[M-IS03] Evolution and variability of the Asian Monsoon and Indo-Pacific climate during the Cenozoic Era

[MIS03-06] Do nearby sediment cores (<10 km) in the northern Okinawa Trough result in the same paleotemperature reconstruction?

★Invited Papers

*Ru-Yun Tung¹、Pierrick Fenies¹、Masanobu Yamamoto²、Jens Hefter³、Yoshimi Kubota⁴、Chuan-Chou Shen^5,6、Sze Ling Ho¹ (1.Institute of Oceanography, National Taiwan University, Taiwan、2.Graduate School of Earth and Environmental Science, Hokkaido University, Japan、3.Alfred Wegener Institute, Helmholtz Center for Polar and Marine Research, Germany、4.Department of Geology and Paleontology, National Museum of Nature and Science, Japan、5.High-Precision Mass Spectrometry and Environment Change Laboratory (HISPEC), Department of Geosciences, National Taiwan University, Taiwan、6.Research Center for Future Earth, National Taiwan University, Taiwan)

キーワード：Okinawa Trough, Mg/Ca, UK’37, TEX86

Geochemical proxies are often used to reconstruct past ocean temperatures. Due to limitations in resources, these reconstructions are typically based on a single sediment core. Therefore, it remains unknown to what extent paleotemperature records are representative of a locality and the study region. Here, we aim to assess the spatial representativeness of commonly applied paleotemperature using an opportunistic setup involving four sediment cores located within 10 km of each other in the northern Okinawa Trough, located in the same influence of oceanographic setting. We both compiled and generated new temperature estimates based on techniques of Mg/Ca, U^K’₃₇, and TEX₈₆. The mean absolute deviations for nearby records of the same proxy showed that U^K’₃₇ has the best reproducibility, followed by Mg/Ca and lastly TEX₈₆. However, inter-proxy offsets and the Holocene trend of each proxy are not reproducible among nearby sites, probably due to noise in the proxy records stemming from instrumental error and sediment heterogeneity. All proxies display the same broad direction of glacial-interglacial changes, albeit with differences in their absolute values, especially between Mg/Ca and U^K’₃₇ records, which were previously interpreted as temperature signal from different seasons. Nonetheless, the Mg/Ca-U^K’₃₇ offsets are within the uncertainty range and not statistically significant. Both Mg/Ca and U^K’₃₇ records indicated a ~3°C glacial cooling, while a stronger glacial cooling of approximately ~6-8°C was observed in TEX₈₆ SST records. TEX₈₆ records in this region have been interpreted as either surface or subsurface (50–100 m) temperatures that coincides with the depth of the thermocline. However, when calibrated to subsurface temperature, the TEX₈₆ subT record does not resemble the Mg/Ca records of thermocline-dwelling foraminifer species, Neogloboquadrina dutertrei and Globoconella inflata, suggesting that these proxies do not reflect the same water depth. Instead, there is a far better agreement with that of benthic foraminiferal Mg/Ca records of Uvigerina spp. (~700 m) and the intermediate temperature derived from radiolarian assemblage (~500 m), hinting at a deeper-than-thermocline recording depth for TEX₈₆. In summary, our results show that proxy noise may affect the trend determination during a relatively stable interval like the Holocene and inter-proxy comparison of paleotemperature records but not the direction of glacial-interglacial shifts. We recommend horizontal temperature gradients, commonly used to reconstruct past changes in Kuroshio Current or East Asian Monsoon in the East China Sea, be based on the same proxy to minimize uncertainty in paleotemperature records. Future research should focus on constraining the recording depth of paleotemperature proxies and reducing the calibration uncertainty to achieve a more accurate and precise quantitative paleotemperature reconstruction.