LAUSR

Abstract Our understanding of past climate conditions largely comes from paleoclimate proxies, such as oxygen isotope ratios ( δ 18 Oc) in marine fossils. The marine δ 18 Oc signal primarily reflects a mixture of seawater temperature and oxygen isotopic composition of seawater ( δ 18 Ow) at the time of calcification. Knowledge of δ 18 Ow is critical for the interpretation of marine δ 18 Oc records but remains poor for past hothouse climates. Here, we conduct water isotope-enabled simulations of the early Eocene using CO2 levels of 1×, 3×, 6×, and 9× the preindustrial value. We calculate model δ 18 Oc using simulated δ 18 Ow and ocean temperature, and make direct comparison with proxy records. Model δ 18 Oc matches the proxy values well for the early Eocene and Paleocene–Eocene Thermal Maximum with root-mean-squared errors approaching the standard error in individual records. Eocene δ 18 Ow in the model exhibits strong variation depending on states of the hydrological cycle and ocean circulation. Differences in the mean δ 18 Ow between regions of net evaporation and precipitation increase monotonically with the magnitude of the net atmospheric moisture transport that connects them; however, this relationship breaks at the regional scale due to ocean circulation changes. In particular, an increase in ocean ventilation brings more 18O-enriched deep water into the mixed layer, increasing sea-surface δ 18 Ow near the ventilation site and in certain remote regions through fast upper ocean currents. δ 18 Ow variations and the linkage to both hydrological cycle and ocean circulation bring challenges for an accurate interpretation of marine δ 18 Oc records. Our study illustrates the value of using water isotope-enabled simulations and model-data comparison for learning past climate changes.