the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 03 Sep 2025
Natural wetland methane emissions simulated by ICON-XPP
Abstract. Methane emissions from natural wetlands account for about 1/3 of global methane emissions, and thus have a significant climatic impact due to methane’s high global warming potential. Among the sources of methane, those from natural wetlands have the highest uncertainty, and it is thus of key importance to understand and reduce the uncertainties in the estimates of the wetland emissions to be able to close the global methane budget. Using a coupled land and atmosphere setup, a new implementation of interactive wetlands and wetland methane emissions into the ICON-XPP Earth System Model was used to test the sensitivity of the wetland methane emissions to a number of hydrological and biogeochemical model assumptions of which some are mimicing anthropogenic influences on the Earth System. Averaged over the historical period (1855–2014) the simulated emissions are 166.2 (156.3–181.2) Tg(CH4) yr-1. For the 2000–2012 period the equivalent numbers 182.3 (154.3–205.1) Tg(CH4) yr-1 are in good agreement with estimates from other studies. Wetland methane emissions rise by about 12 % during the historical period, mainly since 1980, an increase which is due to an enhanced carbon cycle caused by the CO2 fertilization associated with rising atmospheric CO2 concentrations. The modeled emissions are very sensitive to changes and assumptions in the model hydrology, some dependencies only revealed through the interaction with the atmosphere (changes in the moisture recycling patterns). Therefore offline land models are only of limited value to test the influence of changes in model hydrology, which is also influenced by changes in terrestrial vegetation.
- Preprint
(2259 KB) - Metadata XML
-
Supplement
(2287 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-3601', Anonymous Referee #1, 27 Oct 2025
-
RC2: 'Comment on egusphere-2025-3601', Anonymous Referee #2, 11 Dec 2025
General comments: In this study, the authors incorporated wetlands and wetland methane emissions into the ICON-XPP model to estimate global methane emissions. They tested the sensitivity of the results using a coupled model simulation and identified some dependencies with the atmosphere. Although there have been some studies that have estimated wetland methane emissions using models, these are mostly carried out offline. Therefore, this study is an interesting addition to our current knowledge. However, some aspects need improving before publication. One of the main concerns is that the method requires more detail. For example, the processes of methane production, oxidation and transport, and the model simulation setups. Without these details, the results were difficult to understand. Specific comments follow:
As mentioned in the introduction, human activities significantly modify surface hydrology (line #32). Incorporating all these activities will be a huge challenge, but would it be possible to include wetland drainage? For example, could we include the percentage of the wetland area that has been drained each year since 1900, if this better represents the wetland area?
Line #198: Could you add the number from Saunois et al. (2016)?
Figure 4: Could you extend the axis range to make the ends of the bars more visible?
DrySurf experiment: The SWR scheme changed the surface water dynamics as well as the precipitation patterns. Did including SWR in the model lead to more realistic wetland extent and precipitation patterns?
I find the WetCrop setup a bit confusing: it could be due to the experiment name, the definition of wetlands or the lack of detailed information. Switching off the drainage means that croplands can be converted to wetlands (with a change in vegetation), increasing both the area of wetlands and their methane emissions. However, what does it mean when you say there are no effects of drainage on the basic soil hydrology below the crops in line #103 and no changes in the carbon cycle compared to the 'Base' in line #259? Are croplands kept wet without any vegetation change from crops to wetland plants? If the croplands are 'wet', why does the wetland area increase together with methane emissions from wetlands?
The same applies to ConstVeg: is the land cover of 1855 kept constant throughout the simulation despite changes to surface water? If so, how do wetland area and methane emissions change?
Section 2.1: What were the temporal and spatial resolutions used to calculate the wetland area fraction? Was the resolution high enough to capture the heterogeneity of permafrost areas such as polygonal tundra or palsa?
Section 2.2: What are the exact processes of methanogenesis? SOM decomposes to 100% CO₂ in aerobic conditions and to 50% CO₂ and 50% CH₄ in anaerobic conditions with 35% efficiency. What is the proportion between acetoclastic methanogenesis and hydrogenotrophic methanogenesis? Do you assume that acetoclastic methanogenesis is dominant? A more detailed description of methanogenesis could be provided.
Section 4.1: To what extent are the wetlands comparable with those in global and regional inventories? If not, what could be causing the uncertainty? Despite the wetland extent potentially being underestimated, global methane emissions are comparable with those in other syntheses. Is methane emission per area then overestimated? A more detailed description of methane production and transport processes would help us to understand this issue.
Line #269: I would be careful with the comment that offline simulations can lead to the wrong conclusion. Depending on the purpose and how the model works, offline and online simulations can have different strengths and weaknesses. For example, interactions between wetlands and the climate can only be observed through online simulations. However, if the area of wetlands and wetland methane emissions, as well as their interaction with the climate, are not simulated well, there can be higher errors in online simulations than in offline simulations, where the extent of wetlands and the climate are well constrained.
Section 5.4: Could you add the numbers of wetland extent from each reference?
There are many typos. These are the ones I found, but there may be more. Please read through the manuscript carefully and correct them.
1980’ies to 1980s in line #18
CO2 to CO2 in line #20
wetland extend to wetland extent in line #68
it’s to its in line #98
were to where in line #112
in creasing to increasing in line #123
1940’ies to 1940s in line #165
the paragraph needs to be organized better to avoid two sentences in parenthesis in line #196
Citation: https://doi.org/10.5194/egusphere-2025-3601-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,717 | 56 | 23 | 1,796 | 42 | 26 | 20 |
- HTML: 1,717
- PDF: 56
- XML: 23
- Total: 1,796
- Supplement: 42
- BibTeX: 26
- EndNote: 20
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
I have several minor, editorial comments:
You say “interactive” as a key feature of your model, but I am almost sure that you mean “iterative”
An interactive model requires some back and forth response of the user, whereas iterative model uses a calculation that repeats estimating the solution until convergence to an accurate solution.
You did not describe any interface for collecting interactive user response. I therefore assume that you mean “iterative”.
On second reading, I think that by “interactive” you mean that the atmosphere is driving and getting feedback from the surface emission model. That will make it a “coupled atmosphere-land surface model”. You do use the term “coupled model” in the discussion. I still don’t think “interactive” is the best term to use here.
Abstract
As your main goal is stated “to test the sensitivity of the wetland methane emissions”, and especially because your model is not globally validated to the methane emissions, please provide more information on what you found in that regard to sensitivity to different anthropogenic effects, and less details about what the exact numbers for methane emission were as determined by the model.
General editing
In L68 and everywhere else in the manuscript, correct “wetland extend” to “wetland extent”. “Extend” is a verb. The noun is “extent”.
You need a comma before every “e.g.”, and before most “which”, after “therefore”, and around “however”.
Methods
L66 – what do you mean by “5”? (“...presented in Marthews et al. (2015) and 5.”)
Subsection 2.2 – were these parameters constant EVERYWHERE? That is highly unrealistic
L97 “via diffusion, ebullition and plant aerenchyma”. Where did you come up with the diffusion parameters, and especially the aerenchyma transport parameters?
Discussion
L268-278 this sentence is way too long. I got lost. Please break it to shorter statements.