A strong mitigation scenario maintains climate neutrality of northern peatlands

•Northern peatlands remain a CO2 sink of ?0.1 Pg C year?1 until 2300 under RCP2.6•Northern become source ?0.2 by RCP8.5•CH4 emissions from northern will increase 5-fold RCP8.5•Modeling peatland resilience, vegetation, and peat quality changes should be improved Intact remove carbon dioxide (CO2) the atmosphere through photosynthesis store in soils waterlogged conditions, while emitting methane (CH4) to atmosphere. The net climate impact depends on relative magnitude these two greenhouse gases. Here, we assess future CH4 balance using five large-scale, process-based models. Our results suggest that policies action, are likely neutral because climate-warming effect is offset cooling sinks. However, if action change not taken, could accelerate global warming projected substantially, may turn sinks sources driven strong drying. Northern 300–600 C, which approximately half underlain permafrost. Climate and, some regions, soil drying enhanced evaporation progressively threatening this large stock. fluxes land surface models explicitly include representation processes. Under Representative Concentration Pathways (RCP) 2.6, for next three centuries. A shift substantial RCP8.5, exacerbate 0.21°C (range, 0.09–0.49°C) year 2300. true might higher owing processes simulated direct anthropogenic disturbance. study highlights importance understanding how trigger high losses peatlands. Global mean temperatures 0.3–4.8°C (relative 1986–2005) end 21st century.1IPCCClimate Change 2013. Physical Science Basis. Working Group I Contribution Fifth Assessment Report Intergovernmental Panel Change. Cambridge University Press, 2013Google Scholar Unabated gas (GHG) emissions, such as those RCP8.5 scenario its extension (see experimental procedures), result 3.0–12.6°C 2300.1IPCCClimate Given amplified mid- latitudes compared with average,2Seneviratne S.I. Donat M.G. Pitman A.J. Knutti R. Wilby R.L. Allowable based regional impact-related targets.Nature. 2016; 529: 477-483Crossref PubMed Scopus (325) Google stability organic (SOC) stocks (300–600 C)3Yu Z. Loisel J. Brosseau D.P. Beilman D.W. Hunt S.J. dynamics since last glacial maximum.Geophys. Res. Lett. 2010; 37: 1-5Crossref (627) Scholar, 4Yu dynamics: review.Biogeosciences. 2012; 9: 4071-4085Crossref (404) 5Hugelius G. Chadburn S. Jackson R.B. Jones M.C. MacDonald G.M. Marushchak M.E. Olefeldt D. Packalen M. Siewert M.B. et al.Large nitrogen vulnerable permafrost thaw.Proc. Natl. Acad. Sci. U S A. 2020; 117: 20438-20446Crossref (90) particular concern. remains, however, poorly understood, only few studies attempting quantification century6Leifeld Wüst-Galley C. Page managed GHGs 1850 2100.Nat. Clim. 2019; 945-947Crossref (57) 7Qiu Zhu Ciais P. Guenet B. Peng role cycle century.Glob. Ecol. Biogeogr. 29: 956-973Crossref (15) 8Müller Joos F. Committed peatlands-continued transient model simulations maximum.Biogeosciences. 2021; 18: 3657-3687Crossref (1) 9Stocker B.D. Roth Spahni Steinacher Zaehle Bouwman L. Prentice I.C. Multiple greenhouse-gas feedbacks biosphere scenarios.Nat. 2013; 3: 666-672Crossref (141) 10Alexandrov G.A. Brovkin V.A. Kleinen T. Yu capacity long-term sequestration.Biogeosciences. 17: 47-54Crossref (10) 11Chaudhary N. Miller P.A. Smith Modelling past, present accumulation across pan-Arctic region.Biogeosciences. 2017; 14: 4023Crossref (27) 12Chaudhary Westermann Lamba Shurpali Sannel A.B.K. Schurgers past pan-Arctic.Glob. Biol. 26: 4119-4133Crossref (18) 13Spahni Stocker Z.C. Transient peatlands: Last Glacial Maximum century.Clim. Past. 1287-1308Crossref (68) Scholar; even less work has been done predicting beyond 2100.8Müller Scholar,14Loisel Gallego-Sala Amesbury Magnan Anshari Benavides Blewett Camill Charman al.Expert assessment vulnerability sink.Nat. 11: 70-77Crossref (39) Scholar,15Gallego-Sala A.V. D.J. Brewer S.E. Friedlingstein Moreton M.J. Björck al.Latitudinal limits predicted warming.Nat. 2018; 8: 907-913Crossref (101) Using evidence literature expert elicitation, al.14Loisel found experts anticipate gains regions degradation al.15Gallego-Sala examined relationships between rates millennium different parameters; they positive relationship growing-season cumulative photosynthetically active radiation (PAR) They would both RCP2.6 scenarios, Arctic continuously temperate decrease. much greater faster than millennium1IPCCClimate Scholar,16Marcott S.A. Shakun J.D. Clark P.U. Mix A.C. reconstruction temperature 11,300 years.Science. 339: 1198-1201Crossref (884) accompanied more frequent severe droughts.17Dai Increasing drought observations models.Nat. 52-58Crossref (2325) With drastic anticipated hydroclimate (particularly water-table position, one most important controls balance), pushed their natural envelope, abiotic factors appear have correlated cannot expected dominant future.18Evans C.D. Peacock Baird Artz R.R.E. Burden Callaghan Chapman P.J. Cooper H.M. Coyle Craig E. al.Overriding water table control emissions.Nature. 593: 548-552PubMed 19Swindles G.T. Morris Mullan Payne R.J. Roland T.P. Lamentowicz Turner T.E. Sim al.Widespread European recent centuries.Nat. Geosci. 12: 922-928Crossref (70) 20Huang Y. Luo Wang Qiu Goll D.S. Makowski De Graaf I. al.Tradeoff drawdown.Nat. 618-622Crossref (5) Process-based models, parameterize hydrological, thermal, biogeochemical processes, can account complex interactions among driving ecosystems (e.g., temperature, moisture, vegetation type, thaw depth, fertilization). useful tools explore responses changes. Yet, represented current generation earth system (ESMs), considered carbon-climate coupled peatland-climate models.9Stocker Previous independent offline already provided insight into it difficult draw robust conclusions critical inputs (i.e., forcing, extent, initiation time) were resulting divergent projections.7Qiu Scholar,8Müller Scholar,11Chaudhary In addition, previous attempts accumulation/storage or focused fluxes,7Qiu 14Loisel 15Gallego-Sala contribution versus CH4) essential information understand influence Earth’s system.9Stocker Furthermore, timescale century too short look at want elucidate full warming, long (decades centuries) lags response slow-turnover pools forcing.8Müller To address above research gaps, conducted multi-model intact (north 30°N latitude) state-of-the-art large-scale ORCHIDEE-PEAT,21Qiu Krinner Aurela Bernhofer Brümmer Bret-Harte al.ORCHIDEE-PEAT (revision 4596), CO2, water, energy daily annual scales.Geosci. Model Dev. 497-519Crossref (29) Scholar,22Qiu Tootchi Ducharne Hastie area Holocene ORCHIDEE-PEAT (SVN r5488).Geosci. 2961-2982Crossref LPJ-MPI,23Kleinen V. Schuldt dynamic wetland extent accumulation: Holocene.Biogeosciences. 235-248Crossref (93) LPX-Bern,8Müller Scholar,13Spahni Scholar,24Wania Ross Integrating model: 1. Evaluation sensitivity physical processes.Glob. Biogeochem. Cycles. 2009; 23: 1-19Google 25Wania 2. 1-15Google 26Müller – investigation.Biogeosciences. 5285-5308Crossref (4) 27Stocker DYPTOP: cost-efficient TOPMODEL implementation simulate sub-grid spatio-temporal wetlands peatlands.Geosci. 2014; 7: 3089-3110Crossref (51) LPJ-GUESS,28Smith Warlind Arneth Hickler Leadley Siltberg Implications incorporating N cycling limitations primary production an individual-based model.Biogeosciences. 2027-2054Crossref (293) Scholar,29McGuire A.D. Christensen T.R. Hayes Heroult Euskirchen Kimball J.S. Koven Lafleur Oechel W. al.An tundra: comparisons observations, process atmospheric inversions.Biogeosciences. 3185-3204Crossref (209) LPJ-GUESS_dynP (“dynP” multi-peat layers)11Chaudhary Scholar,12Chaudhary Scholar,30Chaudhary 2571-2596Crossref (14) Note S1 Tables S1–S3 detailed about each model). forced fixed (Figure 1)31Xu Liu Holden PEATMAP: refining estimates distribution meta-analysis.Catena. 160: 134-140Crossref (187) integrated 10,000 years before following common simulation protocol procedures). same concentration bias-corrected gridded projections IPSL-CM5A-LR general circulation (GCM)32Frieler K. Lange Piontek Reyer C.P.O. Schewe Warszawski Zhao Chini Denvil Emanuel Assessing impacts 1.5 warming–simulation inter-sectoral intercomparison project (ISIMIP2b).Geosci. 10: 4321-4345Crossref (221) mitigation (RCP2.6) high-end emission (RCP8.5) used drive all over reach 1.6°C (1.9°C >30°N) 2100 1986–2005 GCM, followed steady but small trend contrast, dramatic air increasing ?5.9°C (6.7°C ?14.7°C (15.5°C Figure S1). Simulated present-day (1986–2005) total SOC stock 3.2 million km2 covered peatlands31Xu ranges 200 870 S2), brackets C).3Yu ensemble (572 C) falls close observations. Hereafter, report values, parentheses, unless stated otherwise. study, termed biome (NBP), calculated productivity (NPP) minus heterotrophic respiration—anthropogenic disturbances fires modeled. NBP thus represents flux land. 0.11 (0.01–0.22) 1A), matching Hemisphere (0.10–0.15 year?1).5Hugelius Scholar,34Frolking Talbot Treat C.C. Kauffman J.B. Tuittila E.-S. Roulet N.T. Peatlands system.Environ. Rev. 2011; 19: 371-396Crossref (253) trajectory change. RCP2.6, relatively stable uptake rate 1A). By within coming 100–150 varies subregions precipitation RCP8.5. For four out main complexes Hemisphere, i.e., continental Western Canada (CWC) 1B), Hudson Bay Lowlands (HBL) 1C), Europe (NOE) 1D), West Siberian (WSL) 1E), (or nearly neutral) Two (LPX-Bern LPJ-GUESS), cycling, larger future, opposed NPP limited available (ORCHIDEE, LPJ-MPI LPJ-GUESS_dynP). Russian Far East (RFE) 1F), where largest (110% RFE respect 5%–75% other subregions) S1), (LPJ-GUESS) predicts future. All toward shallower tables S3), indicating remains preserved anoxic conditions below table. good agreement empirical extrapolation made RCP2.6. peatlands, persistent projected, comparable estimate (CWC, HBL, NOE, WSL), al. slight contrast mechanistic applied here. accumulated during ignored decomposition deeper (older) peat. When drops level, our exposure aerobic warmer loss carbon.18Evans Scholar,35Dorrepaal Toet Van Logtestijn R.S.P. Swart Weg T.V. Aerts Carbon respiration subsurface accelerated subarctic.Nature. 460: 616-619Crossref (472) Scholar,36Fenner Freeman Drought-induced peatlands.Nat. 4: 895-900Crossref (361) There variation trajectories simulations. This due differences parameterization hydrological thermal (Note S2) consequently wide range terms (Figures S3 S4), S5), S6), Figures S7–S9 show capability reproduce position site level. Peatland development strongly governed local therefore impossible exactly site. LPJ-GUESS better captured interannual, among-sites variability position. lies far outside validate there no way ascertain simulates accurately extreme elevated concentrations quantify cycle, compare (>30°N) lands 2). Estimates (LSM) Inter-Sectoral Impact Intercomparison Project (ISIMIP2b; https://doi.org/10.5880/PIK.2019.012),32Frieler Pion