Climate change in the Arctic is occurring rapidly, and projections suggest the complete loss of summer sea ice by the middle of this century1. The sensitivity of permanently frozen ground (permafrost) in the Northern Hemisphere to warming is less clear, and its long-term trends are harder to monitor than those of sea ice. Here we use palaeoclimate data to show that Siberian permafrost is robust to warming when Arctic sea ice is present, but vulnerable when it is absent. Uranium–lead chronology of carbonate deposits (speleothems) in a Siberian cave located at the southern edge of continuous permafrost reveals periods in which the overlying ground was not permanently frozen. The speleothem record starts 1.5 million years ago (Ma), a time when greater equator-to-pole heat transport led to a warmer Northern Hemisphere2. The growth of the speleothems indicates that permafrost at the cave site was absent at that time, becoming more frequent from about 1.35 Ma, as the Northern Hemisphere cooled, and permanent after about 0.4 Ma. This history mirrors that of year-round sea ice in the Arctic Ocean, which was largely absent before about 0.4 Ma (ref. 3), but continuously present since that date. The robustness of permafrost when sea ice is present, as well as the increased permafrost vulnerability when sea ice is absent, can be explained by changes in both heat and moisture transport. Reduced sea ice may contribute to warming of Arctic air4,5,6, which can lead to warming far inland7. Open Arctic waters also increase the source of moisture and increase autumn snowfall over Siberia, insulating the ground from low winter temperatures8,9,10. These processes explain the relationship between an ice-free Arctic and permafrost thawing before 0.4 Ma. If these processes continue during modern climate change, future loss of summer Arctic sea ice will accelerate the thawing of Siberian permafrost.
Subscribe to Journal
Get full journal access for 1 year
only $3.83 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
The chronology data from this article are available at the NGDC database, at https://doi.org/10.5285/d56addfe-b7b8-4b70-9617-fadc1f7666b4. All data that support the findings of this study are available from the corresponding author on request as well.
IPCC Climate Change 2013: The Physical Science Basis (eds Stocker, T. F. et al.) (Cambridge Univ. Press, 2013).
Bell, D. B., Jung, S. J. A. & Kroon, D. The Plio-Pleistocene development of Atlantic deep-water circulation and its influence on climate trends. Quat. Sci. Rev. 123, 265–282 (2015).
Cronin, T. M. et al. Enhanced Arctic amplification began at the Mid-Brunhes Event ~400,000 years ago. Sci. Rep. 7, 14475 (2017).
Ballantyne, A. P. et al. The amplification of Arctic terrestrial surface temperatures by reduced sea-ice extent during the Pliocene. Palaeogeogr. Palaeoclimatol. Palaeoecol. 386, 59–67 (2013).
Dai, A., Luo, D., Song, M. & Liu, J. Arctic amplification is caused by sea-ice loss under increasing CO2. Nat. Commun. 10, 121 (2019).
Screen, J. A. & Simmonds, I. The central role of diminishing sea ice in recent Arctic temperature amplification. Nature 464, 1334–1337 (2010).
Lawrence, D. M., Slater, A. G., Tomas, R. A., Holland, M. M. & Deser, C. Accelerated Arctic land warming and permafrost degradation during rapid sea ice loss. Geophys. Res. Lett. 35, L11506 (2008).
Liu, J., Curry, J. A., Wang, H., Song, M. & Horton, R. M. Impact of declining Arctic sea ice on winter snowfall. Proc. Natl Acad. Sci. USA 109, 4074–4079 (2012).
Park, H., Walsh, J. E., Kim, Y., Nakai, T. & Ohata, T. The role of declining Arctic sea ice in recent decreasing terrestrial Arctic snow depths. Polar Sci. 7, 174–187 (2013).
Zhang, T. Influence of the seasonal snow cover on the ground thermal regime: an overview. Rev. Geophys. 43, RG4002 (2005).
Serreze, M. C. & Stroeve, J. Arctic sea ice trends, variability and implications for seasonal ice forecasting. Phil. Trans. R. Soc. Lond. A 373, 20140159 (2015).
Schuur, E. A. G. et al. Vulnerability of permafrost carbon to climate change: implications for the global carbon cycle. Bioscience 58, 701–714 (2008).
Anisimov, O. & Reneva, S. Permafrost and changing climate: the Russian perspective. Ambio 35, 169–175 (2006).
Vaks, A. et al. Speleothems reveal 500,000-year history of Siberian permafrost. Science 340, 183–186 (2013).
Lauriol, B., Ford, D. C., Cinq-Mars, J. & Morris, W. A. The chronology of speleothem deposition in northern Yukon and its relationships to permafrost. Can. J. Earth Sci. 34, 902–911 (1997).
Brown, J., Ferrians, J. O. J., Heginbottom, J. A. & Melnikov, E. S. Circum-Arctic Map of Permafrost and Ground-ice Conditions Circum-Pacific Map Series CP-45 (US Geological Survey & Circum-Pacific Council for Energy and Mineral Resources, 1997).
Peel, M. C., Finlayson, B. L. & McMahon, T. A. Updated world map of the Köppen–Geiger climate classification. Hydrol. Earth Syst. Sci. 11, 1633–1644 (2007).
Cheng, H. et al. Improvements in 230Th dating, 230Th and 234U half-life values, and U–Th isotopic measurements by multi-collector inductively coupled plasma mass spectrometry. Earth Planet. Sci. Lett. 371–372, 82–91 (2013).
Lisiecki, L. E. & Raymo, M. E. A Pliocene-Pleistocene stack of 57 globally distributed benthic δ18O records. Paleoceanography 20, PA1003 (2005).
Hansen, J. et al. Global temperature change. Proc. Natl Acad. Sci. USA 103, 14288–14293 (2006).
Laskar, J. et al. A long-term numerical solution for the insolation quantities of the Earth. Astron. Astrophys. 428, 261–285 (2004).
Kender, S. et al. Closure of the Bering Strait caused Mid-Pleistocene Transition cooling. Nat. Commun. 9, 5386 (2018).
Maroon, E. A., Kay, J. E. & Karnauskas, K. B. Influence of the Atlantic meridional overturning circulation on the Northern Hemisphere surface temperature response to radiative forcing. J. Clim. 31, 9207–9224 (2018).
Lawrence, K. T., Sosdian, S., White, H. E. & Rosenthal, Y. North Atlantic climate evolution through the Plio-Pleistocene climate transitions. Earth Planet. Sci. Lett. 300, 329–342 (2010).
Cronin, T. M. et al. Quaternary ostracode and foraminiferal biostratigraphy and paleoceanography in the western Arctic Ocean. Mar. Micropaleontol. 111, 118–133 (2014).
Vandenberghe, J. et al. Eurasian permafrost instability constrained by reduced sea-ice cover. Quat. Sci. Rev. 34, 16–23 (2012).
Kurita, N. Origin of Arctic water vapor during the ice-growth season. Geophys. Res. Lett. 38, (2011).
Bintanja, R. & Selten, F. M. Future increases in Arctic precipitation linked to local evaporation and sea-ice retreat. Nature 509, 479–482 (2014).
O’Neill, H. B. & Burn, C. R. Impacts of variations in snow cover on permafrost stability, including simulated snow management, Dempster Highway, Peel Plateau, Northwest Territories. Arctic Sci. 3, 150–178 (2017).
Barnhart, K. R., Miller, C. R., Overeem, I. & Kay, J. E. Mapping the future expansion of Arctic open water. Nat. Clim. Change 6, 280–285 (2016).
Tchebakova, N. M., Rehfeldt, G. E. & Parfenova, E. I. in Permafrost Ecosystems: Ecological Studies (Analysis and Synthesis) Vol. 209 (eds Osawa, A. et al.) 427–446 (Springer, 2010).
Ford, D. & Williams, P. W. Karst Geomorphology and Hydrology (Unwin Hyman, 1989).
Seligman, B. The English language edition of the geocryological map of Russia and neighbouring republics. P.J. Williams and I.M.T. Warren. 1999. Ottawa: Collaborative Map Project. Handbook: 32 p, soft cover. ISBN 0-9685013-0-3. Map: 16 sheets (95 × 66 cm per sheet), scale 1:2,500,000. $US350.00. Polar Record 36, 259–260 (2000).
Mason, A. J., Henderson, G. M. & Vaks, A. An acetic acid-based extraction protocol for the recovery of U, Th and Pb from calcium carbonates for U-(Th)-Pb geochronology. Geostand. Geoanal. Res. 37, 261–275 (2013).
Hiess, J., Condon, D. J., McLean, N. & Noble, S. R. 238U/235U systematics in terrestrial uranium-bearing minerals. Science 335, 1610–1614 (2012).
Mason, A. J. & Henderson, G. M. Correction of multi-collector-ICP-MS instrumental biases in high-precision uranium-thorium chronology. Int. J. Mass Spectrom. 295, 26–35 (2010).
Cheng, H., Edwards, R. L., Murrell, M. T. & Benjamin, T. M. Uranium-thorium-protactinium dating systematics. Geochim. Cosmochim. Acta 62, 3437–3452 (1998).
Steiger, R. H. & Jäger, E. Subcommission on geochronology: convention on the use of decay constants in geo- and cosmochronology. Earth Planet. Sci. Lett. 36, 359–362 (1977).
Ludwig, K. A. (ed.) Berkeley Geochronological Centre Special Publication Vol. 5 (Berkeley Geochronological Centre, 2012).
Rolph, G., Stein, A. & Stunder, B. Real-time Environmental Applications and Display sYstem: READY. Environ. Model. Softw. 95, 210–228 (2017).
Stein, A. F. et al. NOAA’s HYSPLIT atmospheric transport and dispersion modeling system. Bull. Am. Meteorol. Soc. 96, 2059–2077 (2015).
Bulygina, O. N., Veselov, V. M., Razuvaev, V. N. & Aleksandrova, T. M. Baseline Climatological Data Sets (RIHMI-WDC, 2014); http://meteo.ru/english/climate/cl_data.php.
Bulygina, O. N., Veselov, V. M., Razuvaev, V. N. & Aleksandrova, T. M. Description of hourly data set of meteorological parameters obtained at Russian stations. Meteo http://meteo.ru/english/climate/descrip12.htm (2014).
We thank I. G. Kazanzev, V. A. Stol’arov, D. S. Sokol’nikov and other Speleoclub Arabica members for their help with expeditions’ organizations, mapping the caves and collecting samples. We thank S. V. Alexeev, E. A. Kozireva, V. A. Pellinen and A. A. Svetlakov from Institute of Earth’s Crust of the Russian Academy of Sciences (IEC RAS) in Irkutsk for help in the field and with samples shipment. We thank V. A. Balaev, and V. V. Alexioglo from Lensk, for their help during the fieldwork in Ledyanaya Lenskaya Cave. We thank O. Kwiecien from Northumbria University, Newcastle upon Tyne, F. Lechleitner from the University of Oxford, UK, and N. Brall from University of Lyon, France, for help with samples collecting in Botovskaya Cave. We thank B. Cohen from the Geological Survey of Israel for the help with graphic work. We thank M. Deininger, R. Parrish and L. Schirrmeister for useful comments that helped to improve the quality of this paper. We also acknowledge the NOAA Air Resources Laboratory (ARL) for the provision of the HYSPLIT transport and dispersion model and/or READY website (https://www.ready.noaa.gov) used in this publication. This work was funded by NERC Fellowship NE/G013829/1, NERC Standard Grant NE/K005057/1, Royal Society grant JP080831 and RFBR joint grant 09-05-92605 КО_а.
The authors declare no competing interests.
Peer review information Nature thanks Michael Deininger, Randall Parrish and Lutz Schirrmeister for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
A, Map and cross-section of Ledyanaya Lenskaya Cave, with speleothem and temperature logger positions. B, Map of Botovskaya Cave with locations of speleothems and temperature loggers.
A, Ledyanaya Lenskaya Cave air temperatures between March 2010 and December 2013. B, Comparison of air temperature in Ledyanaya Lenskaya Cave (from A), mean annual surface temperature and surface air temperature changes (Lensk meteorological station 24923 data). C, Monitoring of air temperatures inside Botovskaya Cave (February 2010–February 2016). Between 2010 and 2014, temperatures were monitored deep inside the cave, whereas between 2015 and 2016 loggers were placed closer to the entrance. D, Comparison of Botovskaya Cave temperatures with mean surface temperature and surface temperature changes between February 2010 and July 2014 (data from temperature logger outside the cave).
A, Permafrost map of the Ledyanaya Lenskaya Cave area (reproduced with permission from ref. 33, Cambridge Univ. Press). The cave site is marked by a magenta circle. Lensk and Olekminsk are marked by black circles. B, Permafrost map of the Botovskaya Cave area33. The cave location is marked by magenta circle. Zhigalovo is marked by black circle. Types and thickness of the permafrost are shown below. In places with continuous permafrost (types 18–20), it covers >95% of the area, but may contain small unfrozen units (taliks), mainly under permanent bodies of water. Taliks may go through the entire permafrost layer (through taliks), or through part of it (not through taliks).
Extended Data Fig. 4 Pictures of speleothems’ cross-sections with ages (in ka) from Ledyanaya Lenskaya Cave.
Age uncertainties are 2σ. The U–Pb ages shown in black and a single U–Th age in stalactite SLL9-214 is shown in brown. Hiatuses are shown with red arrows. All speleothems are composed of calcite.
Extended Data Fig. 5 Pictures of speleothems’ cross-sections with ages (in ka) from Botovskaya Cave.
Age uncertainties are 2σ. The U–Pb ages shown in black and U–Th ages14 are shown in brown. Most speleothems are composed of aragonite, except stalagmite SB-6919, and layers A–D and G in stalagmite SB-01112, which are calcitic.
A, Calcitic stalagmite SLL10-6 with its layered structure. B, Magnified area (5 to 3 mm—rectangle on the top of A) in crossed polar light, showing columnar crystals. C, Aragonitic stalagmite SB-7497(3) with fibrous crystals. D, Magnified area (3 to 1.5 cm) shown by black rectangle in C in plain polar light.
Extended Data Fig. 7 Common Pb composition assessment for Ledyanaya Lenskaya (upper) and Botovskaya (lower) caves.
234U–206Pb and 238U–206Pb appear on the left, and 235U–207Pb on the right, showing consistent common Pb values for each cave. The y intercept represents the common lead ratio; yellow bars show the assigned range for common Pb in age calculations. Groups refer to the regression groups in Supplementary Table 1 with solid black lines showing regression fits. Grey contours show the percentage change in calculated age resulting from changing the common Pb composition by its assigned uncertainty. Dashed black lines show reference isochrons and ungrouped data are shown by grey circles. Uncertainties (2σ) are sometimes smaller than the symbol size. Particular details for each plot are as follows. A, Estimate of the common 208Pb/206Pb for Ledyanaya Lenskaya Cave in 234U/206Pb–208Pb/206Pb isotope space. 234U is used in the plot instead of 238U to suppress scatter in the U/206Pb ratio due to variations in the 234U/238Uinitial ratio. All but two of the ungrouped data have age uncertainties due to common Pb of <1% and are thus insensitive to the common Pb correction. Most data that are more sensitive to the common Pb composition (that is, those between the 1% and 3% contours) are included in the regressions to estimate the common Pb composition. B, Equivalent plot in 235U/207Pb–208Pb/207Pb isotope space for the estimation of the common 208Pb/207Pb ratio for Ledyanaya Lenskaya cave. Groups 3 and 4 correspond to the clump of ungrouped data close to the horizontal axis at about 0.225 on A. Note that 235U/207Pb ages are substantially more sensitive to the common Pb correction; these ages are used as a check on U–Pb concordance rather than to derive the dates used in the paper. C, Estimate of the common 208Pb/206Pb for Botovskaya Cave in 238U/206Pb–208Pb/206Pb isotope space. Groups 1, 3, 4 and 6 (plotted as squares) are unpublished data from about 2 Ma samples included here only to provide additional constraint on the common Pb composition (Supplementary Table 1). Ages used here are from the data plotted as circles. As for Ledyanaya Lenskaya Cave, the common Pb estimate includes the data for which the correction matters most as far as is possible. D, Equivalent plot in 235U/207Pb–208Pb/207Pb isotope space for the estimation of the common 208Pb/207Pb ratio for Botovskaya Cave. 235U/207Pb data are more sensitive to the common Pb correction.
A, The chronology of Ledyanaya Lenskaya Cave speleothems with 95% confidence errors. Each column in the plot represents one individual speleothem named in bottom horizontal axis. In each column, the U–Pb ages (purple circles) appear in stratigraphic order from left (young) to right (old). For each U–Pb age, the corresponding proportion of radiogenic 206Pb (right vertical axis) is shown by red circles above. The U–Th age of the youngest layer A in the stalactite SLL9-214 is shown by blue circle (bottom left). The two isochron ages of layers F top and G in SLL10-6 stalagmite are shown by olive circles. Several replicate age determinations for similar layers were performed and appear in the plot in the same order as shown in Supplementary Table 1. B, Botovskaya Cave speleothems’ U–Th ages (dark blue circles, left y axis; Vaks et al.14 and current work) and U–Pb ages (light blue circles, left y axis), the latter are given with percentages of radiogenic 206Pb (red circles, right y axis). All age uncertainties are shown by 95% confidence error bars. The stratigraphic age of the layers dated from stalagmite SB-01112 increases from left to right.
Extended Data Fig. 9 238U–234U–206Pb ages and concordance of 235U–207Pb ages for Ledyanaya Lenskaya Cave (based on Supplementary Table 1).
Ages and age uncertainties are shown in black with the corresponding coloured error bar indicating the degree of concordance (235U–207Pb age as a percentage of the 238U–234U–206Pb age; blue horizontal line indicates perfect concordance) of the 235U–207Pb age. Blue error bars indicate the analysis is concordant within error. Orange error bars denote slightly discordant analyses, where the apparent discordance is likely due to common Pb correction on 207Pb. Red error bars denote analyses that are discordant with markedly low 235U–207Pb ages attributed to overcorrection of the common Pb due to a residual interference on 208Pb. The numbers and bars in red show the percentage bias in the 238U–234 U–206Pb ages if the discordance of these samples is attributed to an interference on 208Pb, based on the relative sensitivity of the 235U–207Pb and 238U–234U–206Pb ages to the common Pb correction (Supplementary Table 1). Grey horizontal lines indicate where ages replicate in two or more speleothems. The biases in the 238U–234U–206Pb ages attributed to the interference on 208Pb do not change these ages outside of error, hence, the tendency of these ages to replicate well in other samples. The discordance arises almost entirely from the 235U–207Pb ages (which are not used), due to their vastly greater sensitivity to over-correction of the common Pb than the 238U–234U–206Pb ages. Errors are 95% confidence. For each speleothem, the stratigraphic age of the dated layers increases from left to right.
Extended Data Fig. 10 HYSPLIT-model-based seven-day back-trajectories of 12 snow events at the Ledyanaya Lenskaya Cave site.
A–L, These snowfall events were accompanied by a substantial decrease in air temperature indicating that Arctic air was involved in the synoptic event. Six snow events occurred in October and first half of November 2012 (A–F), and other six events occurred in October and first half of November of 2013 (G–L). For each of 12 events, three back trajectories for elevations of 500, 1,500 and 5,000 m above sea level were calculated, with at least one of them originating in the Arctic Ocean in each case. The dates and times of the events are as follows (dd/mm/yyyy): 04/10/2012, 21:00 (A); 07/10/2012, 21:00 (B); 19/10/2012, 09:00 (C); 21/10/2012, 15:00 (D); 05/11/2012, 09:00 (E); 11/11/2012, 09:00 (F); 01/10/2013, 09:00 (G); 09/10/2013, 00:00 (H); 19/10/2013; 09:00 (I); 25/10/2013, 09:00 (J); 02/11/2013, 09:00 (K); 10/11/2013, 09:00 (L). Identification information: Job ID: 164576; Job Start: Tue 3 Sep, 12:19:42, UTC 2019; Meteorology: 0000Z, 1 Oct 2012 – GDAS1 (A); Job ID: 165119; Job Start: Tue 3 Sep, 12:46:49, UTC 2019; Meteorology: 0000Z, 1 Oct 2012 – GDAS1 (B); Job ID: 169518; Job Start: Sun 8 Sep 14:04:06, UTC 2019; Meteorology: 0000Z, 15 Oct 2012 – GDAS1 (C); Job ID: 169478; Job Start: Sun 8 Sep 13:58:40, UTC 2019; Meteorology: 0000Z, 22 Oct 2012 – GDAS1 (D); Job ID: 171083; Job Start: Sun 8 Sep 15:42:40, UTC 2019; Meteorology: 0000Z, 1Nov 2012 – GDAS1 (E); Job ID: 171195; Job Start: Sun 8 Sep 15:47:51, UTC 2019; Meteorology: 0000Z, 8 Nov 2012 – GDAS1 (F); Job ID: 187728; Job Start: Fri 30 Aug 12:07:03 UTC 2019; Meteorology: 0000Z, 1 Oct 2013 – GDAS1 (G); Job ID: 172040; Job Start: Sun 8 Sep 16:46:37 UTC 2019; Meteorology: 0000Z, 8 Oct 2013 – GDAS1 (H); Job ID: 172169; Job Start: Sun 8 Sep 16:54:44 UTC 2019; Meteorology: 0000Z, 15 Oct 2013 – GDAS1 (I); Job ID: 172846; Job Start: Sun 8 Sep 17:16:55 UTC 2019; Meteorology: 0000Z, 22 Oct 2013 – GDAS1 (J); Job ID: 173152; Job Start: Sun 8 Sep 17:24:37 UTC 2019; Meteorology: 0000Z, 1 Nov 2013 – GDAS1 (K); Job ID: 173482; Job Start: Sun 8 Sep 17:32:06 UTC 2019; Meteorology: 0000Z, 8 Nov 2013 – GDAS1 (L). Parameters equal for all calculations in Extended Data Fig. 10: Source, latitude 60.371000, longitude 116.946472; heights, 500 m, 1,500 m, 5,000 m above ground level (AGL); trajectory direction, backward; duration, 168 h; vertical motion calculation method, model vertical velocity.
Table 1a shows U-Pb data from Ledyanaya Lenskaya and Botovskaya caves, where 238U-206U chronology data appear in columns A-X, initial 234U/238U data – in columns Y-AA, 235U-207Pb chronology data – in columns AB-AY, concordance % - in columns AZ-BB, U concentration (ppm) in column BC and the comments in column BD. Further details appear in footnotes below the table. Table 1b shows common Pb estimates for Ledyanaya Lenskaya and Botovskaya caves. Further details appear in footnotes below the table.
Table 2 shows U-Th chronology of speleothems from Botovskaya Cave with raw data in columns A-P, and data corrected for initial 230Th in columns Q-AA. Further details appear in footnotes below the table.
About this article
Cite this article
Vaks, A., Mason, A.J., Breitenbach, S.F.M. et al. Palaeoclimate evidence of vulnerable permafrost during times of low sea ice. Nature 577, 221–225 (2020). https://doi.org/10.1038/s41586-019-1880-1