Challenges in 14C dating towards the limit of the method inferred from anchoring a floating tree ring radiocarbon chronology to ice core records around the Laschamp geomagnetic field minimum

- An independent high-accuracy test of the GICC05 ice core time scale.
- An independent test for the presently available 14C calibration curve.
- Indications for a systematically biased calibration curve at 40 000 kyr BP.
- A solid basis for connecting 14C dated records to ice core time scales.

The 14C dating method is the cornerstone for inferring age estimates for natural archives covering the last 50 000 yrs. However, 14C age calibration for the last ice age relies mostly on records that only indirectly reflect the atmospheric 14C concentrations. In consequence, calendar age estimates are significantly more uncertain for the period of the last ice age compared to the past 14 000 yrs where tree-ring based calibration records exist. Here we connect a 14C tree-ring chronology from Kauri trees in New Zealand to ice core 10Be records via the common signal in the galactic cosmic ray flux around the period of the Laschamp geomagnetic field minimum (ca. 41 000 yrs BP). Synchronous changes of modelled 14C and 14C inferred from U/Th-dated speleothems support the ice core chronology independently and suggest that the published ice core time scale errors are rather conservative for this period. Our analysis puts 14C age determinations directly into the context of ice core climate records and it shows that the 14C records underlying the 14C calibration curve overestimate the atmospheric 14C concentration by more than 200‰. Consequently, 14C age calibration presently yields too old calendar age estimates by about 1200 yrs for this period.

100