Fig. 3: Scatter plot of Δ¹⁴C data from Dispilio against reference from Brehm et al.²² and IntCal20⁷⁰, and best last ring-fit for the dated wood samples (χ²).

a Measured ¹⁴C concentrations represented as Δ¹⁴C, vertical bars represent 1σ errors (Supplementary Data 1); legend labels marked with “DISP-“ and larger symbols refer to wood samples from Dispilio and corresponding ¹⁴C measurements obtained in this study; other labels represent data from Brehm et al.²², measurements from Siberian larch, Bristlecone pine, and Irish oak measured at two different labs, symbol shapes according to Lab; Bristlecone pine ¹⁴C data are shifted forward by 1 year from the original BR22 publication following a correction to the dating of the bristlecone master tree-ring chronology (Supplementary Note 2); shaded band represents IntCal20⁷⁰. Inset in right corner same as in main panel a. but at higher resolution, spanning 14 years centred around the ¹⁴C spike. Lower two panels: χ2 tests of Dispilio measurements against the average from BR22²², b. χ² test results for wood samples DISP-10,070 and DISP-10063 (χ² crit. value = 9.49), and c for wood samples DISP-10206 and DISP-10611 (χ² crit. value = 15.51), legend as in panel above. Figure panels produced in R¹⁰¹, code and source data available in Supplementary Data 3.