the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Jan 2024
Future methane fluxes of peatlands are controlled by management practices and fluctuations in hydrological conditions due to climatic variability
Abstract. Peatland management practices, such as drainage and restoration, have a strong effect on boreal peatland methane (CH4) fluxes. Furthermore, CH4 fluxes are strongly controlled by local environmental conditions, such as soil hydrology, temperature and vegetation, which are all experiencing considerable changes due to climate change. Both management practices and climate change are expected to influence peatland CH4 fluxes during this century but the magnitude and net impact of these changes is still insufficiently understood. In this study, we simulated the impacts of two forest management practices, rotational forestry and continuous cover forestry, as well as peatland restoration on hypothetical forestry-drained peatlands across Finland using the land surface model JSBACH coupled with the soil carbon model YASSO and peatland methane model HIMMELI. We further simulated the impacts of climatic warming using two RCP (Representative Concentration Pathway) emission scenarios, RCP 2.6 and RCP 4.5. We investigated the response of CH4 fluxes, soil water-table level (WTL), soil temperatures, and soil carbon dynamics to changes in management practices and climate. Our results show that management practices have a strong impact on peatland WTLs and CH4 emissions continuing for several decades, with emissions increasing after restoration and clearcutting. Towards the end of the century, WTLs increase slightly likely due to increasing precipitation. CH4 fluxes have opposing trends in restored and drained peatlands. In restored peatlands, CH4 emissions decrease towards the end of the century following the decomposition harvest residue in the top peat layers, while in drained peatland forests sinks get weaker and occasional emissions become more common, likely due to rising WTL and soil temperatures. The strength of these trends vary across the country, with CH4 emissions from restored peatlands decreasing more strongly in southern Finland and forest soil CH4 sinks weakening most in northern Finland.
- Preprint
(2375 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2023-3037', Anonymous Referee #1, 02 Apr 2024
This manuscript aims to model methane fluxes over peatlands under different management and climate scenarios. In general, the manuscript is well written and easy to read. It is important to understand variation in methane fluxes considering its potent nature.
However, I would like the authors to discuss one important aspect of such modelling studies. From the manuscript, I can see that this is a quite complex modelling exercise involving lot of coupled models and formulations (e.g. for biomass). A lot of published papers are cited to support the modeling framework that was adopted. In reality, one cannot read all those papers, even to understand the models on a stand-alone basis. Understanding the complex coupling between them is even more challenging. There could be significant uncertainties in model outputs even for a stand alone model. Given this, how confident one can be in the results obtained (given the limited number of observation sites) and further, how others can replicate such studies? Even a minor change in model parameters can lead to some major changes in the results!
In addition to this, I believe that the peatlands scenarios modelled in this manuscript are hypothetical in nature (please correct me if I am wrong). Then how beneficial this study will be?
In a more generic manner, given the increase in computational resources, anyone can assume some scenarios, model it and try to publish a paper. How such modelling studies will benefit the scientific community as a whole? If such studies are not justified strongly, this can potentially lead to hundreds of modelling papers. (To the authors: This is a more generic comment I have in my mind seeing the proliferation of modelling studies in the scientific literature. Some of the modelling studies do not even consider the uncertainties or performance of the underlying datasets but make strong claims. Please do not take this as a personal comment. You may please provide justifications to show that your manuscript is not just any one modelling study but it makes a real strong case).
Now, let me discuss some points specific to this manuscript:
Page 3, line 75: What is a nutrient-rich and nutrient poor peatland? How do they differ in terms of functioning?
Page 4, section 2.1: The JSBACH model uses PFT's to represent the vegetation functions in the model. Since this is a case study specific to Finland, are the vegetation PFTs prescribed in the model correspond to the vegetation found in Finland. For example, a same type of vegetation may exhibit different growing phases/phenology etc. depending on the site. Hence, to learn about how CH4 fluxes vary in Finland, the vegetation types and the peatland characteristics coded in the model should match the ground conditions. How to ensure this?
Section 2.1.1, line 100 (and several other places): Reference to a paper under preparation is given. Please do not cite a paper under preparation to support your work. How one can check the scientific validity of the work that is under preparation? Technically, all the novelties in the manuscript under preparation must be presented here.
Page 7: Line 176: Can you give a mean value of the carbon stock as simulated by the model after the spinup, for different regions in Finland, along with standard deviation?
Section 2.2: The management practices assumed for the modelling simulations adhere to the actual ground conditions followed in Finland?
Section 2.3: Can you provide a map of Finland with the different regions marked (as the region names are used frequently in the paper, it is difficult to attach a location to them) and showing the location of the flux measurement sites?
Section 3.1: The first line is too generic and qualitative in describing the model's performance. Please provide some useful numerical metrics to understand the same. From Figure 2, I can see significant differences between the model simulated results and the ground observed fluxes.
Page 17: line 329: Is the amount of methane contained in the fast decaying pools fixed? Then only an initial overestimation can lead to later underestimation. Won't there be any CH4 cycle to replenish this pool? Also, in the same line, the word 'latter' should be changed to 'later'.
Page 4, line 91: In this work....simulations with only ONE PFT per site (the word one is missing).
Citation: https://doi.org/10.5194/egusphere-2023-3037-RC1 - AC1: 'Reply on RC1', Vilna Tyystjärvi, 04 Jun 2024
-
RC2: 'Comment on egusphere-2023-3037', Anonymous Referee #2, 19 Apr 2024
This modeling study investigates how the management of peatlands may change methane fluxes and how this could change in the future. The authors have made an enormous modeling effort to combine different modules in order to simulate water level fluctuations and methane emissions and sinks on peatlands. This is precisely why it is essential to describe the model in more detail, because almost all the equations for all the assumptions are missing. I was not able to recapitulate how things are related and which are the most important equations. The figure helps, but equations are definitely needed too. I even thought that modeling effort could be a separate technical article.
As this is a very local and even management-specific application of this model combination, a more comprehensive evaluation is required. The authors only show a very superficial evaluation on flux towers, not even showing measurements and simulations on a specific site. The only conclusion I can take away is that the model is within the range of the measurements, which is relatively easy to achieve. The study does not show that specific points can be reproduced, nor does it show that management option can reproduce the behavior that was observed. I would also like to see if the changes in water table height are in the right range. I know this is not easy to observe, but it seems easier to do on a local scale.
A general remark, I find it really hard to believe that peatlands are not methane sources from what I know. Anoxic decomposition definitely leads to higher methane concentration in soils. What process drives the sink behavior? Diffusion would merely equalize the concentration in the atmosphere and in the soil. Yes, atmospheric methane concentration will increase, but I couldn’t find out i if this is accounted for. The present study pays particular attention to this.
Specific comments:
Methods: Which PFT’s are meant to be peatland-PFT’s and which plants do they represent?
Experiment design: If you would run the model for longer than 10000 years would the soil carbon pool be different? Is the model able to reach an equilibrium? And did you get an discontinuity in 1951 when switching to JSBACH-FOM? I’m not sure if it is even possible to change from one model to another.
A table for the different scenario would be helpful to understand what the difference of the management scenarios mean. I still not understand what restoration means. Degraded peat usually mean that the peat has dried out due to drainage and the organic matter is decomposing extremely quickly.
Fig. 2 Here I would prefer scatter sites to see if the model has a bias or not. Additionally, I would like to see the dynamics on specific sites as well, because this would support modeled projection result of these fluxes. References are missing for the flux tower, please provide all of them, including the link. It is also not mentioned which period for the measurement were taken. Is that an annual mean or sum? Please be more precise in describing the data. The scale of a) and b) is very different, meaning that the sink is extremely small, this reflect what I expected. Is that really significant for all years that these sites are mainly methane sinks? Peatlands are the largest natural source of methane emissions, so I would have expected this for all sites.
Environmental controls:
It's just a suggestion, but wouldn’t it be more powerful to explain the dynamics based on the measured data first . This would improve the study considerably .
line 260 Do you mean a carbon sink of 2kg/ha or Methane, you wrote C. I could imagine that it will remain a carbon sink. Since it comes up here, I would prefer to see the full carbon dynamics, including organic carbon, carbon dioxide, and methane and importantly also the removal of biomass. As peatlands are carbon sinks, this should be reflected some how, not only in the increase of methane emissions.
Discussion:
line 300 : Whether or not methane becomes a source after clear-cutting depends heavily on the organic material that has been removed. Whether all or part of it is left on the land or removed has an extreme impact on carbon and methane dynamics. How was this addressed in your model and in the model cited? What does YASSO pools mean? This is an important management option that I think affects the results the most, at least in vegetation models.
Citation: https://doi.org/10.5194/egusphere-2023-3037-RC2 - AC2: 'Reply on RC2', Vilna Tyystjärvi, 04 Jun 2024
Status: closed
-
RC1: 'Comment on egusphere-2023-3037', Anonymous Referee #1, 02 Apr 2024
This manuscript aims to model methane fluxes over peatlands under different management and climate scenarios. In general, the manuscript is well written and easy to read. It is important to understand variation in methane fluxes considering its potent nature.
However, I would like the authors to discuss one important aspect of such modelling studies. From the manuscript, I can see that this is a quite complex modelling exercise involving lot of coupled models and formulations (e.g. for biomass). A lot of published papers are cited to support the modeling framework that was adopted. In reality, one cannot read all those papers, even to understand the models on a stand-alone basis. Understanding the complex coupling between them is even more challenging. There could be significant uncertainties in model outputs even for a stand alone model. Given this, how confident one can be in the results obtained (given the limited number of observation sites) and further, how others can replicate such studies? Even a minor change in model parameters can lead to some major changes in the results!
In addition to this, I believe that the peatlands scenarios modelled in this manuscript are hypothetical in nature (please correct me if I am wrong). Then how beneficial this study will be?
In a more generic manner, given the increase in computational resources, anyone can assume some scenarios, model it and try to publish a paper. How such modelling studies will benefit the scientific community as a whole? If such studies are not justified strongly, this can potentially lead to hundreds of modelling papers. (To the authors: This is a more generic comment I have in my mind seeing the proliferation of modelling studies in the scientific literature. Some of the modelling studies do not even consider the uncertainties or performance of the underlying datasets but make strong claims. Please do not take this as a personal comment. You may please provide justifications to show that your manuscript is not just any one modelling study but it makes a real strong case).
Now, let me discuss some points specific to this manuscript:
Page 3, line 75: What is a nutrient-rich and nutrient poor peatland? How do they differ in terms of functioning?
Page 4, section 2.1: The JSBACH model uses PFT's to represent the vegetation functions in the model. Since this is a case study specific to Finland, are the vegetation PFTs prescribed in the model correspond to the vegetation found in Finland. For example, a same type of vegetation may exhibit different growing phases/phenology etc. depending on the site. Hence, to learn about how CH4 fluxes vary in Finland, the vegetation types and the peatland characteristics coded in the model should match the ground conditions. How to ensure this?
Section 2.1.1, line 100 (and several other places): Reference to a paper under preparation is given. Please do not cite a paper under preparation to support your work. How one can check the scientific validity of the work that is under preparation? Technically, all the novelties in the manuscript under preparation must be presented here.
Page 7: Line 176: Can you give a mean value of the carbon stock as simulated by the model after the spinup, for different regions in Finland, along with standard deviation?
Section 2.2: The management practices assumed for the modelling simulations adhere to the actual ground conditions followed in Finland?
Section 2.3: Can you provide a map of Finland with the different regions marked (as the region names are used frequently in the paper, it is difficult to attach a location to them) and showing the location of the flux measurement sites?
Section 3.1: The first line is too generic and qualitative in describing the model's performance. Please provide some useful numerical metrics to understand the same. From Figure 2, I can see significant differences between the model simulated results and the ground observed fluxes.
Page 17: line 329: Is the amount of methane contained in the fast decaying pools fixed? Then only an initial overestimation can lead to later underestimation. Won't there be any CH4 cycle to replenish this pool? Also, in the same line, the word 'latter' should be changed to 'later'.
Page 4, line 91: In this work....simulations with only ONE PFT per site (the word one is missing).
Citation: https://doi.org/10.5194/egusphere-2023-3037-RC1 - AC1: 'Reply on RC1', Vilna Tyystjärvi, 04 Jun 2024
-
RC2: 'Comment on egusphere-2023-3037', Anonymous Referee #2, 19 Apr 2024
This modeling study investigates how the management of peatlands may change methane fluxes and how this could change in the future. The authors have made an enormous modeling effort to combine different modules in order to simulate water level fluctuations and methane emissions and sinks on peatlands. This is precisely why it is essential to describe the model in more detail, because almost all the equations for all the assumptions are missing. I was not able to recapitulate how things are related and which are the most important equations. The figure helps, but equations are definitely needed too. I even thought that modeling effort could be a separate technical article.
As this is a very local and even management-specific application of this model combination, a more comprehensive evaluation is required. The authors only show a very superficial evaluation on flux towers, not even showing measurements and simulations on a specific site. The only conclusion I can take away is that the model is within the range of the measurements, which is relatively easy to achieve. The study does not show that specific points can be reproduced, nor does it show that management option can reproduce the behavior that was observed. I would also like to see if the changes in water table height are in the right range. I know this is not easy to observe, but it seems easier to do on a local scale.
A general remark, I find it really hard to believe that peatlands are not methane sources from what I know. Anoxic decomposition definitely leads to higher methane concentration in soils. What process drives the sink behavior? Diffusion would merely equalize the concentration in the atmosphere and in the soil. Yes, atmospheric methane concentration will increase, but I couldn’t find out i if this is accounted for. The present study pays particular attention to this.
Specific comments:
Methods: Which PFT’s are meant to be peatland-PFT’s and which plants do they represent?
Experiment design: If you would run the model for longer than 10000 years would the soil carbon pool be different? Is the model able to reach an equilibrium? And did you get an discontinuity in 1951 when switching to JSBACH-FOM? I’m not sure if it is even possible to change from one model to another.
A table for the different scenario would be helpful to understand what the difference of the management scenarios mean. I still not understand what restoration means. Degraded peat usually mean that the peat has dried out due to drainage and the organic matter is decomposing extremely quickly.
Fig. 2 Here I would prefer scatter sites to see if the model has a bias or not. Additionally, I would like to see the dynamics on specific sites as well, because this would support modeled projection result of these fluxes. References are missing for the flux tower, please provide all of them, including the link. It is also not mentioned which period for the measurement were taken. Is that an annual mean or sum? Please be more precise in describing the data. The scale of a) and b) is very different, meaning that the sink is extremely small, this reflect what I expected. Is that really significant for all years that these sites are mainly methane sinks? Peatlands are the largest natural source of methane emissions, so I would have expected this for all sites.
Environmental controls:
It's just a suggestion, but wouldn’t it be more powerful to explain the dynamics based on the measured data first . This would improve the study considerably .
line 260 Do you mean a carbon sink of 2kg/ha or Methane, you wrote C. I could imagine that it will remain a carbon sink. Since it comes up here, I would prefer to see the full carbon dynamics, including organic carbon, carbon dioxide, and methane and importantly also the removal of biomass. As peatlands are carbon sinks, this should be reflected some how, not only in the increase of methane emissions.
Discussion:
line 300 : Whether or not methane becomes a source after clear-cutting depends heavily on the organic material that has been removed. Whether all or part of it is left on the land or removed has an extreme impact on carbon and methane dynamics. How was this addressed in your model and in the model cited? What does YASSO pools mean? This is an important management option that I think affects the results the most, at least in vegetation models.
Citation: https://doi.org/10.5194/egusphere-2023-3037-RC2 - AC2: 'Reply on RC2', Vilna Tyystjärvi, 04 Jun 2024
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 448 | 169 | 32 | 649 | 21 | 23 |
- HTML: 448
- PDF: 169
- XML: 32
- Total: 649
- BibTeX: 21
- EndNote: 23
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1