Assessing multiproxy approaches (Sr/Ca, U/Ca, Li/Mg, and B/Mg) to reconstruct sea surface temperature from coral
Yang Wu1, Stewart J Fallon1, Neal E Cantin2
1Research School of Earth Sciences, The Australian National University, Mills Road, Canberra, ACT 2601, Australia.
Abstract:
Due to the increasing concerns of global warming and short instrumental records of sea surface temperature (SST), coral-based proxies, such as δ18O, Sr/Ca, U/Ca, and Li/Mg have been developed and applied to reconstruct SST in paleoclimate studies. However, these proxies are not universally applicable in different environments, because they are affected by coral physiology and various environmental factors. In this study, seven long-lived Porites corals were collected from the southern sector of the Great Barrier Reef (GBR) off the coast of Gladstone and the central sector of the GBR within the Whitsunday Islands in 2017 and 2018. Coral sites were selected to cover a wide latitudinal range with different annual temperature ranges. Century-long geochemical records (Li/Ca, B/Ca, Mg/Ca, Sr/Ca, and U/Ca) were generated using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) at weekly resolution. This study has tested the robustness of two traditional temperature proxies (Sr/Ca and U/Ca), a recently developed temperature proxy (Li/Mg), and an additional potential temperature proxy (B/Mg). U/Ca was found to be the most robust and stable temperature proxy for corals from the GBR over long-term timescales. Sr/Ca is a close second, however due to the lower response of Sr fractionation per 1 °C, it is more sensitive to analytical methods and less sensitive to annual SST changes than U/Ca. Li/Mg and B/Mg have clearer periodicity compared to Li/Ca and B/Ca. Both Li/Mg and B/Mg are strongly correlated with SST, which is due to the cancellation of temperature-independent commonality. Empirical calibrations established from this multi-proxy approach increase the certainty of temperature reconstruction when a single proxy does not perform well. These site- and colony-specific SST calibrations also provide an opportunity to revisit the universal multi-trace element calibration of sea surface temperatures (UMTECS) model, which does not require the knowledge of local SST for calibration.
