Introduction

Wetlands are important when assessing the climatic influence of land cover because of their multifold role in climate forcing. As discussed below in more detail, wetlands emit methane (CH4) and volatile organic compounds (VOCs), such as isoprene and monoterpenes, and represent an important sink for atmospheric carbon dioxide (CO2). Through the atmospheric oxidation of VOCs and following gas-to-particle conversion, wetlands are also source of secondary organic aerosols. However, virtually no studies exist on aerosol formation from wetland emissions1. CH4 is the second most important greenhouse gas after CO2. Approximately, 40% of CH4 emissions are from natural sources, mainly wetlands2. Northern peatlands3 (i.e., latitude 40°–70°N) emit about 36 Tg CH4-C per year4,5, which is equivalent to 11% of the total CH4 emission6. CH4 emissions from bog or swamp are highly variable and sensitive to soil temperature, while emissions from fen are additionally controlled by the dominating vegetation species7. Other factors that control the emissions are water table depth and pH. Complex interactions between temperature, other environmental variables and vegetation determine the CH4 emissions from northern peatlands8.

Biogenic volatile organic compounds (BVOC) emitted from wetlands are equivalent to only 1.5–3% of the carbon emitted as CH4. Earlier studies have reported varying amounts of monoterpene emissions from wetlands9,10,11,12. Some studies have reported emissions below the detection limit11 but emissions also up to 146 mg C m−2 h−1 13 have been observed. Monoterpenes were found to be the second most emitted group of BVOCs after isoprene when 82 compounds were measured from 2 to 10 cm depth in boreal peatland13.

Atmospheric aerosol particles and their effects on cloud properties and lifetime are the most uncertain aspect of radiative forcing, and especially is so the effect of secondary aerosol particles on the forcing2. The atmospheric secondary aerosol formation, either from biogenic or anthropogenic origin, is responsible for producing about half of the cloud condensation nuclei (CCN)14.

It has been shown that aerosol formation can be initiated by clustering of sulphuric acid and stabilizing bases15,16,17, highly oxidized organic vapors18 or iodine oxides19. Highly oxidized organic compounds (HOM), that have a very low volatility, have been shown to originate from the oxidation reaction of VOCs20. The formation of organic condensable material from the reaction between alpha-pinene and ozone has been known for long time21, but more exact knowledge of the gas-phase reactions that produce the low volatile compounds has been developed recently18,20,22,23. Several VOCs have been shown to undergo fast auto-oxidation reaction that leads to the formation of low-volatile compounds24,25. After the first molecular-level discovery of extensive involvement of biogenic organic compounds in new particle formation20 many more locations have been found26, where the formation of HOMs is detected in the gas and particulate phase. Although it was suggested already 60 years ago that very low volatile organics are responsible on new particle formation and subsequent growth27,28, the formation mechanism and chemical composition was unknown.

Here we introduce previously not recognized climatically relevant process in the atmosphere above wetlands. In addition to the large emissions of CH4 wetlands emit also VOCs among others monoterpenes and sesquiterpenes. Even though the amounts are small, they are enough to initiate new particle formation and consecutively cloud formation and thereby, potentially, influencing the radiation balance of the planet. With accelerating global warming, large areas under permafrost will thaw and introduce new wetlands with carbon emissions29,30, but also increase the number of newly formed particles. The particle formation aspect of wetlands is not well studied, only one publication was found where new particle formation was observed in the tundra ecosystem in a subarctic birch forest mixed with wetlands with minimal anthropogenic and boreal forest influence1.

Results and discussion

Specific meteorological conditions characterized by clear skies and low wind speeds occur regularly throughout the year. At these conditions, a decoupled layer (or multiple superimposed layers) having a combined depth of about 0.5–4 m form over the wetland and the overlying inversion layers effectively isolate them from the rest of the nocturnal stable boundary layer (SI Figs S1 and S3). We observed the formation of this layer by monitoring turbulence and concentrations of CO2 and CH4 from different heights (1.5 m, 3 m) (see details in SI). By utilizing these measurements, we divided all nights into two groups; decoupling nights and normal nights (during our measurement period decoupling was detected on 40% of the nights). During the decoupling periods, all the emissions from the wetland are captured inside the stable boundary layer and nearly no mixing with the rest of the atmosphere occurs. Such “closed box” conditions made it possible to observe the potential of wetlands to produce new particles.

Due to surface emissions and subsequent chemical reactions, the concentration of 1.2–1.7 nm clusters/particles detected with the Particle Size Magnifier (PSM) increased at the same time with the CH4 concentration (Fig. 1). Since this happens inside the isolated stable layer over the wetland, we conclude that CH4 and the observed clusters originate from the same source, the wetland surface. Also, the nighttime formation rate of 1.5 nm atmospheric clusters increases significantly as a function of CH4 concentration during the decoupling nights (Fig. 2). The formation rate of clusters increased by a factor of 4 as the decouple layering develops.

Fig. 1: Connection between methane and aerosol particle concentrations.
figure 1

Methane concentration (ppm) and 1–2.4 nm particle concentrations (cm−3) during nights with decoupling (blue circles) and nights without decoupling (red circles). Median values of hourly means during night time (0:00–9:00, 18:00-00:00). Panel a) particle size 1–1.2 nm; b) particle size 1.2–1.4 nm; c) particle size 1.4–1.7 nm; d) particle size 1.7–2.4 nm.

Full size image
Fig. 2: Atmospheric aerosol particle formation rate and methane concentration.
figure 2

Connection between the 1.5 nm particle formation rate (cm−3 s−1) and methane concentration (ppm). Diurnal median values of hourly values during the night time (18:00–9:00). Strong correlation does not imply that methane is responsible for enhanced formation rates, instead the emission of methane is a proxy of all emissions from wetland and the hour in consideration for decoupling nights marked in the figure. Blue fitting is only for decoupling nights (J1.5 = 0.5544*CH4–1.033, R = 0.84, p < 0.05) and black fitting for all nights (J1.5 = 0.6026*CH4–1.1355, R = 0.85, p < 0.05).

Full size image

During daytime the aerosol concentrations in sub 3 nm size range (Fig. 3) and formation rate of 1.5 nm clusters/particles became significantly enhanced (Fig. 4) as a function of the CH4 emissions (fluxes). This shows clearly that the CH4 emissions and new particle formation (NPF) are connected to each other, but it does not mean the CH4 is chemically involved in NPF. We can assume that the processes in wetlands, such as microbial activity in the peat, produces both CH4 and terpenes31, and therefore CH4 can be used as proxy for terpene production. However, based on current knowledge the CH4 does not play a role in the chemistry behind NPF, rather the same process in wetland is producing both, CH4 and monoterpenes. In this study, we measured monoterpene emissions of 30 mg C m−2 h−1 (9.5 ng m−2 s−1) (Supplementary Note 2, Fig. S4).

Fig. 3: Day time methane flux and atmospheric aerosol particle concentrations.
figure 3

Daily median values of methane flux (μmol s−1m−2) and cluster concentrations (cm−3) during the light hours (09:00–15:00). Good correlation between the variables indicates the same source and not the causality. Color scale denotes the concentration of sulphuric acid, where no systematic features could be observed. This indicates a minor role of sulphuric acid in small particle formation over wetland. Panel a) particle size 1–1.2 nm; b) particle size 1.2–1.4 nm; c) particle size 1.4–1.7 nm; d) particle size 1.7–2.4 nm.

Full size image
Fig. 4: Connection between methane flux and particle formation rate.
figure 4

Daytime (09:00–15:00) median values of methane flux (\({F}_{C{H}_{4}}\), μmol s−1m−2) and formation rate (J1.5, cm−3 s−1) of 1.5 nm particles measured with PSM. Correlation does not indicate direct causality, but instead an indication of the same origin for the both parameters. Depicted trendline follows an equation \({J}_{1.5}={10}^{(0.91\cdot \log ({F}_{C{H}_{4}})+0.311)}\), R=0.63, p < 0.05.

Full size image

For the climatic relevance, the particles of sub-3 nm in size need to grow substantially larger. The material for the growth is provided by the formation of low volatile compounds. Here we observed correlation between concentrations of HOMs and sub-3 nm particles during the night time over the wetland (Fig. 5 and detailed scatterplots in Supplementary Note 3, Fig. S5–S7). The atmospheric mixing was minimal during the decoupling nights and this created conditions inside the isolated layer, where we observed strong increase in HOM concentrations that originated from the reaction between monoterpenes and O3. These formed HOMs do not have a nitrogen atom in the molecule (denoted in the Fig. 5 as CHO monomer and dimer). In addition, also the HOMs from the sesquiterpene and O3 reaction show clear elevated concentrations during decoupling nights and a strong correlation with sub-3 nm particles. On the contrary, the HOMs with nitrogen in the molecule (denoted as CHON and CHON2) do not have significant increase during the decoupling nights and only the dimers of the HOM group with one nitrogen show a mild correlation with the sub-3 nm particle concentration (Fig. 5). The formation of nitrogen-containing HOMs only occurs when conditions are turbulent and a good mixing with the rest of the atmosphere exists. The same can be observed if looking only at one example day (20th–21th May, 2016 Figs. 6 and 7), where new particle formation together with an increase in the concentration of HOMs without nitrogen was observed. Nitrogen-containing HOMs and sulphuric acid concentration were decreasing during the particle formation event and it can be concluded that in this case, they were not participating in the particle formation and growth.

Fig. 5: Particle number concentration smaller than 2.5 nm and concentrations of different type of HOMs.
figure 5

Median values of hourly means during night hours (18:00–09:00). The HOMs are separated by the formation mechanism (see Table 1 in SI). HOMs formed from monoterpene and sesquiterpene in absence of NOx are clearly responsible in particle formation, while HOMs that contain one or two nitrogen atoms are not well correlated with particles with the exception of CHON dimers. Correlation coefficient (R) and statistical significance (p) are calculated for all nights combined. Panel a) monoterpene oxidation products containing C, H, and O, 10 carbon-containing molecules; b) same as panel a) but 20 carbon molecules; c) monoterpene oxidation products containing C, H, O, and N. 10 carbon molecules; d) same as panel c) but 20 carbon molecules; e) monoterpene oxidation products containing C, H, O, and 2 N atoms, 10 carbon molecules; panel f) sesquiterpene oxidation products containing C, H, and O, 15 carbon molecules.

Full size image
Fig. 6: Nighttime new particle formation and subsequent growth.
figure 6

Measured in Siikaneva 20.5.2016 with NAIS. Panel a) negative ions, panel b) positive and panel c) for neutral particles. Start of the event and the maximum growth period marked with black lines that correspond to the lines in Fig. 3c.

Full size image
Fig. 7: HOM and sulphuric acid time series during NPF period.
figure 7

Mass spectrometric measurements with CI-APiTOF. Monoterpene CHO monomer and dimer HOMs have two peaks one at the start of the event and the other during the maximum growth period. Panel a) monoterpene oxidation products containing C, H, O, and N atoms, 10 carbon atoms containing molecules; panel b) same as panel a) but 20 carbon molecules; c) monoterpene oxidation products containing C, H, and O, 10 carbon atoms containing molecules; d) same as panel c) but 20 carbon molecules; e) sesquiterpene oxidation products containing C, H, and O, 15 carbon molecules; f) isoprene oxidation products containing C, H, and O, 5 carbon molecules; g) monoterpene oxidation products containing C, H, O, and 2 N atoms, 10 carbon atoms containing molecules; h) sulphuric acid.

Full size image

Figure 6 shows that the clusters formed in the decoupled layer above the wetlands contribute to the formation of large particles (instrument upper detection is 40 nm). In contrast with the clustering in the boreal forest during nighttime32 (Supplementary note 4, Fig. S8) these cluster actually grow to sizes larger than 20 nm, contributing to the regional aerosol number budget, as in the morning hours these particles and remaining precursors are mixed upwards into the planetary boundary layer. Although we discovered the phenomena during the decoupling nights, peatlands emits monoterpenes all the time and contributes to total concentrations. In areas where wetlands are dominating, wetland-induced aerosol production can be a significant source of climatically relevant CCN. The whole process from clustering to 100 nm particles will take several hours (typically 10–40 h)33.

In order to upscale the results, the spatial extent of the wetlands now and in the future climate scenarios must be taken into consideration. Current estimates for wetlands north of 30°N are 3.88–4.08 million km2 34,35,36. In the boreal region (north of 50°N) wetlands make up more than 0.5 million km2 (SI Fig. S9). This translates to 13 000 t C h−1 of monoterpene emissions from the boreal region, further, assuming 7% yield for conversion to HOMs20,24,37 and all the HOM is condensing, about 900 t h−1 of organic carbon is condensing onto or forming new particles in the atmosphere over boreal wetlands. Considering the maximum boundary layer heights between about 1 and 2 km in boreal environments, such amounts of condensable organic material would be capable of producing 100-nm diameter particles at rates >1000 cm−3 h−1 over boreal wetlands. This is far more than the particle source strength needed for maintaining the observed populations of >100-nm particles in boreal environments38, indicating that the rate-limiting step for the CCN formation from wetland emissions is the particle formation rate rather than the growth of these particles to larger sizes.

As a summary, we have shown that wetlands emit terpenes and form atmospheric clusters and new aerosol particles. The freshly formed particles contribute to regional NPF events and, due to their further growth by condensable organic material formed from terpenes emitted by wetlands or nearby forested areas, to regional the CCN population. We further showed that the monoterpene emissions and resulting NPF occur concurrently with CH4 emissions from wetlands. The wetland CH4 emissions causes a warming effect on climate while, based on our results, the aerosol formation through wetland monoterpene emissions and oxidation leads to a cooling effect (Fig. 8). The overall magnitude of this balancing effect is difficult to estimate as it is constrained by wetland soil microbiology, boundary layer dynamics, atmospheric chemistry and cluster/aerosol dynamics. The interactions connect the warming (CH4) and the cooling (aerosols, clouds) radiative forcing components together. Large-scale model studies have identified boreal forests as a major source of CCN to the atmosphere, with a potentially important cooling effect due to aerosol-cloud interactions24,39. This study demonstrates that, compared with boreal forests, boreal wetlands are even a stronger areal source of atmospheric clusters. Subsequential growth of these clusters is expected to lead to the formation of new CCN (CCN produced per m2 of land) in the atmosphere, indicating that this effect needs to be considered when estimating the influences of wetlands on regional and global climate. The observed close connection between wetland CH4 emissions and particle formation rate (Fig. 4) offers a means to incorporate the cooling effect due to wetland terpene emissions into any large-scale model capable of simulating atmospheric aerosol dynamics.

Fig. 8: Pathways of potential climatic effects of a wetland.
figure 8

Schematic figure connecting methane (CH4) and volatile organic compounds (VOC, namely mono- and sesquiterpenes) in wetland and the potential (gray) climatic effect. The warming effect depicted in red and cooling one in blue.

Full size image

Methods

We deployed state-of-the-art instrumentation to Finnish wetland, Siikaneva (61°49'59.4“N 24°11'32.5“E, 162 m a.s.l.) where is located a class II ecosystem ICOS (European Integrated Carbon Observation System) station40 and to SMEAR II station (Station for Measuring Ecosystem-Atmosphere Relations)41, in Hyytiälä (61°50'47.1“N 24°17'43.2“E, 181 m a.s.l.) and investigated all the relevant components that are known to influence the new particle formation. The observations were performed on 10th May–15th June 2016. We monitored direct VOC and CH4 emissions from wetland and the concentrations of oxidation products of VOCs, SO2, and O3. We monitored concentrations and chemical composition of atmospheric clusters, aerosols, and air ions from the smallest sizes (0.5 nm) up to 40 nm approaching sizes which can be activated to CCN. As a reference, we utilized SMEAR II station in Hyytiälä, located 5 km east of these measurements. The SMEAR II station is monitoring over 1200 variables, including also the ones measured in the Siikaneva wetland.

The Hyytiälä site is a relatively homogeneous Scots pine stand surrounded by evergreen coniferous forests41, while the Siikaneva site is located in a pristine boreal fen. Peat started to accumulate in Siikaneva after the latest ice age about 9000 years ago and peat depth at the measurement site is approximately 4 meters42,43. Siikaneva fen is characterized by relatively flat topography with a number of vegetation communities and some surface patterning featuring drier hummocks and wetter lawns.

The measurement site consisted of a small hut containing all the instrumentation, which was equipped with sampling inlets at heights of approximately 1.5 m and 3 m. The CI-APi-TOF and APi-TOF, NAIS, PSM, O3 measurements were conducted with the inlet at 1.5 m, while all the meteorological, CH4, CO2, and VOC data were obtained at 3 m.

Data sets from the SMEAR II station at Hyytiälä can be obtained from the AVAA smartSMEAR website (https://avaa.tdata.fi/web/smart)44. A detailed description of the SMEAR II station at Hyytiälä can be found elsewhere41,45. Siikaneva station is part of ICOS (European Integrated Carbon Observation System) network that includes two classes of Ecosystem stations, referred to as Class 1 (complete) and Class 2 (basic) stations. They differ in costs of construction, operation, and maintenance due to the reduced number of variables measured at the Class 2 stations. Siikaneva station is classified as the class 2 ecology site.

Air temperature and relative humidity (RH) were measured with Rotronic HC2 sensor (Rotronic AG, Switzerland) at 2-meter height in Siikaneva. The air temperature was measured at 2 min and RH one minute time resolution. Photosynthetically active radiation (PAR) was measured once in a minute by a Li-Cor Li-190SZ quantum sensor (LI-COR, Inc., USA). Wind speed and direction were measured with Metek USA-1/Gill HS 50 anemometer at 3 meters height. The averaging period for all auxiliary measurements was 30 minutes.

VOC concentrations were measured with a proton transfer time-of-flight mass spectrometer (PTR-TOF, Ionicon) which consists of a proton transfer reaction ion source (PTR) and a TOF-MS46. The PTR instrument is described in detail in literature47,48 and only short description is given here. The PTR consists of a H3O + ion source (hollow cathode discharge in water vapor) and a drift tube where protonated water is mixed with the sample and protons are transferred to the VOC species according to Eq. 1:

$${{{{{{\rm{H}}}}}}}_{3}{{{{{{\rm{O}}}}}}}^{+}+{{{{{\rm{VOC}}}}}}\to {{{{{{\rm{VOCH}}}}}}}^{+}+{{{{{{\rm{H}}}}}}}_{2}{{{{{\rm{O}}}}}}$$
(1)

This charging mechanism works for VOCs with higher proton affinity than that of water, most atmospheric VOC fulfill this requirement47.

The ionized VOCH+ are then passed to the TOF and the mass is determined with an accuracy of 20ppt and resolving power of 3000Th/Th. The VOC is identified using the accurate mass and the prior made calibration. The concentrations of VOCs can be computed from the calibration as the ratio of sample to reagent ion using equation Eq. 2:

$$[{{{{{\rm{VOC}}}}}}]=[{{{{{{\rm{VOC}}}}}}}^{+}]/([{{{{{{\rm{H}}}}}}}_{3}{{{{{{\rm{O}}}}}}}^{+}]\cdot {{{{{\rm{kt}}}}}})$$
(2)

where [H3O +] is the concentration of H3O + in the absence of reacting neutrals, k is the reaction coefficient of the proton transfer reaction and t is the average time the ions spend in the reaction region47. Product kt is obtained from calibration.

Terpene and isoprene emissions are depended on temperature and light49. Accordingly, an increase in both concentrations is observed when approaching summer, indicating an increase in biogenic emissions (Supplementary note 6 and 7. Fig. S10-S12).

The chemical composition of air ions was measured with atmospheric pressure interface (APi) time of flight mass spectrometer50 (APi-TOF, Tofwerk AG). The sample was driven to the instrument through 10 mm electropolished stainless steel tube with a flow rate of 6lpm. The sample was further introduced to APi through a critical orifice with a sample flow of 0.8 l min−1, ions are transported into the TOF to determine their mass to charge ratio(m/Q). The ion beam is focused by two guiding quadrupoles and an ion lens assembly, in three separate differentially pumped chambers, leading into the TOF. The instrument has resolving power of >3000 Th/Th and mass accuracy <20ppm.

Second APi-TOF equipped with chemical ionization (CI) inlet was used to measure the concentration of highly oxidized organic molecules and sulphuric acid. The design of the CI-inlet is similar to one used earlier51,52. The sample was drawn though a ¾” electropolished stainless-steel tube with a flow rate of 10 lpm. Nitrate ions are created by exposing clean air containing nitric acid to a soft x-ray radiation. This sheath flow is then introduced in an ion reaction tube concentric to the sample flow. Nitrate ions in the sheath flow are directed into the sample flow by means of an electric field. The interaction time between ions and sample gas is approximately 200 ms.

The signal of highly oxidized organic compounds (HOMs) is distributed over multiple mass peaks. Example of spectrum is shown in supplementary note 7. Marker peaks were selected to represent each HOM group (Table 1). Concentrations of HOMs were calculated by assuming collision limited charging and the same calibration coefficient (5·1010) as for sulphuric acid according Eq. 3

$${C}_{HOM}=\zeta \cdot \frac{\mathop{\sum }\limits_{m=1}^{n}{I}_{m}}{\mathop{\sum }\limits_{c=1}^{3}{I}_{c}},$$
(3)

where, ζ - calibration coefficient, Im – signal intensity of HOM marker, Ic – signal intensity of charger ion, n – number of marker compounds of corresponding HOM group.

Table 1 Marker peaks for highly oxidized molecules.
Full size table

Calibration for sulphuric acid was performed prior the campaign in the laboratory by oxidation of SO2 with OH to produce the sulphuric acid53. The hydroxyl radical is produced by UV photolysis of water vapor. The final H2SO4 concentration is calculated by a numerical model at the outlet of the calibration source. Comparison of this modelled concentration and the signals measured by CI-APi-TOF yields a calibration factor.

Comparison of Hyytiälä and Siikaneva sulphuric acid concentrations and diurnal patterns are shown in supplementary note 9, Fig. S17 and S18.

Atmospheric ions and total particles were measured with a Neutral cluster and Air Ion Spectrometer54 (NAIS), an instrument for measuring mobility and size distribution. The range for electrical mobility for NAIS is 3.2-0.0013 cm2 V−1 s−1 corresponding to a mobility diameter range of 0.8–42 nm. In total particle mode, the size distribution starts from 2 nm since the charger ions are indistinguishable at smaller particle size ranges55,56.

For measurements of sub-3 nm particles, a Particle Size Magnifier (PSM) in series with a Condensation Particle Counter (CPC) was deployed. This setup enables the detection of single particles without charging the particles57. PSM uses diethylene glycol for activating and growing the particles before entering the CPC. This enables the detection of particles as small as ∼1 nm in mobility diameter57.

Growth Rates (GR) were determined using the 50% appearance time method58,59. This method uses particle size distributions measured by NAIS or DMPS instruments to determine the time when half the concentration maximum is reached in different size bins. The growth rate was obtained by following the evolution of the 50% appearance times as a function of the size distribution during clustering events.

Condensation Sinks (CS) for both measurement sites, Hyytiälä and Siikaneva, were calculated using the particle size distributions measured with the NAIS and DMPS instruments. Since there were measured at ambient conditions a parameterization to correct for ambient hygroscopicity was used60. The CS describes the sink for condensing vapors arising from the available surface area of pre-existing aerosols and can be determined using the Eq. 4:

$$CS=4\pi D\mathop{\sum}\limits_{{d}_{p}}{\beta }_{m,{d}_{p}}{d}_{p}N_{{d}_{p}}$$
(4)

where D is the diffusion coefficient of the condensing vapor, H2SO4 in our case, βm, dp refers to the transition regime correction61, while Ndp is the particle number concentration.

Condensation sink (CS) is usually calculated from DMPS data, since unlike NAIS (2 – 42 nm) it detects aerosol population up to 1000 nm. However, the DMPS system was not available at Siikaneva, and it was verified that calculating CS from NAIS data introduces larger systematic error than using CS calculated from Hyytiälä DMPS (Supplementary note 10, Fig. S19). The comparison of CS calculated from the NAIS shows a good linear correlation between the CSs determined for the two measurement sites (Supplementary note 10, Fig. S20). Error from using CS for Siikaneva calculated from Siikaneva NAIS compared to the Hyytiälä DMPS is larger and justifies the use of the CSs from Hyytiälä DMPS when determining formation rates at Siikaneva. Example clustering event measured simultaneously at two stations is depicted in supplementary note 11 in Fig. S21.

Formation rates (Jdp) were calculated for the Siikaneva measurement site based on the particle number size distribution, the coagulation sink and the determined GR62 with Eq. 5

$${J}_{{d}_{p}}=\frac{dN_{{d}_{p}}}{dt}+CoagS_{{d}_{p}}\;\ast\; N_{{d}_{p}}+\frac{GR}{\varDelta {d}_{p}}N_{{d}_{p}}$$
(5)

Where Jdp is the formation rate of the cluster with diameter dp while CoagSdp is the coagulation sink arising from collisions with larger particles63.

Greenhouse gas flux measurements, in this case carbon dioxide (CO2) and methane (CH4), were conducted with eddy covariance (EC) method at 3 meters height. CO2 and water vapor (H2O) flux measurements were done with high-frequency optical gas analyzer (LI- 7200, LI-COR Biosciences) and 3D sonic anemometer (USA-1, Metek GmbH; CSAT3 Campbell Scientific, Inc.). CH4 flux was measured with CH4 analyzer RMT-200 (Los Gatos Research Inc., Mountain View, California, USA) and 3D sonic anemometer (USA-1, Metek GmbH; CSAT3 Camp- bell Scientific, Inc.). CO2 and CH4 fluxes were measured every half an hour. Uncertain values of CO2 and CH4 fluxes were filtered out with friction velocity values less than 0.1 m s−1. In Hyytiälä, net ecosystem exchange (NEE, CO2 flux) was directly measured using Metek USA-1 anemometer and LI-COR LI-7000 gas analyzer.

Map product used to estimate the area of wetland was produced by European Space Agency Climate Change Initiative (ESA CCI) Land Cover project. The CCI-LC project delivers a consistent global land cover maps at 300 m spatial resolution on an annual basis from 1992 to 2015. The Coordinate Reference System (CRS) used for the global land cover databases is a geographic coordinate system (GCS) based on the World Geodetic System 84 (WGS84) reference ellipsoid.

The CCI-LC combined spectral data from 300 m and 1000 m resolution ENVISAT (MERIS) surface reflectance to classify land cover into 36 land cover types following the United Nations Land Cover Classification System (UNLCCS) legend64,65. The whole archive of MERIS data was first pre-processed for radiometric and geometric corrections, cloud screening, and atmospheric correction with aerosol retrieval. An automated classification process, combining supervised and unsupervised algorithms, was then applied to the full-time series to serve as a baseline to derive land cover maps that were representative of three 5-year periods66. The current updated product is provided in 1-year periods (ESA CCI67).

From CCI-LC maps we calculated the extent of wetlands located north from 30°N to be 1.2 million km2. Equivalent values reported in the literature are 3.88–4.08 million km2 34,35,36. The reliability of land cover estimations based on remote sensing data depends on the sampling, preprocessing and interpretation of the data. The observed discrepancy with earlier studies is therefore reasonable. The area of wetland in boreal region (north form 50°N) is 0.5 million km2.

The highest density of wetland in north from 50°N is located in Siberia (31–180°E, 3.8·105 km2), especially in areas of Khanty-Mansi and Yamalo-Nenets regions (60–90°E, 2.2·105 km2), and in Scandinavia (5–31°E, 6.2·104 km2) (SI Fig. 9).