Miocene temperature gradients

Climate models have difficulty simulating the weak meridional temperature gradients reconstructed from proxy records of past “greenhouse” climates. These models also normally fail to produce the right globally averaged temperatures when driven by CO2 concentrations reconstructed from proxies. Either models are missing key processes or proxy interpretations incorporate deep, persistent biases.
In this project, with an interdisciplinary international (Swiss+US) team, we will reconcile climate models and proxy interpretations for an extreme and well-characterized greenhouse climate in the Miocene period (23 to 5 million years ago). Using new proxies, we will test if meridional temperature gradients were consistently flat in the Miocene. Using an updated proxy approach, we will test if atmospheric CO2 levels are consistent with the inferred temperatures across the Miocene. Finally, we will evaluate which climate processes in current models contribute to greatest agreement with robust proxy data.
We will improve proxy interpretation by developing a proxy system model (PSM) for coccolithophore-based CO2 and temperature (alkenone and clumped isotope) proxies. Then, we will evaluate this proxy system model using preindustrial climate simulations and corresponding global proxy core top archives and subsequently apply it to Miocene simulations. For improved CO2 estimation, we will provide new orbitally resolved records of phytoplankton isotope fractionation replicated at two sites, for four Miocene timeslices of critical climate transitions spanning contrasting climate states. To test the robustness of proxy estimates of latitudinal temperature gradients, for these key timeslices, we will produce new estimates of surface ocean temperature in high latitudes and tropical regions from the clumped isotope ratio of coccoliths produced by phytoplankton in the surface ocean. This new temperature indicator is validated by recent core top and culture calibrations, and is one of few proxies produced exclusively in the ocean's photic zone with capacity to record warmer than modern tropical temperatures as well as cool high latitude temperatures. We will conduct a series of new Miocene simulations using the Community Earth System Model Version 2 (CESM2), the latest generation of the coupled climate/Earth system models developed primarily by National Center for Atmospheric Research (NCAR), which has contributed to the Coupled Model Intercomparison Project (CMIP6). Our cases evaluate the simulated climate resulting from range of atmospheric pCO2, Antarctic ice sheet states and evaluate two eccentricity configurations. Finally, we will compare simulated climates and the proxy values generated by the proxy system model for these Miocene climates, with the new CO2 proxy data and new and compiled surface ocean temperature proxy record. This comparison will allow us to identify the model forcing and processes consistent with proxy records. If our work indicates that proxy interpretations are correct, then climate model prediction for the future may underestimate polar amplification and climate sensitivity. On the other hand, if climate models and data can be reconciled in the past, model predictions in the future will have enhanced credibility and proxy system models will improve paleoclimate reconstructions.