Re-evaluating <SUP>14</SUP>C dating accuracy in deep-sea sediment archives

International audienceThe current geochronological state of the art for applying the radiocarbon (14C) method to deep-sea sediment archives lacks key information on sediment bioturbation. Here, we apply a sediment accumulation model that simulates the sedimentation and bioturbation of millions of foraminifera, whereby realistic 14C activities (i.e. from a 14C calibration curve) are assigned to each single foraminifera based on its simulation time step. We find that the normal distribution of 14C age typically used to represent discrete-depth sediment intervals (based on the reported laboratory 14C age and measurement error) is unlikely to be a faithful reflection of the actual 14C age distribution for a specific depth interval. We also find that this deviation from the actual 14C age distribution is greatly amplified during the calibration process. Specifically, we find a systematic underestimation of total geochronological error in many cases (by up to thousands of years), as well as the generation of age-depth artefacts in downcore calibrated median age. Even in the case of "perfect" simulated sediment archive scenarios, whereby sediment accumulation rate (SAR), bioturbation depth, reservoir age and species abundance are all kept constant, the 14C measurement and calibration processes generate temporally dynamic median age-depth artefacts on the order of hundreds of years - whereby even high SAR scenarios (40 and 60 cm kyr-1) are susceptible. Such age-depth artefacts can be especially pronounced during periods corresponding to dynamic changes in the Earth's Δ14C history, when single foraminifera of varying 14C activity can be incorporated into single discrete-depth sediment intervals. For certain lower-SAR scenarios, we find that downcore discrete-depth true median age can systematically fall outside the calibrated age range predicted by the 14C measurement and calibration processes, thus leading to systematically inaccurate age estimations. In short, our findings suggest the possibility of 14C-derived age-depth artefacts in the literature. Furthermore, since such age-depth artefacts are likely to coincide with large-scale changes in global Δ14C, which themselves can coincide with large-scale changes in global climate (such as the last deglaciation), 14C-derived age-depth artefacts may have been previously incorrectly attributed to changes in SAR coinciding with global climate. Our study highlights the need for the development of improved deep-sea sediment 14C calibration techniques that include an a priori representation of bioturbation for multi-specimen samples