A probabilistic assessment of the rapidity of PETM onset

Knowledge of the onset duration of the Paleocene-Eocene Thermal Maximum - the largest known greenhouse-gas-driven global warming event of the Cenozoic - is central to drawing inferences for future climate change. Single-foraminifera measurements of the associated carbon isotope excursion from Maud Rise (South Atlantic Ocean) are controversial, as they seem to indicate geologically instantaneous carbon release and anomalously long ocean mixing. Here, we fundamentally reinterpret this record and extract the likely PETM onset duration. First, we employ an Earth system model to illustrate how the response of ocean circulation to warming does not support the interpretation of instantaneous carbon release. Instead, we use a novel sediment-mixing model to show how changes in the relative population sizes of calcareous plankton, combined with sediment mixing, can explain the observations. Furthermore, for any plausible PETM onset duration and sampling methodology, we place a probability on not sampling an intermediate, syn-excursion isotopic value. Assuming mixed-layer carbonate production continued at Maud Rise, we deduce the PETM onset was likely