Magnetic assessment of OSL and radiocarbon ages of sediments beneath a lava in Jeju Island, Korea: Implication of possible resetting of OSL signals and age constraint of the late Quaternary lava

Abstract

Better chronological constraining of Late Pleistocene-Holocene volcanic eruptions is essential for figuring out spatio-temporal frequency distribution of ancient volcanism and assessing volcanic eruptions probability in a volcanic field. For this chronological frame work, radiocarbon ( C-14) and luminescence dating methods are being attempted alternatively, instead of K-Ar and Ar-40/ Ar-39 techniques with lower resolutions and age limits of applicability. For the sediments capped by a lava flow in a Quaternary basaltic volcanic field of Jeju Island, Korea, we here introduce an interesting and probably frequently-encountered case of the discrepancy in age estimates between C-14 and optically stimulated luminescence (OSL) chronologies. For samples with different positions in the sediments, the C-14 ages of organic material extracts have a stratigraphically consistent but widely dispersed values between 2.25 +/- 0.09 and 27.2 +/- 0.5 calibrated (cal.) kyr BP, whereas the OSL ages of quartz extracts have a narrow range between 13.4 +/- 1.3 and 16.1 +/- 0.7 ka. Multiple magnetic parameters including Koenigsberger ratio, ratio of natural remanent magnetization (NRM) to saturation isothermal remanent magnetization intensities, and similarity in NRM direction compared with paleomagnetic direction for the upper lava comprehensively infer the paleo-heating of the sediments by the lava. Further, paleo-heating temperatures estimation by two independent magnetic methods, based on progressive thermal demagnetizations of NRM and repeated magnetic susceptibility versus temperature measurements with increasing peak temperatures, undoubtedly reveals that the sample positions where the OSL ages were yielded had paleo-temperatures between 200 and 250 degrees C and much higher than 550 degrees C, which are sufficiently high to reset their luminescence signals. Therefore, we interpret that a weighted mean OSL age of 14.8 +/- 0.7 ka is the most probable age for the lava. Accordingly, the C-14 ages older than similar to 15.5 cal kyr BP are interpreted to be the deposition ages for their horizons, whereas the three exceptionally young C-14 ages (similar to 2.3-similar to 7.7 cal kyr BP) for the upper sediments were contaminated by modern carbon. We suggest that magnetic applications, when combined with C-14 and luminescence dating techniques, can help greatly improve the reliance of chronologies for late Quaternary volcanism of the world.