CLUMPED ISOTOPE THERMOMETRY CONSTRAINTS ON LATE PLEISTOCENE HYDROCLIMATES FOR LAKE SURPRISE, CALIFORNIA

The response of regional precipitation and evaporation rates to changing climate forcings are a major uncertainty. We use a novel approach, clumped isotope thermometry of lake carbonates, in concert with terminal basin water balance modeling to constrain past temperatures and place controls on the water balance of late Pleistocene Lake Surprise during the Last Glacial Maximum (LGM) and the subsequent deglaciation. We analyzed previously dated shoreline tufa from six localities spanning 50 to 200 m of lake level change across the eastern side of the valley. Preliminary carbonate clumped isotope analyses revealed a warm month mean water temperature depression of 10.2 ± 2.2 °C (1 SE, n=15) during the LGM and 5.5 ± 2.5 °C (1 SE, n=5) during Heinrich Stadial 1 (HS1), relative to modern conditions associated with lake carbonate precipitation. These temperature changes should be analogous to changes in summer air temperatures, and place minimum constraints on changes in mean annual temperature. Using the clumped isotope derived temperatures and reconstructed lake water oxygen isotope compositions, we model lake evaporation rates, basin average precipitation rates, and relative runoff into the lake. We test the sensitivity of our water balance calculations to several different lake water evaporation models and modeling assumptions. Initial work indicates that during the LGM precipitation rates were 115 ± 6% (1 SE, n=15) of modern and HS1 precipitation rates were 129 ± 4% (1 SE, n=5) of modern. The LGM temperature depression at Surprise Valley simulated by the Paleoclimate Modeling Intercomparison Project 3 (PMIP3) model ensemble mean is -7.0 °C (with an ensemble range of -3.9 to -9.8 °C, n=9), and the annual precipitation rate change is 107% (85 to 150%) of modern, in broad agreement with our clumped isotope based hydroclimate estimates. Our approach demonstrates a promising method to use lacustrine records for quantitative terrestrial hydroclimate reconstructions, and to assess the accuracy of climate model simulations.