Multi-proxy reconstruction of sea surface temperatures in the Pacific Southern Ocean over the last glacial-interglacial cycle

The Southern Ocean, especially the Subantarctic region, plays a critical role in anthropogenic carbon uptake, heat exchange, and nutrient transfer between high and low latitudes. Pronounced past changes in surface ocean properties, particularly sea surface temperatures (SSTs), reflect this region’s exceptional sensitivity to external and internal forcings over millennial and orbital timescales. This underscores its importance in understanding past and future climate change and its key role as major link between Antarctica and the low latitudes.

Accurate reconstructions of past SSTs are crucial for understanding past climate dynamics and validating models for future projections. To achieve this, various temperature proxies based on physical, chemical, and biological properties preserved in marine sediments have been developed. However, all proxies carry uncertainties due to environmental factors that may bias the signals archived in the sedimentary record. A multiproxy approach helps to mitigate these uncertainties, providing a more robust and comprehensive interpretation of past climate conditions.

This study presents a high-resolution SST reconstruction from IODP Expedition 383 Site U1539 and pre-site survey PS75/054 in the Subantarctic South Pacific, near the modern Subantarctic Front (SAF), using three different temperature proxies: a diatom transfer function and two organic proxies based on coccolithophorid alkenone lipids (U^K’₃₇) and archaeal glycerol dialkyl glycerol tetraether (GDGT) lipids (TEX₈₆). This location is characterized by unusually high sedimentation-rates (~10-50 cm/kyr), mainly because the northerly extended opal belt reaches Site U1539 during glacials with high diatom ooze deposition.

Our record spans the last 150 ka with a centennial to millennial resolution. The general temperature pattern follows the overall glacial/interglacial succession. The alkenone SSTs range from minimum values approaching zero around the Last Glacial Maximum to ~10°C in MIS 5e (and 6-7°C during the Holocene equivalent to modern austral summer SST). These exceptionally high glacial/interglacial amplitudes are much less pronounced in the diatom summer SST reconstruction with an amplitude of only~3-4°C. Much higher SSTs are shown by the TEX₈₆ with equally maximum values during MIS 5e (absolute SST higher than the alkenone SST). However, the TEX₈₆ temperature record shows very high amplitudes at millennial time-scales. Within dating uncertainties, these changes follow the Antarctic temperature pattern as recorded in ice-core.

We are discussing the palaeoceanographic implications of our SST records and potential reasons for the partial mismatches of the different SST proxies. These include varying seasonality sensitivity, depth habitat, and SST calibration and transfer function uncertainties.