T327 Paleoclimate Proxies and Evolving Human Landscapes

Wednesday, 30 October 2013
8:00 AM - 12:00 PM
Room 605

NEW EVIDENCE FOR LGM GLACIATION OF THE ARID CHAJNANTOR PLATEAU, NORTHERN CHILEAN ANDES

Paper number: T327-5
Program: Technical Sessions
Time: 8:50 AM – 9:10 AM

Authors

Abstract

The processes controlling climate change in the arid, subtropical Andes are of critical importance in understanding the long history of human subsistence in one of the driest regions on Earth. Some studies attribute increased moisture in the subtropical Andes to orbitally-forced changes in summer insolation over land, while others have argued that millennial-scale variability in Pacific SST gradients more strongly influences the regional climate. The two viewpoints are not mutually exclusive, but their reconciliation requires a detailed understanding of the hydroclimatic history of the subtropical Andes. In this region, the presence of glaciers in the past is a strong indicator of much wetter conditions at that time.

The history of Andean glaciation is poorly dated in the central Andes, and no direct dating of deposits on the Chilean Altiplano has been reported. We present the first cosmogenic 10Be exposure ages for glacial features on the Chajnantor Plateau (23ºS), an arid, presently unglaciated part of the Andes in the Chilean Altiplano. The distribution of fourteen new exposure ages from moraine boulders and glaciated bedrock implies that the most recent extensive glacial occupation here was during the global Last Glacial Maximum (LGM). The ages range from 7 ka to 117 ka, with clusters at 16-26 ka and 32-42 ka. These periods coincide with wet periods in drill core proxies from the nearby Salar de Uyuni (Bolivia) and Salar de Atacama (northern Chile). This suggests an orbital modulation of moisture delivery to the subtropical Andes from the eastern lowlands, possibly by changing the position or intensity of the South American Summer Monsoon. Previous glacier modeling has suggested that a 4-5x increase in precipitation is required to glaciate the study area. We present preliminary regional climate model output for this area at 21 ka, which supports a several-fold increase in snowfall in conjunction with a 4-5ºC temperature depression at that time relative to the modern climate.