Neukom, Steiger, Gómez-Navarro, Wang and Werner (Nature 571, 550–554; 2019) reconstructed Common Era (past 2,000 years) temperature from the global PAGES 2k proxy network using a multi-method ensemble of reconstruction techniques, and found no evidence for globally coherent preindustrial warm and cold epochs: the coldest epoch of the last millennium occurred at different times in different regions, and the spatial coherence that does exist is statistically consistent with stochastic (unforced) climate variability, implying preindustrial forcing was insufficient to produce globally synchronous multidecadal-to-centennial temperature extremes. The conclusion rests, however, on a proxy network that is sparse and spatially unequal — especially in the first millennium CE and the Southern Hemisphere — and on reconstruction methods whose low-frequency skill is hard to validate, leaving two linked open questions: (1) how much of the inferred absence of global epoch coherence would survive expanded proxy networks and validated reconstruction methods, i.e., is it a property of the climate system or an artifact of data and method limitations; and (2) what mixture of unforced internal climate variability and regionally varying responses to external forcing (volcanic aerosols, solar variability, regional ocean–atmosphere feedbacks) produced the observed regional multidecadal extremes. Proxy databases and two-millennium reanalyses have substantially expanded since 2019, but no published work has re-run the global epoch-coherence test with the new data, and attribution of the regional extremes remains inconsistent across studies. The problem is open.

Background

Earth's climate over the Common Era (CE; the past ~2,000 years) has conventionally been described through named climatic epochs — most prominently the Medieval Climate Anomaly (MCA, a putatively warm interval roughly 950–1250 CE) and the Little Ice Age (LIA, a putatively cold interval roughly 1450–1850 CE) — which have often been assumed to have occurred simultaneously across the globe. Because widespread instrumental measurements only begin in the mid-19th century, CE climate is known through palaeoclimate proxy records (natural archives such as tree rings, corals, ice cores, lake and marine sediments, and speleothems, whose measurable properties record past temperature or hydroclimate) and historical documents. The PAGES 2k (Past Global Changes, 2,000-year) Consortium assembled the standard global proxy database for this period. Converting these sparse, noisy records into estimates of past climate fields is the task of climate field reconstruction (CFR). Prominent approaches include regression-based methods and offline data assimilation (DA), in which proxy observations are combined with ensembles of climate-model states through proxy system models (PSMs) — forward models translating climate into the proxy-observed quantity. Because the 'true' past climate is unknown, CFR methods are benchmarked in pseudoproxy experiments, which emulate the real proxy network with known climate from model simulations in order to test reconstruction skill.

Neukom et al. (2019) applied a multi-method ensemble of global CFRs (including analogue, regression, covariance-regularization and DA techniques) to the PAGES 2k network specifically to test the assumed global synchrony of CE epochs. They found no evidence for preindustrial globally coherent cold and warm periods: the coldest epoch of the last millennium most likely occurred in the fifteenth century in the central and eastern Pacific Ocean, in the seventeenth century in northwestern Europe and southeastern North America, and in the mid-nineteenth century over most other regions; the warmest period of the past two millennia occurred during the twentieth century for more than 98 per cent of the globe; and the spatial coherence that does exist over the preindustrial CE is consistent with the spatial coherence of stochastic climate variability. A companion paper published simultaneously, the PAGES 2k Consortium (2019) analysis, showed the complementary face of the result: global-mean multidecadal temperature fluctuations over the CE are coherent across seven different reconstruction methods and consistent with forced model simulations, with a substantial contribution from volcanic aerosol forcing during 1300–1800 CE. Together the two papers sharpen the puzzle: the global mean varies coherently at multidecadal timescales, while the regional timing of epoch extremes is asynchronous.

Two limitations of this evidence define the open problem. First, the proxy network underlying the 2019 analysis is spatially and temporally unequal — particularly sparse during the first millennium CE and across the Southern Hemisphere — and the low-frequency (multidecadal to centennial) skill of reconstruction methods is difficult to assess because instrumental calibration and validation windows are short; a follow-up DA study (Franke et al. 2020) showed that reconstruction skill depends strongly on proxy quantity, quality, screening and PSM assumptions. It is therefore unknown how much of the inferred regional asynchrony reflects the climate system rather than data and method limitations. Second, the mechanism behind the observed regional timing is unresolved: candidate explanations are unforced internal variability (stochastic ocean–atmosphere dynamics producing asynchronous regional extremes) versus regionally varying responses to external forcing (explosive volcanism, solar variability, and regional ocean–atmosphere feedbacks), which requires explicit detection-and-attribution analysis against last-millennium model ensembles such as the PMIP (Paleoclimate Modelling Intercomparison Project) past1000 simulations.

Problem Statement

Determine whether the absence of globally coherent warm and cold periods over the preindustrial Common Era is a robust property of the climate system, and explain mechanistically the observed regional timing of multidecadal-to-centennial temperature extremes. Concretely, building on Neukom et al. (2019):

  1. Robustness of the coherence structure. Re-assess the spatiotemporal coherence of preindustrial Common Era temperature at multidecadal-to-centennial timescales using the expanded proxy networks now available together with reconstruction or reanalysis methods whose low-frequency skill has been explicitly validated, quantifying the effects of each of proxy screening, proxy system model assumptions, and covariance stationarity. The question is how much of the previously inferred regional asynchrony of epoch extremes — such as the staggered regional timing of the coldest epoch of the last millennium — survives these improvements, and whether globally synchronous preindustrial epochs are excluded or the earlier null result reflected data and method limitations.

  2. Attribution of the regional extremes. Attribute the regional multidecadal-to-centennial temperature extremes of the preindustrial Common Era by quantitatively separating the contributions of unforced internal climate variability from regionally varying responses to external forcing (volcanic aerosols, solar variability, and regional ocean–atmosphere feedbacks), such that the stochastic structure of the null hypotheses — including nonstationary forcings and realistic internal variability — is explicitly justified. The question is what mixture of unforced internal climate variability and regionally varying forced responses produced the regional multidecadal-to-centennial extremes, and how those extremes partition quantitatively between the two.

The two components share one scientific objective: establishing the true space-time structure of preindustrial temperature variability and its causes. They are linked open questions arising from the same source finding — the paper's null result is itself stated as consistency of the observed preindustrial coherence with stochastic climatic variability and as insufficiency of preindustrial forcing to produce globally synchronous extremes — and neither component alone resolves the question of the true space-time structure of preindustrial temperature variability and its causes. The question concerns the full preindustrial Common Era and the full globe, not a single hemisphere, region, sub-period, or global-mean series alone.

References

  1. Raphael Neukom, Nathan Steiger, Juan José Gómez-Navarro, Jianghao Wang & Johannes P. Werner (2019), 'No evidence for globally coherent warm and cold periods over the preindustrial Common Era', Nature 571, 550–554. doi:10.1038/s41586-019-1401-2, https://doi.org/10.1038/s41586-019-1401-2
  2. PAGES 2k Consortium: Raphael Neukom, Luis A. Barboza, Michael P. Erb, Feng Shi, Julien Emile-Geay, Michael N. Evans, Jörg Franke, Darrell S. Kaufman, Lucie Lücke, Kira Rehfeld, Andrew Schurer, Feng Zhu, Stefan Brönnimann, Gregory J. Hakim, Benjamin J. Henley, Fredrik Charpentier Ljungqvist, Nicholas McKay, Veronika Valler & Lucien von Gunten (2019), 'Consistent multidecadal variability in global temperature reconstructions and simulations over the Common Era', Nature Geoscience 12, 643–649. doi:10.1038/s41561-019-0400-0, https://doi.org/10.1038/s41561-019-0400-0
  3. Scott St. George (2019), 'The aberrant global synchrony of present-day warming', Nature 571, 483–484. doi:10.1038/d41586-019-02179-2, https://doi.org/10.1038/d41586-019-02179-2
  4. Jörg Franke, Veronika Valler, Stefan Brönnimann, Raphael Neukom & Fernando Jaume-Santero (2020), 'The importance of input data quality and quantity in climate field reconstructions – results from the assimilation of various tree-ring collections', Climate of the Past 16, 1061–1074. doi:10.5194/cp-16-1061-2020, https://doi.org/10.5194/cp-16-1061-2020
  5. Jianglin Wang, Bao Yang, Miao Fang, Zhiyuan Wang, Jingjing Liu & Shuyuan Kang (2023), 'Synchronization of summer peak temperatures in the Medieval Climate Anomaly and Little Ice Age across the Northern Hemisphere varies with space and time scales', Climate Dynamics 60, 3455–3470. doi:10.1007/s00382-022-06524-6, https://doi.org/10.1007/s00382-022-06524-6 (published online 6 October 2022)
  6. Heinz Wanner, Christian Pfister & Raphael Neukom (2022), 'The variable European Little Ice Age', Quaternary Science Reviews 287, 107531. doi:10.1016/j.quascirev.2022.107531, https://doi.org/10.1016/j.quascirev.2022.107531
  7. Bronwen L. Konecky, Nicholas P. McKay, Georgina M. Falster, Samantha L. Stevenson, Matt J. Fischer, Alyssa R. Atwood, Diane M. Thompson, Matthew D. Jones, Jonathan J. Tyler, Kristine L. DeLong, Belen Martrat, Elizabeth K. Thomas, Jessica L. Conroy, Sylvia G. Dee, Lukas Jonkers, Olga V. Churakova (Sidorova), Zoltán Kern, Thomas Opel, Trevor J. Porter, Hussein R. Sayani, Grzegorz Skrzypek & the Iso2k Project Members (2023), 'Globally coherent water cycle response to temperature change during the past two millennia', Nature Geoscience 16, 997–1004. doi:10.1038/s41561-023-01291-3, https://doi.org/10.1038/s41561-023-01291-3
  8. Jason E. Smerdon, Edward R. Cook & Nathan J. Steiger (2023), 'The Historical Development of Large-Scale Paleoclimate Field Reconstructions Over the Common Era', Reviews of Geophysics 61, e2022RG000782. doi:10.1029/2022RG000782, https://doi.org/10.1029/2022RG000782
  9. Feng Zhu, Julien Emile-Geay, Kevin J. Anchukaitis, Nicholas P. McKay, Samantha Stevenson & Zilu Meng (2023), 'A pseudoproxy emulation of the PAGES 2k database using a hierarchy of proxy system models', Scientific Data 10, 624. doi:10.1038/s41597-023-02489-1, https://doi.org/10.1038/s41597-023-02489-1
  10. Rachel M. Walter, Hussein R. Sayani, Thomas Felis, Kim M. Cobb, Nerilie J. Abram, Ariella K. Arzey, Alyssa R. Atwood, Logan D. Brenner, Émilie P. Dassié, Kristine L. DeLong, Bethany Ellis, Julien Emile-Geay, Matthew J. Fischer, Nathalie F. Goodkin, Jessica A. Hargreaves, K. Halimeda Kilbourne, Hedwig Krawczyk, Nicholas P. McKay, Andrea L. Moore, Sujata A. Murty, Maria Rosabelle Ong, Riovie D. Ramos, Emma V. Reed, Dhrubajyoti Samanta, Sara C. Sanchez & Jens Zinke (2023), 'The CoralHydro2k database: a global, actively curated compilation of coral δ18O and Sr/Ca proxy records of tropical ocean hydrology and temperature for the Common Era', Earth System Science Data 15, 2081–2116. doi:10.5194/essd-15-2081-2023, https://doi.org/10.5194/essd-15-2081-2023
  11. Nicholas P. McKay, Darrell S. Kaufman, Stéphanie H. Arcusa, Hannah R. Kolus, David C. Edge, Michael P. Erb, Chris L. Hancock, Cody C. Routson, Maurycy Żarczyński, Leah P. Marshall, Georgia K. Roberts & Frank Telles (2024), 'The 4.2 ka event is not remarkable in the context of Holocene climate variability', Nature Communications 15, 6555. doi:10.1038/s41467-024-50886-w, https://doi.org/10.1038/s41467-024-50886-w
  12. Veronika Valler, Jörg Franke, Yuri Brugnara, Eric Samakinwa, Ralf Hand, Elin Lundstad, Angela-Maria Burgdorf, Laura Lipfert, Andrew Ronald Friedman & Stefan Brönnimann (2024), 'ModE-RA: a global monthly paleo-reanalysis of the modern era 1421 to 2008', Scientific Data 11, 36. doi:10.1038/s41597-023-02733-8, https://doi.org/10.1038/s41597-023-02733-8
  13. Liang Ning, Zhengyu Liu, Michael E. Mann, Jian Liu, Mi Yan, Weiyi Sun, Kefan Chen, Yanmin Qin & Qin Wen (2025), 'Decadal climate variability during the pre-industrial Common Era: characteristics and mechanisms', Science Bulletin 70, 2190–2203. doi:10.1016/j.scib.2025.04.056, https://doi.org/10.1016/j.scib.2025.04.056
  14. Fen Wu, Liang Ning, Zhengyu Liu, Jian Liu, Wenqing Hu, Mi Yan, Fangmiao Xing, Lili Lei, Haohao Sun, Kefan Chen, Yanmin Qin, Benyue Li & Chuanxi Xu (2025), 'A new last two millennium reanalysis based on hybrid gain analog offline EnKF and an expanded proxy database', npj Climate and Atmospheric Science 8, 62. doi:10.1038/s41612-025-00961-w, https://doi.org/10.1038/s41612-025-00961-w
  15. Dominik Collet, Sam White, Scott Bremer, Brita Brenna, Håkon Glørstad, Ingar Mørkestøl Gundersen, Heli Huhtamaa, Kirstin Krüger, Hans W. Linderholm, Fredrik Charpentier Ljungqvist, Astrid E. J. Ogilvie, Helene Løvstrand Svarva & Bergsveinn Þórsson (2025), 'The Little Ice Age: The History and Future of a Traveling Concept', WIREs Climate Change 16, e70023. doi:10.1002/wcc.70023, https://doi.org/10.1002/wcc.70023
  16. Michael N. Evans, Lucie J. Lücke, Kevin J. Fan & Feng Zhu (2026), 'A database of databases for Common Era paleoclimate applications', Earth System Science Data 18, 1185–1202. doi:10.5194/essd-18-1185-2026, https://doi.org/10.5194/essd-18-1185-2026