Greenhouse gases

Warming from freezing soils

Freezing and thawing of soils leads to large pulses of nitrous oxide release. An empirical model shows that cropland winter nitrous oxide emissions are substantial, calling for a revision of the global nitrous oxide budget.

Nitrous oxide (N2O) is a trace gas that is responsible for about 7% of anthropogenic global warming, and also contributes to the destruction of stratospheric ozone. Its atmospheric concentration has increased from 270 ppbv in the pre-industrial era to 328 ppbv in 20151. This increase is largely a consequence of the massive use of inorganic and organic fertilizers in agriculture, which on the one hand has boosted food production and sustained rapid global population growth, but on the other hand has also stimulated microbial N2O production and increased emission from soils2. N2O fluxes from agricultural fields are typically dominated by emission pulses — so-called hot moments — which are often observed after fertilization or thaw events. Writing in Nature Geoscience, Wagner-Riddle and colleagues3 suggest that N2O pulses from freeze–thaw events in croplands contribute a substantial proportion of agricultural N2O emissions.

Emissions due to fertilizer use in agriculture are currently estimated to be 5.3 Tg per year or 64 % of the global anthropogenic N2O budget2. Agricultural N2O emissions are expected to rise, with nitrogen (N) fertilizer use projected to rise by 0.7% per year in the coming decades4. This increase is driven by a growing world population and enhanced consumption of animal products, which require substantial N inputs5. It has become evident that only reducing greenhouse gas emissions from the industrial, transport and energy sectors will not achieve the 1.5 °C target by 2100. The 21st Conference of the Parties (COP21) to the United Nations Framework Convention on Climate Change, and the resulting Paris Agreement, opened the door to directly include greenhouse gas emissions from the agricultural sector in national mitigation strategies. However, to develop tailored N2O mitigation strategies, a thorough understanding of the magnitudes and drivers of N2O emission is a prerequisite.

Current estimates of agricultural N2O sources are uncertain by approximately 50%2. Uncertainty arises from a failure to adequately capture the spatial variability of N2O emissions, and a failure to capture hot-moments, especially during winter periods when measurements are often discontinued, but when fluxes can be large.

The freezing and thawing of soils can promote N2O emissions in a number of ways. Freeze–thaw dynamics can disrupt soil aggregates and split open microbial cells, releasing nitrogen and carbon compounds. Microbes partly oxidize these compounds, thereby forming nitrates that can be used by denitrifying bacteria to conduct denitrification, which is the source of most of the N2O emitted during freeze–thaw events. Soil freezing and thawing can also help create and maintain the anaerobic conditions needed for dentrification to occur: thawing soil layers are water saturated, which hampers O2 diffusion into the soil. Upon thaw, the warmer temperatures allow microbial communities to become more active and grow. Denitrifying bacteria and other microbes can take advantage not just of the compounds made available by soil freezing, but also the substrates from crop residues that went unused while low temperatures suppressed microbial activity. In many ecosystems, soils can freeze and thaw multiple times each winter, contributing to multiple pulses of N2O. These winter freeze–thaw emissions can be so large that they dominate annual N2O fluxes from some agricultural soils6.

Wagner-Riddle et al.3 showed that cumulative N2O emissions over the non-growing period were closely related to the duration and magnitude of sub-freezing temperatures in the soil, based on multi-year, continuous micrometeorological flux measurements at two cropland sites in Canada. This relationship was further validated at 11 additional sites in Canada, the US, Japan, China and Germany. In combination with global land use and climate reanalysis data, Wagner-Riddle et al. used a derived relationship between cumulative winter emissions and the duration and magnitude of sub-freezing soil temperatures to estimate winter-period N2O emissions from seasonally frozen croplands globally. The analysis indicates that freeze–thaw events are responsible for emissions of about 1 Tg N2O–N — emissions that have not been considered in previous estimates of greenhouse gas emissions from agriculture. In other words, the global agricultural N2O budget might have been underestimated by 17 to 28%.

The relative lack of validation studies in the relatively low precipitation continental climates of Eurasia may lead to an overestimation if lower soil moisture contributes to lower soil N2O emissions in these climates. On the other hand, Wagner-Riddle et al. only estimated emissions from croplands. Substantial emissions during winter or after soil freezing and subsequent thawing have been observed for managed and natural grasslands, forests, and drained organic soils and natural wetlands (Fig. 1). Owing to this phenomenon's broad occurrence, a revision of the attribution of sources within the entire global N2O budget is needed.

Figure 1: Locations of field studies that report soil N2O emissions during winter or freeze–thaw events.

Wagner-Riddle et al.3 used an empirical model based on continuous measurements of N2O from two agricultural sites in Canada to estimate that winter emissions represent a substantial proportion of the annual N2O flux from agricultural soils globally. The locations shown comprise the field sites in Wagner-Riddle et al. and other field sites from published literature (Supplementary Information). Orange, light green, dark green, brown and blue filled circles represent studies for cropland, grassland, forest, peat (including drained and agriculturally used peat) and wetland/marsh, respectively.

To improve estimates of N2O emission from managed and natural terrestrial ecosystem sources, it is necessary to undertake continuous N2O emission measurements in long-term observation networks, such as the Integrated Carbon Observation System or National Ecological Observatory Network. N2O flux measurements at long-term research sites around the world can provide the high temporal resolution of emission measurements needed to capture hot-moments and the standardized measurements (for example, soil carbon, soil texture and soil inorganic N) that are needed to understand what drives emissions and their temporal variation.

Our understanding of the drivers and processes underlying N2O emissions related to freeze–thaw events is still poor. Key uncertainties in simulating freeze–thaw N2O emission are the development of the snow cover, the effect of subsurface ice formation on soil gas diffusion and water flow during thaw, as well as the dynamics of soil microbial processes and communities involved in the production and consumption of N2O during winter7. An improved understanding of the controls over freeze–thaw N2O emissions and more realistic process based models could help provide insight into potential mitigation options for reducing greenhouse gas emissions, and provide important assistance for policymakers as they refine their national contributions as part of the Paris Agreement.

The high spatio-temporal heterogeneity in N2O emissions from agricultural soils has led many researchers to try to constrain emissions associated with specific management activities such as fertilization, irrigation and tillage. Wagner-Riddle and colleagues3 suggest that at least in croplands that freeze, it may be even more important to understand microbial processes in soils that are freezing and thawing.


  1. 1

    WMO Greenhouse Gas Bulletin: The State of Greenhouse Gases in the Atmosphere Based on Global Observations through 2015 (World Meteorological Organization, 2016).

  2. 2

    Syakila, A. & Kroeze, C. Greenh. Gas Meas. Manag. 1, 17–26 (2011).

    Article  Google Scholar 

  3. 3

    Wagner-Riddle, C. et al. Nat. Geosci. 10, 279–283 (2017).

    Article  Google Scholar 

  4. 4

    Alexandratos, N. & Bruinsma, J. World Agriculture Towards 2030/2050: The 2012 Revision ESA Working Paper No. 12–03 (Food and Agriculture Organization, 2012).

  5. 5

    Herrero, M. et al. Nat. Clim. Change 6, 452–461 (2016).

    Article  Google Scholar 

  6. 6

    Abalos, D. et al. Glob. Chang. Biol. 22, 1244–1255 (2016).

    Article  Google Scholar 

  7. 7

    Wang, C. et al. Glob. Chang. Biol. 22, 2963–2978 (2016).

    Article  Google Scholar 

Download references

Author information



Corresponding authors

Correspondence to Klaus Butterbach-Bahl or Benjamin Wolf.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Butterbach-Bahl, K., Wolf, B. Warming from freezing soils. Nature Geosci 10, 248–249 (2017).

Download citation

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing