the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Mar 2025
A full year of continuous net soil and ditch CO2, CH4, N2O fluxes, soil hydrology and meteorology for a drained fen in Denmark
Abstract. We here present a detailed dataset of automated greenhouse gas (GHG) net soil and ditch fluxes of carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O) from a drained fen in Denmark covering a full year. The dataset resolves small scale spatial and hourly-daily-seasonal dynamics of GHG soil fluxes. The GHG flux dataset is accompanied by simultaneous time series of soil temperature and moisture, as well as groundwater table depth and covers spatiotemporal gradients in soil hydrological and climatic variability. The GHG fluxes of CO2, CH4 and N2O were measured simultaneously by a high-precision cavity ring down laser spectrometer connected with a novel automated GHG system platform called SkyLine2D (Earthbound Scientific Ltd., UK) that allowed up to 27 individual chamber measurement points along a 24 meter transect. In total 47.483 chamber measurements were completed and after quality control 44.631 CO2 fluxes, 44.099 N2O and 42.515 CH4 fluxes remained.
The average net soil CO2 efflux observed at the site (2.55 μmol CO2 m-2 s-1 or 35 tCO2 ha-1 y-1) aligns with findings from similar drained fens in northern Europe. However, this transect average masks substantial spatial variability and highlights the role of episodic emission bursts related to hydrological variability. N2O fluxes measured at this site were similarly variable in space, but displayed a more dynamic flux behaviour than CO2, where increasing groundwater table depth in response to precipitation during warmer seasons lead to emission bursts of N2O that dominated the annual budget and decreased to near-zero fluxes in drier periods. Soil CH4 fluxes were near-zero and the site overall acted as a small net source, although net uptake was observed as well especially in drier conditions.
Diurnal and seasonal patterns of net soil CO2 and N2O fluxes align with expectations of soil temperature driven processes, but no clear patterns were observed for CH4. Compared to soil GHG fluxes, ditch CO2 and N2O fluxes were 4-fold and 27-fold lower, respectively, while CH4 fluxes were more than two orders of magnitude larger, confirming earlier findings that ditches can be CH4 hotspots, where the CH4 is emitted in bursts with little seasonal variability, including emissions as ebullitions.
The data set is well suited for testing and developing biogeochemical models, with emphasis on the soil thermal-hydrology interactions with the peat C and N cycles.
- Preprint
(4284 KB) - Metadata XML
-
Supplement
(1929 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on essd-2025-123', Daniel Epron, 23 May 2025
With their manuscript, the authors would like to share a full year of data on greenhouse gas fluxes from soil and ditch in a drained fen in Denmark. This is really kind of them, especially since the dataset appears to be of very good quality and obtaining it must have involved both financial and human effort. It is highly respectable, but surprising (for me at least) that they don't seek to promote it with a conventional article before sharing it with the rest of the community. I am not used to reviewing data papers and hope my comments will be helpful nonetheless.
I have only one major concern related to the discarded data. Discarding non-significant regressions as explained line 238 is a problem. If the flux is almost 0, the slope is also almost 0 and the regression, by nature, is non-significant. If a significant number of low fluxes were discarded, then means and cumulated fluxes used for annual budget in Table 3 for example are over estimated (in absolute value). Similarly, “wrong windows” (lines 251-259) can be repositioned. Of course, this takes time, but it is worth doing.
Minor comments
L25: replace by “flux was more dynamics”
L43: “changes” (plural)
L59: explain why night-time fluxes are expected to be overestimated
L166: provide the length of the long tube (important that the colleagues that will work with these data understand well how it was measured)
L166: “On the top” is duplicated. Remove one.
L173: clarify “thus” and the link between “net flux” and weed killing
L241: Use “annual cumulated fluxes” instead of “annual budget”
Fig 9,10: thicker green line will be appreciated
L345: was -> were
Fig 11, 18B, 18C: a horizontal line at intercept 0 will help
Section 3.6 seems oddly structured, starting with the annual cumulated fluxes, followed by individual measures (5 per day) and then the monthly sum
Citation: https://doi.org/10.5194/essd-2025-123-RC1 -
RC2: 'Comment on essd-2025-123', Judith Vogt, 13 Jun 2025
I would like to thank the authors for compiling this extensive dataset and corresponding manuscript.
I think the dataset could be very valuable to the research community. Before publication, I think both the dataset and the manuscript need revision.General comments:
The structure and content of the manuscript reflect that of a classic research article, but the authors chose ESSD for publication. Given the focus on datasets in ESSD, I strongly recommend to revise the datasets to make them easy-to-use and comprehensive for potential users.
I find a bit concerning that the site experienced different levels of disturbance and wonder if that may affect the potential for modelers to use this data? Some discussion could be added in the manuscript, and possibly convincing arguments why this dataset would still be valuable for further use in models and experiments.
Generally, the impact of this study could be emphasized further and throughout the manuscript. Some more in-depth analysis would be nice regarding annual budgets of the site, and driving factors of GHG fluxes, for example, with correlation analyses or similar.
Dataset:
- I would recommend to use csv files rather than excel if the data is meant to be broadly used.
- Were the measurements split up in different files because of varying temporal resolutions (I don’t see the need to separate them by figure)? If so, it might make sense to leave them in separate files. In that case, please use uniform column names, e.g. TIMESTAMP instead of Date, DateTime or Time and avoid several columns with the same name. Otherwise, it would also be nice to have all information in a single file, or at least have one file with flux measurements and accompanying WTD, etc.
- I would recommend to give latitude and longitude of the locations in each file. An indication whether the collar is soil or ditch might also be helpful as well as any other indicators that show differences among the collars (maybe elevation?).
- I don’t see the ebullition fluxes in the dataset. Please clarify how they are reflected in the CH4 column and ideally split up into diffusion and ebullition.Specific comments:
Abstract:
- Especially in the abstract and introduction, I suggest to clarify the directions of fluxes, e.g. “soil-atmosphere/ditch-atmosphere fluxes” and “emissions/source to the atmosphere”.
- note that the doi of the dataset should be given in the abstract for ESSD
- add numbers of average fluxes for CH4 and N2O as is done for CO2, and add standard deviations or similar
l. 26: replace “lead” with “led”
l. 27: clarify what is meant by annual budget – N2O or GHG budget?
l. 30: remove “with expectations”Introduction:
- since it is suggested to use this dataset for models, it might be worth to search the literature and include a paragraph about how/why this dataset is currently lacking and would be a valuable addition to the modelling realm
- the second paragraph only mentions a single measurement unit. It might be worth to mention a few others and make the advantages of the system used in this study a bit clearer.
- two paragraphs of the introduction focus on processing methods, although I don’t think that is part of the main points of this manuscript. Therefore, I would recommend to shift the focus a bit and make the impact of this manuscript a bit clearer.
- in the Methods you describe that the site went through a chain of disturbances/changes which likely affect GHG fluxes. Therefore, this should definitely be made clearer in the introduction as well.
l. 48: add “a”, i.e. allow for a standard number of…
l. 65: in Figure 3, the distance is 30 m, not 24 m – please revise.Materials and Methods:
l. 73: add “in Denmark” and the latitude/longitude of the site
l. 79: replace “makes out” with “forms”
l. 86/Figure 1: The scale bar on the lower left is not visible.
l. 90/Figure 2: This figure could be shifted to the appendix or be removed.
l. 93/Figure 3: Lines and text not visible very well, please increase size and visibility, e.g. by adding white background to black text. “30 m” could be red for clarficiation.
l. 95-97: unclear what the JB numbers are without reference
l. 98: yet another transect length is given. Please stick to one throughout.
l. 99: specify analyser, e.g. “greenhouse gas analyser”?
l. 100-101: does the elevation refer to above sea level? Please specify. Remove repetition of “along the transect/across the transect”.
l. 102: replace “an N-S” with “a north-south”
l. 103: briefly mention what the SkyLine2D system is, add a reference or manufacturer
l. 104: clarify whether the the pallet tanks were simply used to stabilize the system
l. 109: replace “was” with “were” and remove repetition of “along the transect”
l. 110: replace “farmer’s field” with “agricultural field”
l. 118-119/Figure 4: The peat depth is not indicated in the figure, so I think the last sentence in the caption could be removed.
l. 120: I don’t find the section heading “Data variables” fitting here since the variables are not clearly presented, but it is rather described how they were measured as would be done in the Methods section of a classic research article.
l. 121: inconsistency with length of dataset – previously it was 12 months, now 13
l. 123: Please indicate whether there was any snow cover during the study period at some point in the manuscript.
l. 124: revise dates to include a comma between day and year and stick to the same date format throughout the manuscript (also in caption of Figure 12, for example)
l. 127/Table 1: The last column is missing a header – should it be “Data availability” or similar? The footnote 2 could be mistaken with exponent 2 – maybe consider choosing a different character such as asterisks instead.
l. 133-135: Worth mentioning here that these measurements were only conducted at specific collars.
l. 143: replace “collecting” with “measuring”
l. 154-160: While I don’t think section 2.5 is essential for the manuscript, it may be helpful for others in the research community. As a note, links should include the date of last access and abbreviations should be explained (such as SigFox).
l. 161: I think section 2.6 should come earlier in the manuscript since the SkyLine2D system has already been mentioned several times before without further explanation. Also, the distinction into subsections 2.6.1-2.6.3 does not seem overly helpful and they could be merged into one since they all describe the measurement setup. Note that at the beginning of a sentence, numbers should be spelled out. A little restructuring of section 2.6 might help to chronologically answer the questions: What was measured and why? How were things measured? I think some information is very detailed (e.g. some data processing steps in l. 219-226) and could be moved to the appendix to shorten this section a bit and focus on the most relevant information.
l. 169: replace “hit” with “sat on” – does this mean that if the water level was low, air could have entered through the holes?
l. 170-173: I think it would be worth to dedicate a subsection to summarize all the disturbances the site faced for clarification. Was the addition of Glyphosate done to reflect common agricultural practices? Or rather to get the true soil flux? Could this treatment have affected GHGs?
l. 185-189: I think this could be removed or are these variables relevant and given in the dataset?
l. 190/Figure 5: This is a nice figure – make sure all text and numbers are visible. I think it could be moved to the appendix since it shows details of the measurement setup.
l. 206: replace “hz” with “Hz”
l. 207: The precisions seem to be for 5-minute intervals? If so, please indicate.
l. 214: remove quotation marks
l. 215: sentence could be removed
l. 217: remove “the procedure outlined in”
l. 218: I think this sentence could be removed, or correct to “converted to micromole per mole”
l. 228 and 233: Why “at time zero”? The slope would cover a range of timesteps.
l. 237: How were you able to detect mechanical malfunction?
l. 238: What do you mean by “non-significant”? Is this based on a p-value? Please indicate. If low fluxes were generally removed, does that mean that you overestimate fluxes?
l. 241-246: I am surprised that this simple approach to estimate annual budgets was chosen with this high-resolution dataset. Was there a reason why the daily values were not summed over the year instead of taking an average? At least a few different estimates using different simple methods to determine annual budgets could be presented. In my opinion, this point could be a strength of this manuscript and would make the data more comparable with other studies.
l. 247/Figure 6: This figure should be moved to the appendix together with the following text in l. 253-261. As a modeler (which seems to be the main target group here), I might not be super interested in how exactly you processed the data, but want to see the clean data presented.
l. 263: Instead of “Ebullition, e.g. mass flow of CH4” I suggest “Methane ebullition flux”
l. 268 and 269: replace “enclosure-1” by “per enclosure” and clarify if you refer to a time or space component here – do you refer to 5 min? Please clarify.
l. 270: Equation 1 results in units nmol * 1e-6 * m-2, so I think 1e-6 should be removed?
l. 275: It is unclear how often ebullition was detected from the ditch, so the frequency remains unknown. Would it not be adequate to extrapolate ebullition fluxes throughout the day based on the measurements? Or is that what was done? Please clarify and elaborate.
l. 278-279: suggest to remove sentence
l. 279-284: I don’t understand this, please clarify. I’m also not sure whether upscaling is the right term here or if you refer to extrapolation? Does this mean that there were 2 actual out of 5 potentially measured ebullition events? And therefore, for 3 out of 5 measurements, the ebullition flux was zero? If so, those should be considered too. Were there any measurements where you were able to determine both diffusive and ebullition fluxes? Was the concentration burst only observed for CH4 or also CO2?
Figure 7: Again, I think this should be placed in the appendix.Data presentation:
l. 300 and 312: I don’t think wind climate is a common term. Wind regime could be used for example.
l. 301: replace “max” with “maximum”
l. 301-303: A pattern is not super clear in Figure 8, so maybe best to remove this sentence?
l. 307-308: Sentence does not seem very meaningful, so it could be removed.
l. 311/Figure 8: In panel A, how can the maximum speed be lower than the mean? Also, error bars for the mean should be inserted. In panel B, reconsider the legend of the color palette. Only blue colors are visible in the wind rose.
l. 323/Figures 9-12: Consider showing the times of GHG measurements with shaded areas spanning the y axis instead of lines on the x axis. Could the x axis ticks be improved to show each first of the month for example? Same comments apply to the following timeseries figures.
l. 327: Sentence is very vague. Seasonal variation in temperature is typical, not only in Denmark. One could either elaborate more on this, or remove the sentence.
l. 329-331: rephrase to “Monthly ranges of air temperatures (Tab. 2) show >20°C variation between minimum and maximum, except for February, pointing towards large diurnal variations.”
l. 333-335/Table 2: Usually, the mean alongside the standard deviation would be given or standard error, for example. Please add that.
l. 339/Figure 10: What is the message of this figure? The soil temperatures would be expected to be very similar along the transect, unless there are clear differences between the locations. It is a bit unclear whether any of the sensors could for example be below the groundwater level at any point. One could show one figure instead that shows the overall seasonal cycle of temperature for all collars combined. Also, the numbers of collars are not consistent with those given in Table 1 for soil temperature measurements, collars 1 and 18 are missing. Is there a reason for that? Add legends for the line colors in the plots. The different blue colors in panels A-E are difficult to distinguish, please choose others.
l. 347-349: Is this information relevant for this study?
l. 356/Figure 11: If the overall water table depth for the site is to be presented, then the figure does not necessarily help much, since the elevation of each collar is different. Figure 4 does a better job. Maybe the authors could clarify the main message of this figure. Is it relevant or would it fit better in the appendix? Also, it would be worth to insert a horizontal line at depth 0.
l. 394: Here and throughout the manuscript, mean fluxes should always be accompanied by an error estimate, such as standard deviation.
l. 396: I think there may be little spatial variation until the start of summer, but after that, most sites take up methane, while some emit. Can the moisture or water level be considered as drivers for methane at this site then?
l. 402: maybe microbes played a role here?
l. 404-406: is there any information available about nutrient content in soils?
l. 407/Figure 14: This figure is a very nice visualization of the study. Consider replacing Figure 4 with Figure 14 to avoid repetition.
l. 426-428 and 430-431: a bit awkwardly phrased sentences, consider to rephrase slightly for better understanding.
l. 436: This is the first time in this section that results were discussed and compared with those from other studies. I definitely miss a bit more discussion in this manuscript. Please add references and discussion for the other gases as well.
l. 439/Figure 16: Small text on right side of the figure to indicate months is barely visible.
l. 443: Maybe better to state “we observed a clear diurnal cycle” or similar.
l. 444-446: Split up or rephrase sentence to avoid long chain of information.
l. 450/Figure 17: What is shown here? Are the fluxes averaged across collars and then summed? Is WTD also averaged across collars? If they are identical for subplots A-C, then there is no need to show it 3 times. Please revise and clarify in the caption.
l. 456: The relation could be shown explicitly in a correlation matrix or plot. In Figure 17, the relation is not clear.
l. 459 and following: Please make sure there is no repetition of this information. It might make sense to restructure or comprise section 3.6 a bit to avoid repetition.
l. 474/Figure 18: Is it possible to partition the diffusive and ebullition CH4 fluxes (considering your methodology) to see whether the higher CH4 fluxes result from ebullition? Please add horizontal lines at zero.
l. 480: Please be more precise about the frequency/occurrence of ebullition at the ditch. How often did it occur? What was the magnitude of the fluxes? And how does this compare to diffusion?
l. 502: Worth mentioning that ebullition fluxes are very sporadic and that there is no clear temporal pattern to be expected.
l. 511-512: The figure reference “(Fig. 19D)” should probably follow after “zero”, and reference to Figure 18C could be added at the end of the sentence.
l. 514/Figure 20: Please add horizontal line for zero or add shaded area or similar.
l. 519: Is the proportionality based on visual inspection of Figure 20?
l. 546-547: I think a reference is needed here when referring to NO3- measurements which are not presented in this study.
l. 554-555: Might be worth to elaborate on this.
l. 559/Table 3: Footnote about GWP can be removed since the info is given in the table caption. Why is the 20-year GWP given only for the ditch? What are the numbers in parenthesis for the ditch GWP for CO2 and N2O? What are the numbers in squared brackets? It could be worth to add further estimates based on different approaches here, as indicated in earlier comments. Consider using <1 or similar for peat soil CH4 GWP since it is not equal to 0. Could the fluxes also be given as area-weighted (in addition to per-m2) fluxes if the area of peat soil and the ditch is known?Conclusion:
Please list limitations of the study and answer the questions: Would the budgets in a different year likely be the same? What effects does the disturbance history of the site have on the measured parameters? I. e. is the data representative in time and among temperate fens?Supplement:
It is unclear which plot belongs to which collar. Please indicate. Also, why was collar 10/the ditch excluded?Citation: https://doi.org/10.5194/essd-2025-123-RC2
Data sets
Supporting Data for: A full year of continuous net soil and ditch CO2, CH4, N2O fluxes, soil hydrology and meteorology for a drained fen in Denmark Annelie S. Nielsen, Klaus S. Larsen, Poul Erik Lærke, Andres F. Rodriguez, Johannes W. M. Pullens, Rasmus J. Petersen, and Jesper R. Christiansen https://doi.org/10.60612/DATADK/BZQ8JE
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 266 | 42 | 9 | 317 | 19 | 7 | 19 |
- HTML: 266
- PDF: 42
- XML: 9
- Total: 317
- Supplement: 19
- BibTeX: 7
- EndNote: 19
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1