Constraints on the Seawater δ18O-Sea Surface Salinity Relationship across the Strong 2015/2016 ENSO Cycle

Reconstructions of preindustrial El Niño-Southern Oscillation (ENSO) variability often rely on oxygen isotope (δ18O) signals in marine carbonates such as corals, but modern-day variations in seawater δ18O (δ18Osw) across ENSO extremes are poorly characterized. In one such coral-based reconstruction, ENSO variability in recent decades was significantly stronger than during the pre-industrial period (Grothe et al., 2019). The dual influence of sea-surface temperature (SST) and δ18Osw on coral δ18O make it difficult to attribute the increased δ18Ocoral signal to an increase in ENSO-related SST variability, δ18Osw variability, or a combination of both, in part owing to a lack of δ18Osw data across ENSO extremes in ENSO centers of action. The scant data that do exist support a strong linear relationship between δ18Osw and sea surface salinity (SSS) in the tropical Pacific, ranging from 0.05 ‰/psu at Kiritimati Island to 0.32 ‰/psu in Koror, Palau (Conroy et al., 2017).

Here, we examine 196 paired in situ SSS-δ18Osw measurements from the NOAA 2016 ENRR Cruise (February-March, 2016) and 53 samples from the P18 US GO-SHIP Cruise (November-December 2016). The data capture the peak of the strong 2015/2016 El Niño and the weakening stage of the ensuing La Niña, and encompass 140°W to 125°W, 10°S to 10°N. Our results show that SSS and surface δ18Osw of the El Niño samples are generally low compared to La Niña values, consistent with mooring data, and with La Niña data exhibiting higher variability north of the equator. A 3% and 61% decrease are observed in SSS and δ18Osw over the ENSO cycle respectively. Further assessment yield two significant relationships with slopes of 0.13 and 0.18 ‰/psu. Conformity with published data from ENSO-neutral conditions suggest consistent SSS-δ18Osw relationship across space and time during the event (Conroy et al., 2017; O'Connor et al., 2021). The expected strong ENSO signal is not resolved likely due to high intrinsic sampling variability. These paired datasets help expand the currently scarce water isotope data network and offer a close look into hydrological changes during ENSO extremes at the heart of ENSO impact. They will help disentangle factors influencing marine carbonate isotopic records and improve climate variability interpretations as tied to anthropogenic climate change.