Modern to Late Pleistocene Stable Isotope Climatology of Alaska

• Main Author: Sloat, Alison
• Format: Thesis
• Language: unknown
• Published: University of Nevada, Las Vegas 2014
• Subjects: Bering Sea; Greenland; Gulf of Alaska; Yukon; Greenland ice core; Ice; ice core; Ice Sheet; North Atlantic; permafrost; wedge*; Alaska
• Online Access: https://dx.doi.org/10.34917/5836162; https://digitalscholarship.unlv.edu/thesesdissertations/2143

Understanding modern controls on climate is necessary to interpret past climatic conditions. This project investigated the modern controls on δ18O and δD values in Alaskan surface waters to interpret the controls on Late Pleistocene climate variability. ArcGIS was used to develop an isoscape of modern δ18O and δD values of over 400 surface water samples collected across Alaska and the Yukon. It was found that winter temperature and precipitation have the greatest controls on δ18O and δD values in Alaska, resulting in high δ18O values along the coast of the Gulf of Alaska and low values inland toward Central Alaska. This isoscape can be applied to paleoenvironmental, modern, and future records to determine isotope-precipitation-temperature values. This is a useful tool in determining paleotemperatures of Alaska which is necessary for the interpretation of the magnitude, timing, and patterns of past response to climatic change. Previous work suggested ancient preserved ice wedges in the CRREL Permafrost Tunnel formed syngenetically during Marine Isotope Stage 3 (MIS 3). However, a new method of determining the timing of ice wedge and pool ice formation using radiocarbon-dated DOC and CO2 reveals that the features are much younger than previously thought and are epigenetic in origin. At least five freeze events and one melt event are apparent in the tunnel, with likely ages between 26 and 12.9 cal ka BP, during MIS 2. The prevalence of inversions in the radiocarbon ages indicates that carbon may persist for as long as 17,000 years within permafrost environments, potentially providing anomalous age dates when dating carbon in permafrost. To better constrain the timing and climatic conditions during ice wedge formation, eight ice wedges and seven ice pools were sampled at high temporal resolution for stable isotopes and were combined with14C ages of DOC from within the ice to estimate the timing of formation and climatic conditions under which they formed. Four intervals of cooling and one interval of warming were recorded in the ice wedge and pool ice, with δ18O values ranging from -28.9 / to -20.4 /. These values range between -6.6 below and +1.9 / above the modern snow δ18O value of -22.3 /, suggesting that some of the ice wedges formed during colder-than-modern conditions. Paleo-winter temperatures ranged from -41.6 to -11.2°C, while paleo-mean annual temperatures (MAT) ranged from -14.1 to 1.4°C, consistent with paleotemperature fluctuations observed in the Bering Sea SST and Greenland ice core records. Based on a visual correlation to the Bering Sea SST record within the ice age limits described in Chapter 3, it is suggested here that Freeze event 1 likely occurred during cooling of -12.3 to -1.2°C between 26.5 - 25.3 cal ka BP, coinciding with Heinrich event 2, while Freeze event 2 occurred during cooling of -8.8 to -2.8°C between 21.0 - 20.5 cal ka BP. One warming interval, M1, when clear pool ice formed in a melt horizon above the F1 and F2 wedges, may have occurred as paleo-MAT ranged between -9.9 and -1.2°C between 19.1 - 18.8 cal ka BP. Freeze event 3 represents the lowest (coldest) ?18O values. These values suggest that paleo-MAT ranged between -14.1 and -9.2°C, which is correlated to the coldest regional climate interval when the Laurentide Ice Sheet reached its local maximum extent in the Yukon between 17.3 - 17.0 cal ka BP, coinciding with Heinrich event 1. Freeze event 4 has high ?18O values that are similar to those of a late Holocene ice wedge in the nearby Vault Creek permafrost tunnel; that wedge was radiocarbon-dated to ca. 3.9 cal ka, and represents the warmest ice-wedge-forming thermal event when paleo-MAT likely ranged between -1.4 and 1.4°C. The combination of low-resolution and low-fidelity radiocarbon dating, high resolution δ18O data, and paleotemperature estimates allow for more robust age constraints when compared with the well-dated Bering Sea temperature record, suggesting that ice wedges in Central Alaska formed in response to North Atlantic DO- and H-type millennial forcing of climate.