STABLE ISOTOPIC EVIDENCE FOR MID-MIOCENE HYPERARIDITY IN THE ATACAMA DESERT

The onset of hyperaridity in the Atacama Desert has been variously estimated between Oligocene to Pliocene in age. In this study we analyzed paleospring travertine at Barrancas Blancas at the foot of the Pacific slope of the Andes. Ground-water in the area today is recharged in the adjacent high Western Cordillera of the Andes to the east and discharges as several small springs into an area of near rainless desert. The spring calcites are extensive both as horizontal sheets and as pure sparry linings of vertical fractures. U-Pb ages from the calcites sampled thus far span 2.08±0.04 to 10.5±0.6 Myrs. Reconstructed d¹⁸O (SMOW) water values (-4.6 to +0.8‰) are as high or higher than modern spring water values in the area today, assuming temperatures of formation of 10°C. Warmer assumed temperatures would make these estimates even more positive. Such high d¹⁸O ground-water values today probably reflect intense sublimation/ evaporation of snow and runoff in the recharge area, and constitute a fingerprint of hyperarid conditions. Hence we suggest that hyperaridity in the Atacama developed prior 10.5±0.6 Ma. This is consistent with previously published paleosol evidence for the onset of hyperaridity sometime in the mid-Miocene. Our results show that since the mid-Miocene significant easterly moisture has not penetrated the area, blocked by an elevated Andes to the east, nor westerly moisture from the Pacific, inhibited by descending mid-latitude air and a cold ocean. The development of hyperaridity predates the large ignimbrite additions (and possible mantle delaminination events?) to the western Cordillera at this latitude in the last 10 Myr.