Deconvolution of the atmospheric radiocarbon record in the last 50,000 years

Here we identify the principal causes of changes in the atmospheric Δ14C record by removing the influence of two discrete large changes in the geomagnetic field between 45 and 30 kyr. The marked transitions in the Δ14C record during this period can be interpreted as being the result of either “zero magnetic field” for periods of the order of 6500 y and 600 y at ∼ 40 kyr and 32 kyr B.P., respectively, or for longer durations if the field was non-zero for appreciable time during the event. Once the effect of these events has been removed, the residual Δ14C record shows a plateau with nearly constant value at ∼ 300‰ during 28–17 kyr B.P, followed by a sharp decrease in Δ14C between 17 kyr B.P. and the present, to a Δ14C value of ∼ 0‰. Estimated global paleomagnetic fields derived from paleointensity measurements in deep sea sediments (SINT-800, NAPIS-75 and Sumatra Basin) can only explain a maximum of 50% of this residual Δ14C record during 30–0 kyr B.P. We propose that the remainder must have resulted from changes in oceanic circulation leading to isolation of ≥ 20% of oceanic radiocarbon inventory from large scale mixing with the atmosphere during 28–17 kyr B.P. Subsequent inclusion of this carbon may be responsible in part for the decrease between 17 kyr B.P. to the present, to a Δ14C value of ∼ 0‰. We also note that the deconvoluted Δ14C record during the past 15,000 y B.P. seems to be primarily influenced by the changes in the atmospheric CO2 concentrations. These possibilities have been discussed in previous works, but our results here point to a seemingly cyclical change in carbon exchange that was initiated prior to the Last Glacial Maximum.