From the astronomical theory of paleoclimates to global warming

The climate predicted to occur over the next centuries by the Intergovernmental Panel on Climate Change appears to be unprecedented over the last 150 years. This requests therefore to go back in the past history of the Earth looking for analogues. As we are presently in an interglacial (the Holocene), the interglacials of the late Pleistocene are particularly well suited. This is why we have investigated the response of the climate system to insolation and CO2 at the peaks of the interglacials over the past 800,000 years using models of different complexity. These simulations show that the relative contributions of insolation and CO2 to the intensity and duration of each interglacial vary from one interglacial to another. They also show that CO2 plays a dominant role on the variations of the global annual mean temperature and of the southern high latitude temperature and sea ice, whereas, insolation plays a dominant role on the variations of monsoon precipitation, vegetation and of the northern high latitude temperature and sea ice. In the explanation of the warmer climate during the interglacials after about 430,000 years ago than those before, boreal winter is found to be a key season, a phenomenon similar to the present-day global warming. If we identify the peaks of the interglacials with Northern Hemisphere summer occurring at perihelion, MIS-1, MIS-11 and MIS-19 (respectively 12, 409 and 788 thousands of years ago - ka) show a pretty similar latitudinal and seasonal distribution of the incoming solar radiation (insolation). When compared to the average of the last 9 interglacials, they are under-insolated over the whole globe during boreal summer and are over-insolated during boreal winter with a maximum at the South Pole. This insolation distribution leads to a cooling over all the continents in boreal summer and to a warming over the whole Earth, except the Arctic, in boreal winter. A warming over the Southern Ocean in austral winter occurs during MIS-1 and MIS-19 due to the summer remnant effect of insolation. However, this does not happen in MIS11 because the large global cooling during this season is dominating the remnant effect of the austral summer. This leads to MIS-11 being a cool insolation-induced interglacials and thus not as good an analogue of MIS-1 as MIS-19, at least as far as insolation is concerned. Its higher CO2 concentration allows it however to be finally classified among the warm interglacials and as such to be compared to MIS-1. Looking now for analogues of the whole Holocene and its future (the Anthropocene), it must be stressed that the shape of the Earth's orbit is approaching a circle. With this and a CO2 concentration at the interglacial level, and even larger under human influence, our interglacial was predicted to be exceptionally long as was MIS-11 in EPICA record. The interglacials MIS-9 and MIS-5 (respectively 334 and 127 ka ago) are the warmest interglacials and, as such, are considered as analogues for our CO2-induced future warm interglacial, but their astronomical forcings are largely different from MIS-1 and its future. The results also show that, compared to today, the past interglacials are warmer during boreal summer and cooler during boreal winter leading to a warmer annual mean with varying length for different interglacials. The best analogue to MIS-1 depends therefore upon the criteria used to select such an analogue. If we compare the climate of these interglacials to the climate of a 2xCO2 atmospheric concentration simulated under Pre-Industrial conditions, both during boreal winter are cooler over the whole Earth, especially over the continents. During boreal summer, they are much warmer over the continents, but remain cooler over the oceans. This shows the importance of the insolation which is much larger at MIS-5 and MIS-9 (even at MIS-11 and MIS-19) with boreal summer occurring at perihelion than in the next future with boreal summer occurring at aphelion. By all means, this sensitivity of climate to the latitudinal and seasonal distribution of insolation must be kept in mind for the climatic projection at the century-millennium time scales. These results underline the diversity of the warm climates of the past one million years and therefore the potential but also the difficulty to find exact analogs for our interglacial and its future