Reconstructing climate during the Paleocene-Eocene Thermal Maximum from soil carbonate stable isotopes in the Green River Basin…

The Paleocene-Eocene Thermal Maximum (PETM) represents a period of abrupt global warming around 56 Ma where temperatures increased between 5 and 9 degreesC. It is globally recognized by a negative carbon isotope excursion (CIE) attributed to rapid greenhouse gas emissions and is considered an analog for modern anthropogenic climate change. While the majority of PETM paleoclimate data come from marine environments and suggest general global warming, continental interiors are relatively understudied, especially with respect to hydroclimate variability. Available data from continental interiors rely heavily on only a handful of sites that may not accurately reflect regional heterogeneities in temperature and hydroclimate. Here, I add to the spatiotemporal resolution of the PETM in North America's Laramide basins by generating a new paleoclimate record based on soil carbonates from the Beaver Creek section of the Green River Basin near present day La Barge, Wyoming (U.S.A.). Previously published soil carbonate stable carbon isotope (δ13Ccarb) data show a CIE, possibly associated with the PETM. Complementary stable oxygen isotope data (δ18Ocarb) show an increase during the main body of the CIE that could reflect evaporative enrichment in 18O of formational soil waters (δ18Osw), but could also reflect warming. To resolve this, I present new clumped isotope (Δ47) based temperatures of select stratigraphic intervals. These data reconstruct the formational temperatures of carbonate nodules, and when combined with δ18Ocarb, allow δ18Osw to be calculated which is then used to reconstruct paleohydrologic conditions. Results show that the timing of PETM warming in the Green River Basin generally parallels that of the Bighorn Basin, but may show a greater warming magnitude. This may indicate the Beaver Creek soil carbonates precipitated during peak summer temperatures or are more sensitive to kinetic isotope effects (KIEs), while Bighorn Basin soil carbonates precipitated later in the fall or under conditions closer to isotopic equilibrium. Meanwhile, calculated δ18Osw is higher in the Beaver Creek samples than the Bighorn Basin across all overlapping times considered in this study, possibly indicating greater evaporation, but also potentially differences in elevation between the two sites. I find contrasting hydrologic responses between the Green River and Bighorn Basins during peak PETM conditions, with the Green River Basin showing wetter conditions while the Bighorn Basin trends towards drier, which could represent a threshold response to subtropical precipitation. Lastly, I present evidence for more intra-annual variability in frequency, and intensity of rainfall in the Green River Basin during the PETM. Comparison with model simulations shows that the greatest agreement occurs when soil carbonates are assumed to have a strong seasonal bias and models have CO2 concentrations that are generally higher than proxies suggest. Furthermore, I present new paleomagnetic data, with the intent of constraining the timing of isotopic trends. Unfortunately, establishing paleomagnetic stratigraphic constraints was complicated by a pervasive remagnetization. A tightly clustered low temperature trend likely reflects chemical remagnetization, although the timing of this event could not be fully constrained. A high temperature component with varying degrees of confidence likely represents the primary paleomagnetic signature. Broadly though, inclination shows a pattern of predominantly reversed polarity that is consistent with the observed CIE being the PETM and not another Eocene hyperthermal. However, it should be further noted that the scattering in these values is large and both declination and inclination are inconsistent with what would be expected for the late Paleocene to early Eocene in southwestern Wyoming. While paleomagnetic data yield ambiguous results, the overall negative inclinations are largely consistent with previous chronostratigraphic constraints that interpret the observed CIE to be the PETM.