摘要：A valuable analogue for assessing Earth's sensitivity to warming is the Last Interglacial (LIG; 129–116 ka), when global temperatures (0 to 2 ∘C) and mean sea level ( 6 to 11 m) were higher than today. The direct contribution of warmer conditions to global sea level (thermosteric) is uncertain. We report here a global network of LIG sea surface temperatures (SST) obtained from various published temperature proxies (e.g. faunal and floral plankton assemblages, Mg ∕ Ca ratios of calcareous organisms, and alkenone U37K′). We summarize the current limitations of SST reconstructions for the LIG and the spatial temperature features of a naturally warmer world. Because of local δ18O seawater changes, uncertainty in the age models of marine cores, and differences in sampling resolution and/or sedimentation rates, the reconstructions are restricted to mean conditions. To avoid bias towards individual LIG SSTs based on only a single (and potentially erroneous) measurement or a single interpolated data point, here we report average values across the entire LIG. Each site reconstruction is given as an anomaly relative to 1981–2010, corrected for ocean drift, and where available seasonal estimates are provided (189 annual, 99 December–February, and 92 June–August records). To investigate the sensitivity of the reconstruction to high temperatures, we also report maximum values during the first 5 millennia of the LIG (129–124 ka). We find mean global annual SST anomalies of 0.2 ± 0.1 ∘C averaged across the LIG and an early maximum peak of 0.9 ± 0.1 ∘C, respectively. The global dataset provides a remarkably coherent pattern of higher SST increases at polar latitudes than in the tropics (demonstrating the polar amplification of surface temperatures during the LIG), with comparable estimates between different proxies. Polewards of 45∘ latitude, we observe annual SST anomalies averaged across the full LIG of > 0.8 ± 0.3 ∘C in both hemispheres with an early maximum peak of > 2.1 ± 0.3 ∘C. Using the reconstructed SSTs suggests a mean LIG global thermosteric sea level rise of 0.08 ± 0.1 m and a peak contribution of 0.39 ± 0.1 m, respectively (assuming warming penetrated to 2000 m depth). The data provide an important natural baseline for a warmer world, constraining the contributions of Greenland and Antarctic ice sheets to global sea level during a geographically widespread expression of high sea level, and can be used to test the next inter-comparison of models for projecting future climate change. The dataset described in this paper, including summary temperature and thermosteric sea level reconstructions, is available at https://doi.org/10.1594/PANGAEA.904381 (Turney et al., 2019).
