PROGRESS TOWARD SIMULATION OF WARM SOUTHERN HEMISPHERE HIGH-LATITUDE CLIMATE DURING PAST WARM CLIMATES (Invited Presentation)

The simulation of warm Southern Hemisphere high-latitude conditions during the extreme climates of the Late Cretaceous and Early Eocene has been a long-standing and persistent challenge. Climate models have consistently simulated surface temperatures that are substantially cooler than proxy temperature estimates. This discrepancy has been attributed to shortcomings in the models including missing or incorrect physics, uncertainties in the proxies, and underestimation of past greenhouse gas concentrations.

Here, we have developed new simulations of the Late Cretaceous and Early Eocene for a range of CO₂ concentrations (1-9x pre-industrial levels) using an isotope-enabled version of the Community Earth System model version 1.2 (CESM1.2). CESM1.2 exhibits a high climate sensitivity (4.2 °C) that increases with warming due to low-cloud feedbacks. As a result, in comparison to previous models, our simulations have warmer high-latitude surface temperatures and generally show much better agreement with Southern Hemisphere high-latitude temperatures inferred from TEX₈₆, Δ47, Mg/Ca, and δ¹⁸O thermometry.

The simulation of δ¹⁸O in CESM1.2 allows a direct comparison between the model and ancient foraminiferal and fish δ¹⁸O. The use of simulated δw to estimate δ¹⁸O paleotemperatures yields a better model-data agreement than other commonly used assumptions. As in earlier isotope-enabled models of the Cretaceous, our CESM1.2 simulations show a systematic decrease in high-latitude surface seawater δ¹⁸O in warmer climates. This effect, in combination with δ¹⁸O anomalies associated with local hydrological conditions, helps reconcile long-standing model-data discrepancies in the Southern Hemisphere high latitudes.