As global temperatures approach levels unprecedented in human history, past warm periods offer insights into climate variations associated with increasing greenhouse gas concentrations. The Early Paleogene (65-49 million years ago) is a prime example of a warmer climate state, with average temperatures 10 to 19 K higher than today. However, our understanding of this period is limited by sparse proxy data and uncertainties in key model parameters, preventing a comprehensive analysis of its climate. In this study we use a new compilation of temperature proxies to select the least biased simulations from the Deep-Time model intercomparison project for each of six Early Paleogene time periods. These ensembles align well with proxy temperature and precipitation data across all six periods of the Early Paleogene. This method also provides an independent estimation of CO2 concentrations, suggesting that Early Paleogene CO2 levels were around 600-1200 ppm, except for the warmest periods, the Paleocene Eocene Thermal Maximum (PETM) and Early Eocene Climatic Optimum where CO2 concentrations were 1100-1800 and 900-1500 ppm, respectively. The ensembles suggest temperatures in high-latitude regions increased by up to 13 K during the PETM hyperthermal, while tropical temperatures rose by approximately 1 K, indicating substantial polar amplification during the nearly ice-free Early Paleogene. Precipitation increased almost everywhere with warming, at a global mean rate of 2% per Kelvin of global warming, matching theorized and modeled estimates from the modern climate. Additionally, precipitation changes suggest a poleward shift in global atmospheric circulation with warming, aligning with theories of modern climate change.

The Earth is transitioning to a state unprecedented in human history. This transition poses a challenge for predicting the future, as climate models require testing and calibration with real-world data from high greenhouse gas climates. Despite significant progress in climate modeling, changes in the precipitation remain highly uncertain. The Paleocene-Eocene Thermal Maximum (PETM) was the warmest period of the Cenozoic Era, and thus serves as a test-bed of how precipitation is altered by extreme greenhouse gas warming. Here, we use paleosol bulk geochemistry methods to quantify changes in precipitation during the PETM in the Uinta Basin, Utah. We find no change in mean annual precipitation during this warming event. However, paleosol mass balance results track increased translocation of carbonates, increased clay illuviation, and increased accumulation of redox-sensitive elements. These results, along with shifts in fluvial stratigraphy provide evidence for increased intensity and intermittency of extreme precipitation events that may be related to changes in the transport direction, seasonality, and moisture transport capability of the North American monsoon. Surprisingly, changes in fluvial stratigraphy continued for 105-106 years after the PETM while paleosol geochemistry returned to pre-PETM conditions almost immediately at the boundary, suggesting persistent changes in precipitation intensity despite a decrease in global temperature. These findings provide further support for an intensification of the hydrological cycle during and after the PETM, provide evidence for a decoupling between mean and extreme precipitation, and indicate the importance of multi-proxy, regional studies for understanding the complexities of climate change.