Permafrost is warming at a global scale

Abstract

Permafrost warming has the potential to amplify global climate change, because when frozen sediments thaw it unlocks soil organic carbon. Yet to date, no globally consistent assessment of permafrost temperature change has been compiled. Here we use a global data set of permafrost temperature time series from the Global Terrestrial Network for Permafrost to evaluate temperature change across permafrost regions for the period since the International Polar Year (2007-2009). During the reference decade between 2007 and 2016, ground temperature near the depth of zero annual amplitude in the continuous permafrost zone increased by 0.39 ± 0.15 °C. Over the same period, discontinuous permafrost warmed by 0.20 ± 0.10 °C. Permafrost in mountains warmed by 0.19 ± 0.05 °C and in Antarctica by 0.37 ± 0.10 °C. Globally, permafrost temperature increased by 0.29 ± 0.12 °C. The observed trend follows the Arctic amplification of air temperature increase in the Northern Hemisphere. In the discontinuous zone, however, ground warming occurred due to increased snow thickness while air temperature remained statistically unchanged.

Fig. 1

Long permafrost temperature records for selected sites. a Location of boreholes with long time-series data. Because some regions lack long temperature records, shorter temperature records from Greenland and Chinese mountains are included for comparison. Depth of measurements is according to the Global Terrestrial Network for Permafrost ID: 24.4 m (ID 356), 20 m (ID 55, 79, 102, 117, 501, 710, 831, 1113, and 1710), 18 m (ID 386), 16.75 m (ID 871), 15 m (ID 854), 12 m (ID 287), 10 m (ID 265, 431), and 5 m (ID 528). The light blue area represents the continuous permafrost zone (>90% coverage) and the light purple area represents the discontinuous permafrost zones (<90% coverage). b Mean annual ground temperature over time. Colors indicate the location of the boreholes in a. Permafrost zones are derived from the International Permafrost Association (IPA) map. World Borders data are derived from http://thematicmapping.org/downloads/world_borders.php and licensed under CC BY-SA 3.0 (https://creativecommons.org/licenses/by-sa/3.0/)

Fig. 2

Permafrost temperature and rate of change near the depth of zero annual amplitude. a, b Mean annual ground temperatures for 2014–2016 in the Northern Hemisphere and Antarctica, n = 129 boreholes. c, d Decadal change rate of permafrost temperature from 2007 to 2016, n = 123 boreholes (Eq. 3). Changes within the average measurement accuracy of ~±0.1 °C are coded in green. Continuous permafrost zone (>90% coverage); discontinuous permafrost zones (<90% coverage). Permafrost zones are derived from the International Permafrost Association (IPA) map. World Borders data are derived from http://thematicmapping.org/downloads/world_borders.php and licensed under CC BY-SA 3.0 (https://creativecommons.org/licenses/by-sa/3.0/)

Fig. 3

Annual permafrost temperature change. a–d Permafrost temperature departure $\mathrm{\Delta }{\overline{T}}_{y,b}$ calculated from mean annual ground temperatures in boreholes near the depth of zero annual amplitude Z^* relative to the 2008–2009 reference period. Mean values calculated as de-clustered, indexed area-weighted averages (Eq. 1). Temperature uncertainties are expressed at 95% confidence. Sample size shown is the number of borehole sites per year and region

Fig. 4

Annual air temperature and snow depth changes. a–d Air temperature anomaly $\mathrm{\Delta }{\widehat{T}}_{y,b}$ relative to the 1981–2010 reference period calculated from mean annual air temperatures at 2-m height above the ground level interpolated from the ERA Interim reanalysis data set. Mean values calculated as de-clustered, indexed area-weighted averages (Eq. 4). Dark colored dashed lines indicate 4-year end-point running means. Snow depth changes $\mathrm{\Delta }{Ŝ}_{y,b}$ in a and b, indicated in gray, calculated as the difference relative to the 1999–2010 reference period from the CMC reanalysis data set (eq. 5). e, f Onset of snow SO, snow insulation maximum SIM (dashed line), and the end of snow melt SE. Uncertainties are expressed as shading at 95% confidence. Sample size n indicates the number of boreholes

Fig. 5

Decadal temperature change rates at permafrost borehole sites. a Boxplots showing the regional (unweighted) distribution of permafrost temperature change rates near the depth of zero annual amplitude Z^* calculated for 2007–2016 in °C per decade (Eq. 3). * indicate significant difference to 0, p < 0.05 defined by the Wilcoxon Signed-Rank test. p values ACP: 0.000, ADP: 0.002, MP: 0.016, and ANP: 0.156 (rounded to three digits). The Kruskal–Wallis test indicated p values > 0.05 for couples that are tied with brackets in the graph. b Air temperature change rates at 2 m height above ground at borehole sites in °C per decade, calculated from the ERA Interim reanalysis data for 2004–2016, separated by regions. Symbols: n number of boreholes, ACP Arctic continuous permafrost, ADP Arctic discontinuous permafrost, MP mountain permafrost, ANP Antarctic permafrost. Boxes represent 25–75% quartiles and whiskers are 1.5 interquartile ranges from the median. Medians are shown as black lines

Fig. 6

Depth distribution of borehole temperatures. a Boxplots showing the depth distribution of temperature measurements (sensor depths) and of the zero annual amplitude Z^* in the boreholes. b Temperature distribution in borehole sensors that are shallower (≤12 m) and deeper (>12 m) than the median of Z^*. * indicate significant difference to 0, p < 0.05 defined by the Wilcoxon Signed-Rank test; p values ≤12 m: 0.000, >12 m: 0.000 (rounded to three digits). The Kruskal–Wallis test between these zones (b) resulted in p = 0.908, indicating that the zones are not significantly different to each other. Boxplots represent 25–75% quartiles and whiskers are 1.5 interquartile ranges from the median. Medians are shown as black lines and labeled with values. The number of boreholes (sensors), and the number of available Z^*values is indicated by n

Fig. 7

Thermal regime of permafrost. Schematic showing the maximum (red line) and minimum ground temperature (blue line) during the year, and their convergence to give the mean annual ground temperature $\overline{T}$ at the depth of zero annual amplitude Z^*. Black dots show the schematic mean temperature for permafrost soils. Compiled guided by French

Fig. 8

Weighting and grouping of boreholes. Map showing the indices and zoning of boreholes prior to area-weighting and calculation of mean temperature changes. a Northern Hemisphere. b Antarctica. Permafrost zones are derived from the International Permafrost Association (IPA) map. World Borders data are derived from http://thematicmapping.org/downloads/world_borders.php and licensed under CC BY-SA 3.0 (https://creativecommons.org/licenses/by-sa/3.0/)