Tipping elements and amplified polar warming during the Last Interglacial

Abstract

Irreversible shifts of large-scale components of the Earth system (so-called 'tipping elements') on policy-relevant timescales are a major source of uncertainty for projecting the impacts of future climate change. The high latitudes are particularly vulnerable to positive feedbacks that amplify change through atmosphere-ocean-ice interactions. Unfortunately, the short instrumental record does not capture the full range of past or projected climate scenarios (a situation particularly acute in the high latitudes). Natural archives from past periods warmer than present day, however, can be used to explore drivers and responses to forcing, and provide data against which to test models, thereby offering insights into the future. The Last Interglacial (129-116,000 years before present)-the warmest interglacial of the last 800,000 yearswas the most recent period during which global temperatures were comparable with low-end 21st Century projections (up to 2°C warmer, with temperature increase amplified over polar latitudes), providing a potentially useful analogue for future change. Substantial environmental changes happened during this time. Here we synthesise the nature and timing of potential high-latitude tipping elements during the Last Interglacial, including sea ice, extent of the boreal forest, permafrost, ocean circulation, and ice sheets/sea level, and review the thresholds and feedbacks that likely operated through this period. Notably, substantial ice mass loss from Greenland, the West Antarctic, and possibly sectors of the East Antarctic drove a 6-9 m rise in global sea level. This was accompanied by reduced summer sea-ice extent, poleward-extended boreal forest, and reduced areas of permafrost. Despite current chronological uncertainties, we find that tipping elements in the high latitudes all experienced rapid and abrupt change (within 1-2 millennia of each other) across both hemispheres, while recovery to prior conditions took place over multi-millennia. Our synthesis demonstrates important feedback loops between tipping elements, amplifying polar and global change during the Last Interglacial. The high sensitivity and tight interconnections between polar tipping elements suggests that they may exhibit similar thresholds of vulnerability in the future, particularly if the aspirations of the Paris Agreement are not met.

A global mean sea surface temperature dataset for the Last Interglacial (129–116 ka) and contribution of thermal expansion to sea level change

“…In this regard, the Last Interglacial (LIG), an interval spanning approximately 129 000 to 116 000 years ago, is of great value (Dutton et al, 2015). Described as a "super-interglacial" (Turney and Jones, 2010;Overpeck et al, 2005), the LIG was one of the warmest periods of the last 800 kyr, experiencing relatively high polar temperatures compared to the global mean ("polar amplification") (Past Interglacials Working Group of PAGES, 2016;Hoffman et al, 2017;Turney and Jones, 2010;Capron et al, 2017), with the most geographically widespread expression of high global mean sea level in the recent geological record (GMSL, +6.6 to +11.4 m) (Dutton et al, 2015;Grant et al, 2014;Kopp et al, 2009;Rohling et al, 2017), abrupt shifts in regional hydroclimate (Wang et al, 2008;Thomas et al, 2015), and elevated atmospheric CO 2 concentrations (relative to the preindustrial period) of ∼ 290 ppm (Köhler et al, 2017;Schneider et al, 2013;Barnola et al, 1987;Petit et al, 1999), suggesting non-linear responses in the Earth system to forcing (Steffen et al, 2018;Thomas, 2016;Dakos et al, 2008;Thomas et al, 2020). Importantly, there remains considerable debate over the contribution of sources to the highstand in global sea level (Dutton et al, 2015;Rohling et al, 2019).…”

“…Quantified temperature reconstruction data for the LIG are often drawn from disparate publications and repositories (usually reported alongside other Late Pleistocene data). To obtain reliable temperature reconstructions, it has until recently proved necessary to determine a global estimate of the magnitude of warming using only a selected number of "high-quality" records; the resulting temperature reconstructions of LIG temperatures ranged from 0.1 to > 2 • C warmer than present (CLIMAP, 1984;White, 1993;Hansen, 2005;Rohling et al, 2008;Turney and Jones, 2010). With the ever-increasing number of quantified temperature reconstructions of the LIG reported in individual publications, it is crucial that these datasets are brought together to derive a comprehensive reconstruction of global change during the LIG.…”

“…A further consideration is that, in contrast to terrestrial sequences, marine records typically provide a continuous record of LIG conditions (Turney andJones, 2010, 2011) and, thus, an opportunity to determine the sensitivity of GMSL to sea surface temperature (SST) conditions during the interglacial (including early maximum temperatures). Given a possible warming of 2 • C (Turney and Jones, 2010;Fischer et al, 2018), the LIG potentially provides insights into the drivers of sea level rise and the long-term impacts under a global temperature target set out in the 2016 Paris Climate Agreement (Schellnhuber et al, 2016).…”

“…Here we present version 1.0 of the Last Interglacial SST database (Turney et al, 2019). This database builds on the previously published 2010 data compilation of Turney and Jones (2010), and includes substantially more records. Importantly, the micro-organisms used to determine SSTs move along with the currents and encounter a range of temperatures during their life cycle (van Sebille et al, 2015;Doblin and van Sebille, 2016;von Gyldenfeldt et al, 2000).…”

“…Müller et al, 1998;Prahl et al, 2003;Sikes et al, 2005;Segev et al, 2016;Elderfield and Ganssen, 2000), and palaeoclimate estimates and model outputs Bakker and Renssen, 2014;NEEM Community Members, 2013;Lunt et al, 2013), including the recently recognized historic (Anthropocene) change in modern plankton communities, which has major implications for calibration studies (Jonkers et al, 2019). The influence of ocean currents has not been explored (or corrected for) in previous studies of the LIG (Hoffman et al, 2017;Capron et al, 2014;Turney and Jones, 2010) and is important for obtaining correct absolute SSTs. This descriptor describes the contents of the database, the criteria for inclusion, and quantifies the relation of each record with instrumental temperature, including the estimated impact of ocean current drift on individual sites and global averages.…”