Résumé
Andean ice cores extracted from high attitude glaciers have provided a wealth of original information dealing with past climate variability both in tropical and subtropical South America. In this chapter, we focus on the transition between the Last Glacial Maximum (LGM) and the Holocene period as recorded by the isotopic composition of the ice (oxygen 18, delta O-18 or deuterium, delta D) of three Andean ice cores spanning at least the last similar to 20,000 years (Sajama, 18 degrees S; Illimani, 16 degrees S; Huascaran, 9 degrees S). We discuss the uncertainties of Andean ice cores dating and compare the three Andean records together and with the isotopic composition of Greenland and Antarctic ice cores (NorthGRIP, 75 degrees N; EPICA Dome C, 75 degrees S). We attribute the common glacial-interglacial isotopic composition increase (delta O-18 similar to 4-6 parts per thousand) recorded in Andean ice cores to the enhanced net precipitation lost by airmasses (similar to+5-7%) at LGM, from the moisture oceanic source to the Andean summits. We discuss the roles of precession and glacial boundary forcings that have been pointed out so far to explain wet conditions in the southern tropics of South America during glacial climate. We also spot that along the deglaciation, strong discrepancies exist between the Sajama ice core and the two other Andean ice cores. A Greenland-like deglacial progression can be observed at Sajama site with a rapid return to near-glacial conditions at around 14 ka, mimicking the Younger Dryas event, whereas an Antarctic-like deglaciation pattern can be identified both at Illimani and Huascaran sites bearing intriguing similarities to the Antarctic Cold Reversal. We discuss different hypotheses including local climate and Pacific influence to examine the possible causes of those discrepancies during the deglaciation.