Long term Paleogene increase in precipitation intermittency and intensity at mid-latitudes

As the world warms due to rising greenhouse gas concentrations, the Earth system moves toward a climate state without societal precedent. Unmitigated scenarios of emissions produce climates like those of the Early Eocene by 2150 CE. Terrestrial records of rivers and floodplains from numerous Paleogene sedimentary basins in the US Western interior and Europe indicate an increase in flash floods and droughts at paleo mid latitudes, indicating increased precipitation intensity and intermittency. A global synthesis of Paleogene precipitation proxies allows us to reconstruct the timing of changes in hydroclimate from Paleocene to the Paleocene-Eocene Thermal Maximum (PETM) and through the Early Eocene Climatic Optimum (EECO). We observe that the largest shifts in hydroclimate are not linked to the PETM but rather occur during the warm late Paleocene and then at the end of the EECO. This is indicated by sedimentological proxies from paleo rivers and floodplains, paleosol geochemical proxies, and biological proxies. The sedimentological proxies indicate a shift from normal rivers, such as are characteristic at mid-latitudes today, to flood-prone rivers in late Paleocene, such as are characteristic in subtropics today. In the flood-prone Paleogene rivers sediment transport occurred during flashy floods. The rivers shifted back to normal at the end of the EECO. Paleocene and early Eocene paleosols indicate sustained droughts and intermittent seasonal rains. Biological proxies indicate large water table fluctuations and shifts in vegetation types. At PETM there is no change in the state of hydroclimate, but rather a further intensification of floods and droughts. These results show that current global warming is likely to cause intensification of precipitation intermittency and intensity at mid-latitudes with significant effects on water availability and agriculture. The most dramatic shifts in hydroclimate were not linked to the largest amplitude of atmospheric drivers at the PETM, but rather suggest a threshold-driven relationship between the atmospheric drivers and hydroclimate. This may suggest that significant changes in hydroclimate are to be expected already before 2150 CE.