Modeling the timing of Patagonian Ice Sheet retreat in the Chilean Lake District from 23–10 ka

Abstract. Studying the retreat of the Patagonian Ice Sheet (PIS) during the last deglaciation represents an important opportunity to understand how ice sheets outside the polar regions have responded to deglacial changes in temperature and large-scale atmospheric circulation. At the northernmost extension of the PIS during the last glacial maximum (LGM), the Chilean Lake District (CLD) was influenced by the southern westerly winds (SWW), which strongly modulated the hydrologic and heat budget of the region. Despite progress in constraining the nature and timing of deglacial ice retreat across this area, considerable uncertainty in the glacial history still exists due to a lack of geologic constraints on past ice margin change. Where the glacial chronology is lacking, ice sheet models can provide important insight into our understanding of the characteristics and drivers of deglacial ice retreat. Here we apply the Ice Sheet and Sea-level System Model (ISSM) to simulate the LGM and last deglacial ice history of the PIS across the CLD at high spatial resolution (450 meters). We present an ensemble of LGM ice sheet model experiments using climate inputs from the Paleoclimate Modelling Intercomparison Project (PMIP4) and a transient simulation of ice margin change across the last deglaciation using climate inputs from the CCSM3 Trace-21ka experiment. We find that although the simulated LGM temperature is primarily responsible for differences in simulated ice geometries, wintertime precipitation also plays an important role in modulating LGM ice sheet volume and extent. The simulated deglaciation is found to match existing geologic constraints that indicate widespread ice margin retreat between 18 to 16.5 ka. Following this interval our simulations suggest that the ice sheet retreated rapidly, and by 15 ka onward, only mountain glaciers remained across the CLD in contrast with sparse geologic data that indicate a local ice cap remaining until 10 ka. Additionally, our results suggest that modest variations in winter precipitation (~10 %) can modulate the pacing of ice retreat by 1–2 ka, which has implications when comparing simulated outputs of ice margin change to geologic reconstructions. Therefore, these LGM and deglacial experiments signify the importance in constraining the deglacial strength, latitudinal position, and extent of the SWW and its influence on the hydrologic and heat budget and also highlight the importance in constraining paleoclimate parameters critical to modelling and understanding the drivers of deglacial PIS behavior.