Comparing classical and Bayesian 210Pb dating models in human-impacted aquatic environments

Chronologies are an essential tool to place natural archives of environmental changes on a calendar scale. Because of this, studies that compare and assess the accuracy and precision of available dating models are essential. Pb-210 is a radioactive isotope which is used to date recent sediments (<150 yr). Here we contrast the chronologies resulting from two different Pb-210 dating models: the Constant Flux model (also known as the Constant Rate of Supply model) and the recently developed Bayesian Plum model. This comparison was implemented by using four sediment cores from contrasting environmental settings, and showed several benefits of using a Bayesian approach. This allows to infer variables, such as the supported levels of Pb-210, crucial to the chronology and commonly estimated through either samples where an asymptotic behaviour is observed or through Ra-226 measurements, which themselves contain some level of uncertainty. Another step of traditional methods is the selection of the equilibrium depth in order to calculate total inventories, which carries strong consequences for the resulting age depth model. Plum, on the other hand, infers the Pb-210 flux, which eliminates the need to select an equilibrium depth and allows for dating cores with incomplete inventory. When traditional methods address cores with missing inventory, Cs-137 time markers are commonly used to correct the chronology, but Plum can use these time markers to improve the chronology and to reduce the uncertainty. These aspects show that the Bayesian approach results in more robust and objective chronologies, and thus in better reconstructions of environmental change over the past centuries.