Talk
Optimising laser ablation-based methods for long sea surface temperature proxy reconstructions from coral skeletons
Thomas Arney, Christopher Standish, Hunter Hughes, Jens Zinke, and Gavin Foster
Geochemistry Group Research in Progress meeting, Bristol, UK ()
Abstract
As reef-forming stony corals grow, they incorporate trace elements into their carbonate skeletons. The proportions of these trace elements to calcium or to each other (e.g. Sr/Ca, Li/Mg) vary in large part due to the ambient water temperature. Well-dated cores can therefore provide long records of sea surface temperature (SST), spanning tens to hundreds of years. After calibration with recent observation-based SST data, these proxies enable the extension of temperature records beyond the range of reliable instrumental data, potentially providing vital quantification of the preindustrial period and beyond. However, such long time-series require high-throughput yet precise measurement of the trace elemental composition of the coral skeleton. Here we show that laser ablation inductively-coupled plasma mass spectrometry (LA-ICPMS) is a viable alternative to established micro-milling and solution-based sampling and measurement techniques, but with higher resolution and much higher throughput. Accuracy and precision of LA-ICPMS data is improved by reporting the average of a small number of parallel, laterally-spaced laser transects, each scaled and/or offset to align uneven growth banding. To counter instrumental drift and improve accuracy and reproducibility within and between long measurement sessions, we implement a data reduction scheme using time-varying regressions of measured and reference element-to-calcium ratios in multiple matrix-matched reference materials. Using these LA-ICPMS methods, a 100-year SST reconstruction at sub-monthly resolution from over a metre of core could be generated in around 24 hours of instrument time and with minimal sample preparation. Given that replication using nearby coral cores is key for reliable reconstructions of SST (Sayani et al., 2019), applying this workflow to long coral cores from suitable locations would enable significantly improved estimations of global temperatures in the pre- and early instrument era.