Extended Data Fig. 10: Removal of prior calcite precipitation effect from measured δ18O.
From: Late Miocene Arctic warmth and terrestrial climate recorded by North Greenland speleothems

Measured δ18O in a, KC19-7 (blue), b, KC19-9 (green), c, KC19-12 (orange), and d, KC19-14 (purple), corrected for the 18O-enrichment from prior calcite precipitation73. Parameters for pCO2, temperature (T), and water film thickness (D) were varied in 7 scenarios to conduct a sensitivity test of the model: 1 - pCO2 = 1500 ppm; T = 1 °C; D = 0.015 mm; 2 - pCO2 = 3550 ppm; T = 1 °C; D = 0.01 mm; 3 - pCO2 = 4370 ppm; T = 4 °C; D = 0.01 mm; 4 - pCO2 = 5000 ppm; T = 6 °C; D = 0.01 mm; 5 - pCO2 = 5750 ppm; T = 8 °C; D = 0.01 mm; 6 - pCO2 = 7080 ppm; T = 11 °C; D = 0.01 mm; 7 - pCO2 = 15000 ppm; T = 15 °C; D = 0.005 mm. e, U-Pb ages plotted as mean and propagated 2 standard error on common depth scale. Individual statistics given in Supplementary Table 1. Composite of measured δ18O data for all KC samples (red). Normalised kernel density function (blue shading) of prior calcite precipitation-corrected δ18O, estimated from all model outputs with parameters in scenarios 1-7 and using a 0.6-mm window to account for unequal sampling resolutions. The bold black line shows the maximal probability at each step. DFT = distance from top.