Past to Future: Defining the states and variability of the ocean

Past2Future (P2F) aims to develop, expand, and leverage the wealth of paleoclimate data to significantly improve existing Earth System Models and deepen our understanding of Earth’s climate response to various types of forcing, with a focus on abrupt climate transitions and tipping points. To achieve this, our work focuses on the compilation, integration, and re-evaluation of past sea surface temperature (SST) data with the aim of defining the states and variability of the ocean temperatures across four pivotal climate intervals: the Mid-Holocene (6.5-5.5 ka), Last Glacial Maximum (23-19 ka), Eemian (130-116 ka), and the mid-Pliocene Warm Period (3.3-3 Ma).

To date, we have identified all published global SST records spanning the Last Glacial Maximum and the Mid-Holocene, reconstructed using both geochemical techniques and faunal assemblages. For the Last Glacial Maximum, we identified 1,426 geochemical and faunal proxy records from over 1,100 cores. For the Mid-Holocene we identified 1,014 geochemical and faunal proxy records from 790 cores. Subsequently, we assessed the suitability of these records for climate model evaluation and tuning by considering: i) the robustness of each record’s age model; ii) the SST reconstruction methodology and associated uncertainties; and iii) site location and representativeness. Consequently, we have prioritised marine sediment records that feature robust age models, high-resolution SST records, low calibration uncertainties, derived from sites minimally influenced by additional climatic or environmental factors (e.g. upwelling), and, where possible, supported by alternative multi-proxy SST reconstructions. To address remaining spatial and temporal data gaps, we will generate new SST records using alkenone (UK’₃₇) and glycerol dialkyl glycerol tetraether (TEX₈₆) proxies, generating datasets that support climate models. The resulting curated and expanded SST datasets will provide a robust benchmark for climate model evaluation and tuning, ultimately contributing to more robust and accurate simulations of past climate states and more reliable projections of future climate change.