Abstract
Peatlands are strategic areas for climate change mitigation because of their matchless carbon stocks. Drained peatlands release this carbon to the atmosphere as carbon dioxide (CO2). Peatland rewetting effectively stops these CO2 emissions, but also re-establishes the emission of methane (CH4). Essentially, management must choose between CO2 emissions from drained, or CH4 emissions from rewetted, peatland. This choice must consider radiative effects and atmospheric lifetimes of both gases, with CO2 being a weak but persistent, and CH4 a strong but short-lived, greenhouse gas. The resulting climatic effects are, thus, strongly time-dependent. We used a radiative forcing model to compare forcing dynamics of global scenarios for future peatland management using areal data from the Global Peatland Database. Our results show that CH4 radiative forcing does not undermine the climate change mitigation potential of peatland rewetting. Instead, postponing rewetting increases the long-term warming effect through continued CO2 emissions.
Subject terms: Climate-change ecology, Environmental impact
Drained peatlands are sources of CO2, and though rewetting could curb emissions, this strategy results in elevated methane release. Here, the authors model peatland emissions scenarios and show that rewetting is a critical way to mitigate climate change despite potential methane increases.
Introduction
Each year, drained peatlands worldwide emit ~2 Gt carbon dioxide (CO2) by microbial peat oxidation or peat fires, causing ~5% of all anthropogenic greenhouse gas (GHG) emissions on only 0.3% of the global land surface1. A recent study states that the effect of emissions from drained peatlands in the period 2020–2100 may comprise 12–41% of the remaining GHG emission budget for keeping global warming below +1.5 to +2 °C2. Peatland rewetting has been identified as a cost-effective measure to curb emissions3, but re-establishes the emission of methane (CH4). In light of the strong and not yet completely understood impact of CH4 on global warming4,5 it may seem imprudent to knowingly create or restore an additional source. Furthermore, there is considerable uncertainty on emissions from rewetted peatlands and some studies have reported elevated emissions of CH4 compared with pristine peatlands6–9.
The trade-off between CH4 emissions with and CO2 emissions without rewetting is, however, not straightforward: CH4 has a much larger radiative efficiency than CO210. Yet, the huge differences in atmospheric lifetime lead to strongly time-dependent climatic effects. Radiative forcing of long-term GHGs (in case of peatlands: CO2 and N2O) is determined by cumulative emissions, because they factually accumulate in the atmosphere. In contrast, radiative forcing of near-term climate forcers (in case of peatlands: CH4) depends on the contemporary emission rate multiplied with the atmospheric lifetime10,11, because resulting atmospheric concentrations quickly reach a steady state of (sustained) emission and decay. Meanwhile, common metrics like global warming potential (GWP) and its sustained flux variants11,12 fail to account for temporal forcing dynamics. These different atmospheric dynamics are relevant for the question how the various management scenarios will influence global climate and whether a scenario will amplify or attenuate peak global warming, i.e., the maximum deviation in global surface temperatures relative to pre-industrial times. An amplification of peak warming increases the risk of reaching major tipping points in the Earth’s climate system13,14.
Here, we explore how the different lifetimes of CO2/N2O vs. CH4 play out when assessing options for peatland rewetting as a climate warming mitigation practice by comparing five global scenarios (Table 1). These scenarios represent extreme management options and exemplify the differences caused by timing and extent of rewetting. For our modeling exercise, we focus on the direct human-induced climatic effects and conservatively assume pristine peatlands to be climate-neutral. Further, we assume that the maximum peatland area to be drained during the 21st century equals the area that is already drained in 2018 (505,680 km², Global Peatland Database15) plus an additional ~5000 km² per year (average net increase of drained peatland area between 1990 and 201716). For all scenarios, we apply IPCC default emissions factors17 as sustained fluxes. To compare the radiative forcing effects of the different GHGs, we use a simplified atmospheric perturbation model that has been shown to provide reliable estimates of the climatic effects of peatlands18 (see Methods). Our results show that total radiative forcing quickly reaches a plateau after rewetting, because of the halted emissions of CO2/N2O of rewetted peatlands and the short atmospheric lifetime of any emitted CH4. In contrast, postponing rewetting has a long-term warming effect resulting from continued CO2 emissions. Warnings against CH4 emissions from rewetted peatlands are therefore unjustified in the context of effective climate change mitigation.
Table 1.
Scenario | Description |
---|---|
Drain_More | The area of drained peatland continues to increase from 2020 to 2100 at the same rate as between 1990 and 2017 |
No_Change | The area of drained peatland remains at the 2018 level |
Rewet_All_Now | All drained peatlands are rewetted in the period 2020–2040 |
Rewet_Half_Now | Half of all drained peatlands are rewetted in the period 2020–2040 |
Rewet_All_Later | All drained peatlands are rewetted in the period 2050–2070 |
Results and Discussion
Radiative forcing dynamics of global scenarios
Rewetting of drained peatlands instantly leads to climatic benefits compared with keeping the status quo (Fig. 1). In case of rewetting all drained peatlands (scenarios Rewet_All_Now and Rewet_All_Later, see Table 1) the radiative forcing stops increasing followed by a slow decrease. Since the response of global temperature is lagging behind changes in total radiative forcing by 15–20 years19, peatlands should be rewetted as soon as possible to have most beneficial (cooling) effects during peak warming, which AR5 climate models expect to occur after ~2060 with increasing probability towards the end of the century (Fig. 1).
The overall climatic effect of peatland rewetting is indeed strongly determined by the radiative forcing of sustained CH4 emissions (Fig. 2). However, because of the negligible or even negative emissions of CO2/N2O of rewetted peatlands and the short atmospheric lifetime of CH4, the total anthropogenic radiative forcing of all three GHGs combined quickly reaches a plateau after rewetting. Meanwhile, differences in radiative forcing between drainage (increased forcing) and rewetting scenarios (stable forcing) are mainly determined by differences in the forcing of CO2 (Fig. 2). Rewetting only half of the currently drained peatlands (Rewetting_Half_Now) is not sufficient to stabilize radiative forcing. Instead, CO2 from not-rewetted peatland keeps accumulating in the atmosphere and warming the climate. Note that in the Rewet_Half_Now scenario CH4 forcing is more than half that of the Rewet_All_… scenarios, because drained peatlands also emit CH4, most notably from drainage ditches. Comparing the scenarios Rewet_All_Now and Rewet_All_Later shows that timing of peatland rewetting is not only important in relation to peak temperature, but also with respect to the total accumulated CO2 and N2O emissions in the atmosphere and the resulting radiative forcing (Fig. 2). These patterns are valid also when considering possible future changes of new drainage rate or emission factors (Fig. 3).
General conclusions for global peatland management
Our simulations highlight three general conclusions: First, the baseline or reference against which peatland rewetting has to be assessed is the drained state with its large CO2 emissions. For this reason, rewetted peatlands that are found to emit more CH4 than pristine ones9 are no argument against rewetting. Moreover, whereas rewetted peatlands may again become CO2 sinks, the faster and larger climatic benefits of peatland rewetting result from the avoidance of CO2 emissions from drained peatlands. Second, the climate effect is strongly dependent on the concrete point in time that rewetting is implemented. This fact is hitherto insufficiently recognized because it remains hidden by the common use of metrics that involve predetermined time horizons (like GWP or sustained flux variants of GWP). Finally, in order to reach climate-neutrality in 2050 (as implied by the Paris Agreement on limiting the increase in global temperature to well below 2 °C), it is insufficient to focus rewetting efforts on selected peatlands only: to reach the Paris goal, CO2 emissions from (almost) all drained peatlands have to be stopped by rewetting2.
Limiting global warming requires immediate reduction of global GHG emissions. It has been suggested that the negative climate effects of drained peatlands could be offset by growing highly-productive bioenergy crops20 or wood biomass21 as substitute for fossil fuels. In this study, we did not include this option because similar biomass-based substitution benefits can also be reached by cultivating biomass on rewetted peatlands22, i.e., without CO2 emissions from drained peat soil.
In conclusion, without rewetting the world’s drained peatlands will continue to emit CO2, with direct negative effects on the magnitude and timing of global warming. These effects include a higher risk of reaching tipping points in the global climate system and possible cascading effects13. In contrast, we show that peatland rewetting can be one important measure to reduce climate change and attenuate peak global warming: The sooner drained peatlands are rewetted, the better it is for the climate. Although the CH4 cost of rewetting may temporarily be substantial, the CO2 cost of inaction will be much higher.
Methods
Scenarios
Drained peatland area was taken from the Global Peatland Database (GPD)15, which includes, among other things, national data from the most recent UNFCCC National Inventory Submissions and Nationally Determined Contributions. We used data separated by IPCC climate zone (boreal, temperate, and tropical) and assigned land use categories. Available land use categories were Forest, Cropland, Deep-drained grassland, Shallow-drained grassland, Agriculture (i.e., either grassland or cropland when the original data source did not differentiate between these two categories), and Peat extraction (see Table 2). Because of their only small area and uncertain emission factors, arctic drained peatlands (~100 kha) were neglected. The average net increase of drained peatland area between 1990 and 201716 assumed for the Drain_More scenario includes the disappearance of drained peatlands that have lost all their peat deposits. Newly drained/rewetted area in the scenarios is distributed across the climatic zones (and land use classes) according to the relative proportions of today’s drained peatland area. As future drainage—similar to the past two decades16—will probably focus on tropical and subtropical peatlands, our Drain_More scenario likely underestimates the climate effects of future drainage. For information on how variations in the assumed drainage rate and uncertainty of emission factors affected the displayed radiative forcing effects of the scenarios please see Fig. 3.
Table 2.
Climatic zone | Land use category | Area (kha) | CO2 (t ha−1 a−1) | CH4 (kg ha−1 a−1) | N2O (kg ha−1 a−1) |
---|---|---|---|---|---|
Boreal | Forest | 5474 | 2.5 | 9.8 | 2.6 |
Cropland | 262 | 27.9 | 58.3 | 19.4 | |
Deep-drained grassland | 426 | 20.2 | 59.6 | 14.2 | |
Shallow-drained grassland | 0 | — | — | — | |
Agriculture | 3420 | 24.1 | 43.0 | 16.8 | |
Peat extraction | 333 | 10.2 | 32.9 | 0.5 | |
Rewetted | — | −1.3 | 123.6 | 0 | |
Temperate | Forest | 6315 | 10.3 | 7.9 | 4.3 |
Cropland | 2528 | 28.6 | 58.3 | 19.4 | |
Deep-drained grassland | 3405 | 22.3 | 73.5 | 12.3 | |
Shallow-drained grassland | 2422 | 13.6 | 63.4 | 2.4 | |
Agriculture | 8389 | 21.0 | 55.8 | 10.1 | |
Peat extraction | 662 | 10.8 | 32.9 | 0.5 | |
Rewetted | — | −0.4 | 205.9 | 0 | |
Tropical | Forest | 7235 | 22.0 | 50.0 | 3.7 |
Cropland | 305 | 45.0 | 118.9 | 4.2 | |
Deep-drained grassland | 70 | 37.4 | 52.0 | 7.7 | |
Shallow-drained grassland | 0 | — | — | — | |
Agriculture | 9314 | 42.5 | 96.6 | 5.4 | |
Peat extraction | 8 | 10.1 | 32.9 | 5.6 | |
Rewetted | — | 1.9 | 166.5 | 0 |
Emission factors assumed for rewetted peatlands are also shown for each climatic zone.
Emissions
Emission factors for each climate zone and land use category were taken from the IPCC Wetland supplement17 that currently presents the most robust and complete meta-study of published emission data. We applied all emission factors as sustained fluxes. Emission factors were averaged for IPCC categories that were given at a higher level of detail (e.g., nutrient-poor vs. nutrient-rich boreal forest) than the available land use categories from the GPD. Equally, we averaged the supplied emission factors for grassland and cropland in order to obtain emission factors of the land use class Agriculture (see Table 2 for final aggregated emission factors and Supplementary Table 1 for exact aggregation steps). We included emissions from ditches and DOC exports by using emission factors and default cover fraction of ditches given by the IPCC17 (Supplementary Table 1). Since the IPCC Wetlands Supplement does not provide an emission factor for CH4 from tropical peat extraction sites, we assumed the same CH4 emissions as for temperate/boreal peat extraction. Values of the emission factors could change slightly when more emission data becomes available. To cover this possibility, we randomly varied all emission factors within a range of 10–20% uncertainty in our sensitivity analysis (Fig. 3). Please note that these uncertainty ranges do not correspond to the confidence intervals given by the IPCC Wetlands Supplement, which describe the observed variability of emissions from individual peatlands. Since our analyses take a global perspective, our sensitivity analyses instead cover possible changes of the mean emissions (i.e., emission factors). In addition, individual studies have discussed the presence of a CH4 peak for the first years after rewetting7,8. Although this is likely not a global phenomenon23, please see Supplementary Fig. 1 for an estimate of the uncertainty related to possible CH4 peaks.
Radiative forcing
The forcing model uses simple impulse-response functions24 to estimate radiative forcing effects of atmospheric perturbations of CO2, CH4, and N2O fluxes12. Perturbations of CH4 and N2O were modeled as simple exponential decays, while CO2 equilibrates with a total of five different pools at differing speeds. For CO2, we adopted the flux fractions and perturbation lifetimes used by ref. 18. In the model, we assume a perfectly mixed atmosphere without any feedback mechanisms but include indirect effects of CH4 on other reagents10.
Climatic effects of CO2 from CH4 oxidation should not be considered for CH4 from biogenic sources10. However, although the large majority of CH4 from peatlands stems from recent plant material (a biogenic source), the proportion of fossil CH4 (from old peat) may be substantial in some cases25. Thus, we conservatively included the climatic effect of CO2 from CH4 oxidation in our analyses. Overall, this forcing comprised only 5–7% of the CH4 radiative forcing and only ~1–3% of total radiative forcing.
We compare the radiative forcing trajectories of the various peatland management scenarios with the global temperature change as projected by all available pathways of IPCC’s AR5 and use the same starting year 2005 as these pathways. Further, we estimated the approximate effects of radiative forcing on global mean temperature as ~1 K per 1.23 W/m² radiative forcing26.
Supplementary information
Acknowledgements
The European Social Fund (ESF) and the Ministry of Education, Science and Culture of Mecklenburg-Western Pomerania funded this work within the scope of the project WETSCAPES (ESF/14-BM-A55-0030/16 and ESF/14-BM-A55-0031/16). G.J. received funding within the framework of the Research Training Group Baltic TRANSCOAST from the DFG (Deutsche Forschungsgemeinschaft) under grant number GRK 2000 (www.baltic-transcoast.uni-rostock.de). This is Baltic TRANSCOAST publication no. GRK2000/0032. V.H. gratefully acknowledges funding by the Federal Agency of Nature Conservation (BfN, grant number: 3516892003) and by the European Regional Development Fund (ERDF) distributed through the NBank.
Author contributions
A.G., J.C., G.J., F.K., and V.H. conceived the study. A.G., A.B., J.C., and H.J. assembled input data. A.G. implemented the simulation model with contributions from J.C. All authors discussed the results and implications. A.G. led writing of the manuscript with comments/edits from all authors.
Data availability
The models for projected temperature change were downloaded from IAMC AR5 Scenario Database (available at https://secure.iiasa.ac.at/web-apps/ene/AR5DB). Emission factors and peatland cover data are entirely included in the paper.
Code availability
The code for the atmospheric perturbation model can be found in the supplementary information. Code was written by A. Günther in the R programming language.
Competing interests
The authors declare no competing interests.
Footnotes
Peer review information Nature Communications thanks Dennis Baldocchi and other, anonymous reviewers for their contributions to the peer review of this work. Peer review reports are available.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Change history
4/12/2024
A Correction to this paper has been published: 10.1038/s41467-024-47604-x
Supplementary information
Supplementary information is available for this paper at 10.1038/s41467-020-15499-z.
References
- 1.Joosten, H., Sirin, A., Couwenberg, J., Laine, A. & Smith, P. in Peatland Restoration And Ecosystem Services (eds Bonn, A., Allott, T., Evans, M., Joosten, H. & Stoneman, R.) (Cambridge University Press, Cambridge, UK, 2016).
- 2.Leifeld J, Wüst-Galley C, Page S. Intact and managed peatland soils as a source and sink of GHGs from 1850 to 2100. Nat. Clim. Change. 2019;9:945–947. doi: 10.1038/s41558-019-0615-5. [DOI] [Google Scholar]
- 3.Leifeld J, Menichetti L. The underappreciated potential of peatlands in global climate change mitigation strategies. Nat. Commun. 2018;9:1–8. doi: 10.1038/s41467-018-03406-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Nisbet EG, et al. Very strong atmospheric methane growth in the four years 2014–2017: Implications for the Paris Agreement. Glob. Biogeochem. Cycles. 2019;33:318–342. doi: 10.1029/2018GB006009. [DOI] [Google Scholar]
- 5.Mikaloff Fletcher SE, Schaefer H. Rising methane: a new climate challenge. Science. 2019;364:932–933. doi: 10.1126/science.aax1828. [DOI] [PubMed] [Google Scholar]
- 6.Abdalla M, et al. Emissions of methane from northern peatlands: a review of management impacts and implications for future management options. Ecol. Evol. 2016;6:7080–7102. doi: 10.1002/ece3.2469. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Franz D, Koebsch F, Larmanou E, Augustin J, Sachs T. High net CO2 and CH4 release at a eutrophic shallow lake on a formerly drained fen. Biogeosciences. 2016;13:3051–3070. doi: 10.5194/bg-13-3051-2016. [DOI] [Google Scholar]
- 8.Hahn J, Köhler S, Glatzel S, Jurasinski G. Methane exchange in a coastal fen in the first year after flooding—a systems shift. PLoS ONE. 2015;10:1–25. doi: 10.1371/journal.pone.0140657. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hemes KS, Chamberlain SD, Eichelmann E, Knox SH, Baldocchi DD. A biogeochemical compromise. The high methane cost of sequestering carbon in restored wetlands. Geophys. Res. Lett. 2018;45:6081–6091. doi: 10.1029/2018GL077747. [DOI] [Google Scholar]
- 10.Myhre, G. et al. in Climate Change 2013 (eds Stocker, T. F. et al.) 659–740 (Cambridge University Press, Cambridge, UK and New York, USA, 2013).
- 11.Allen MR, et al. A solution to the misrepresentations of CO2-equivalent emissions of short-lived climate pollutants under ambitious mitigation. npj Clim. Atmos. Sci. 2018;1:1–8. doi: 10.1038/s41612-018-0026-8. [DOI] [Google Scholar]
- 12.Neubauer SC, Megonigal JP. Moving beyond Global Warming Potentials to quantify the climatic role of ecosystems. Ecosystems. 2015;18:1000–1013. doi: 10.1007/s10021-015-9879-4. [DOI] [Google Scholar]
- 13.Steffen W, et al. Trajectories of the Earth system in the anthropocene. Proc. Natl Acad. Sci. USA. 2018;115:8252–8259. doi: 10.1073/pnas.1810141115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Schellnhuber HJ, Rahmstorf S, Winkelmann R. Why the right climate target was agreed in Paris. Nat. Clim. Change. 2016;6:649–653. doi: 10.1038/nclimate3013. [DOI] [Google Scholar]
- 15.Greifswald Mire Centre. Global Peatland Database. https://greifswaldmoor.de/global-peatland-database-en.html (2019).
- 16.Joosten H. The development of peatland emissions until 2030: a reconnaissance. IMCG Bull. 2017;9:4–8. [Google Scholar]
- 17.Intergovernmental Panel on Climate Change (IPCC). in 2013 Supplement to the 2006 Guidelines for National Greenhouse Gas Inventories: Wetlands (eds Hiraishi, T. et al.) (IPCC, Geneva, Switzerland, 2014).
- 18.Dommain R, et al. A radiative forcing analysis of tropical peatlands before and after their conversion to agricultural plantations. Glob. Change Biol. 2018;24:5518–5533. doi: 10.1111/gcb.14400. [DOI] [PubMed] [Google Scholar]
- 19.Kirschbaum MUF. Climate-change impact potentials as an alternative to global warming potentials. Environ. Res. Lett. 2014;9:1–11. doi: 10.1088/1748-9326/9/3/034014. [DOI] [Google Scholar]
- 20.Järveoja J, et al. Mitigation of greenhouse gas emissions from an abandoned Baltic peat extraction area by growing reed canary grass: life-cycle assessment. Reg. Environ. Change. 2013;13:781–795. doi: 10.1007/s10113-012-0355-9. [DOI] [Google Scholar]
- 21.Minkkinen K, et al. Persistent carbon sink at a boreal drained bog forest. Biogeosciences. 2018;15:3603–3624. doi: 10.5194/bg-15-3603-2018. [DOI] [Google Scholar]
- 22.Wichtmann, W., Schröder, C. & Joosten, H. (eds). Paludiculture - Productive Use of Wet Peatlands. Climate Protection - Biodiversity - Regional Economic Benefits (Schweizerbart Science Publishers, Stuttgart, Germany, 2016).
- 23.Blain, D. et al. in 2013 Supplement To The 2006 Guidelines For National Greenhouse Gas Inventories: Wetlands (eds Hiraishi, T. et al.) 3.1–3.43 (IPCC, Geneva, Switzerland, 2014).
- 24.Joos F, et al. Carbon dioxide and climate impulse response functions for the computation of greenhouse gas metrics. A multi-model analysis. Atmos. Chem. Phys. 2013;13:2793–2825. doi: 10.5194/acp-13-2793-2013. [DOI] [Google Scholar]
- 25.McNicol G, Knox SH, Guilderson TP, Baldocchi DD, Silver WL. Where old meets new: An ecosystem study of methanogenesis in a reflooded agricultural peatland. Glob. Change Biol. 2020;26:772–785. doi: 10.1111/gcb.14916. [DOI] [PubMed] [Google Scholar]
- 26.Stocker, T. F. et al. in Climate Change 2013 (eds Stocker, T. F. et al.) 33–118 (Cambridge University Press, Cambridge, UK and New York, USA, 2013).
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The models for projected temperature change were downloaded from IAMC AR5 Scenario Database (available at https://secure.iiasa.ac.at/web-apps/ene/AR5DB). Emission factors and peatland cover data are entirely included in the paper.
The code for the atmospheric perturbation model can be found in the supplementary information. Code was written by A. Günther in the R programming language.