend which began some 6,000 years ago will continue for the next 5,000 years; this first temperature minimum will be followed by an amelioration at around 15 kyr A.P. (after present), by a cold interval centered at 23 kyr A.P., and by a major glaciation at around 60 kyr A.P.