Methane fluxes measured by eddy covariance on Dutch peatlands
Abstract
Rewetting peatlands is required to limit carbon dioxide (CO2) emissions, however, raising the groundwater level (GWL) will strongly increase the chance of methane (CH4) emissions which have a higher radiative forcing than CO2. Datasets of CH4 from different rewetting strategies and natural systems are scarce, and quantification and an understanding of the main drivers of CH4 emissions are needed to make effective peatland rewetting decisions. We present a large dataset of CH4 fluxes (FCH4) measured across 16 sites with eddy covariance on Dutch peatlands. Sites were classified into six land uses, which also determined their vegetation and GWL range. We investigated the principal drivers of emissions and gap-filled the data using machine learning (ML) to derive annual totals. In addition, Shapley values were used to understand the importance of drivers to ML model predictions. The data showed the typical controls of FCH4 where temperature and the GWL were the dominant factors, however, some relationships were dependent on land use and the vegetation present. There was a clear average increase in FCH4 with increasing GWLs, with the highest emissions occurring at GWLs near the surface. Soil temperature was the single most important predictor for ML gapfilling but the Shapley values revealed the multi-driver dependency of FCH4. Mean annual FCH4 totals across all land uses ranged from 90±11 to 632±65 kg CH4 ha-1yr-1 and were on average highest for semi-natural land uses, followed by paludiculture, lake, wet grassland, and pasture with water infiltration system. The mean annual FCH4 was strongly correlated to the mean annual GWL (R2 = 0.80). The greenhouse gas balance of our sites still needs to be estimated to determine the net climate impact, however, our results indicate that considerable rates of CO2 uptake and long-term storage are required to fully offset the emissions of CH4 from land uses with high GWLs.
https://doi.org/10.5061/dryad.1rn8pk140
Description of the data and file structure
Methane fluxes were measured using eddy covariance systems equipped with Licor LI-7700 gas analysers at several sites across peatlands on the Dutch coastal plain. In addition to the CH4 flux data, supporting measurements of energy and CO2 fluxes, meteorology (air temperature, humidity, pressure), soil temperature and water content, groundwater levels, and redox potential are included. The data covers the years 2020 to 2023, however, data length varies per site.
Files and variables
Files are comma-separated value files (csv) split up by the site ID. Files are provided in half-hourly (hh.zip) and daily timesteps (dd.zip). Half-hourly data contain observed FCH4 fluxes and the gapfilled data. Both baseline- and all-predictor set gapfilled data are provided for sites that were gapfilled. Half-hourly predictor data were filled as described in the manuscript and differed between variables (e.g. meteorology was filled by neighbouring sites and using KNMI data, energy /carbon fluxes by the MDS algorithm, and then mean imputation). Data in the daily timestep were used for figure generation in the manuscript and contain weighted averaged groundwater levels to attempt to control for site representation.
Sites
Site ID | Site name | Land use |
---|---|---|
ANK_PT | Ankeveen | Paludiculture |
ZEG_PT | Zegveld | Paludiculture |
CAM | Camphuys | Semi-natural |
ILP_PT | Ilperveld | Semi-natural |
ONL | Onlanden | Semi-natural |
WRW_SR | Weerribben | Semi-natural |
WRW_OW | Duinigermeer | Lake |
DEM_NT | Demmerik | Wet grassland |
ASD_MP | Assendelft | Pasture WIS |
LAW_MS | Langeweide | Pasture WIS |
BUO | De Burd C | Pasture |
BUW | De Burd T | Pasture |
LDC | Lytse Deelen C | Pasture |
LDH | Lytse Deelen H | Pasture |
HOC | Hommerts C | Pasture |
HOH | Hommerts H | Pasture |
Variables
Flux naming and units generally follow FLUXNET-CH4 conventions.
Missing values are represented by -9999.
Variable | Description | Units |
---|---|---|
TIMESTAMP_END | Timestamp end of the averaging period, used in half-hourly files | YYYYMMDDHHMM |
TIMESTAMP | Timestamp used in daily aggregation files | YYYYMMDD |
FCH4 | Methane flux | nmol m-2 s-1 |
FCH4_F_XGB_BASE | Gapfilled methane flux from XGBoost baseline predictor set | nmol m-2 s-1 |
FCH4_F_XGB_ALL | Gapfilled methane flux from XGBoost all predictor set | nmol m-2 s-1 |
TA_F | Air temperature | deg C |
SW_IN_F | Incoming shortwave radiation | W m-2 |
PA_F | Air pressure | kPa |
WS_F | Wind speed | m s-1 |
PPFD_IN_F | Photosynthetic photon flux density, incoming | umolPhoton m-2 s-1 |
P_F | Precipitation | mm |
RH_F | Relative humidity | % |
VPD_F | Vapour pressure deficit | hPa |
WD_F | Wind direction | degrees |
TS_xxx_F | Soil temperature, depth indicated in cm | deg C |
SWC_xxx_F | Soil water content, depth indicated in cm | % |
REDX_xxx_F | Redox potential, depth indicated in cm | mV |
WTD_x_F | Water table depth, numbered by probe, negative values indicate below the surface) | m |
NEE_F_MDS | Net ecosystem exchange (CO2 flux), gap-filled my MDS | umolCO2 m-2 s-1 |
RECO_NT | Ecosystem respiration | umolCO2 m-2 s-1 |
GPP_NT | Gross primary production | umolCO2 m-2 s-1 |
LE_F_MDS | Latent heat flux | W m-2 |
H_F_MDS | Sensible heat flux | W m-2 |
FCH4_Fx_XGB_xxx | Gapfilled methane flux from XGBoost models 1 to 10, for both baseline and all predictor-sets | nmol m-2 s-1 |