Investigation of the UK37’ – SST relationship for Atlantic Ocean suspended particulate alkenones: An alternative regression model with implications for UK37’ paleothermometry

The U^K₃₇’ index, which describes the proportion of unsaturated alkenones- a molecule uniquely synthesized by prymnesiophytes- is one of the main proxies used to reconstruct past sea surface temperatures (SSTs). This index has been shown to strongly correlate with algal growth temperature, making it possible to quantitatively reconstruct past SSTs from U^K₃₇’ values determined in sediment cores. Although the U^K₃₇’ index is a reliable proxy, uncertainties concerning the relationship between U^K₃₇’ values and phytoplankton growth temperatures remain.

The development of a more reasonably constrained and accurately representative calibration of the relationship between alkenone unsaturation and sea surface temperatures (U^K₃₇’- SST) for use in reconstructing past SSTs is important to improve our overall understanding of past ocean dynamics. This study introduces a new regression model to relate alkenone unsaturation to SST for suspended particulate alkenone samples from the Atlantic Ocean; the Richards curve (Richards, 1959). This non-linear regression model provides a robust calibration of the U^K₃₇’ – SST relationship, and uniquely accounts for both the fact that the U^K₃₇’ index is a proportion, and so must lie between 0 and 1, as well as for the observed reduction in slope at the warm and cold ends of the temperature range. The inverse of this relationship, for the first time, allows estimation of SST from a measured U^K₃₇’ value, including quantitative estimates of uncertainty. Further, we show the regression model is slightly improved when U^K₃₇’ is regressed against monthly mean SST rather than annual mean SST.