This study presents valuable insights into the greenhouse gas fluxes (CH4, N2O, CO2) from a drained peatland forest following clearcutting. The study employs an innovative approach by combining eddy covariance (EC) measurements, statistical modeling, and surface-type classification using UAV imagery to understand the spatial and temporal variability of gas emissions.
Strengths of the Manuscript:
Limited data: There is currently limited EC data from boreal peatland clearcuts. This paper will provide timely information for decision making.
Novel Approach: The integration of surface-type classification with EC and statistical modeling is a notable strength of the manuscript. The use of UAV-based surface classification and Bayesian inference methods adds significant value to the methodology.
Detailed Data Analysis: The thorough analysis of the seasonal and spatial variability of CH4 and N2O emissions is comprehensive, and the modeling framework provides a solid foundation for understanding the drivers of these emissions in post-clearcut peatland ecosystems.
Suggestions for Improvement:
Estimation of Uncertainty: One critical area that requires attention is the estimation of uncertainty in the results. This is particularly important when dealing with gases of small fluxes, such as CH4 and N2O. The study did not provide an analysis on the uncertainty, and this is crucial as there are multiple sources of uncertainty that could impact the results, such as model uncertainty, gap-filling uncertainty, and system uncertainty from the EC systems. A discussion or quantification of these uncertainties would help strengthen the robustness of your findings. I suggest to consider using Monte Carlo and provide a standard deviation to the flux estimates.
Spatial Partitioning of EC Data: The study rightly acknowledges the spatial nature of the EC source area; however, I believe more attention is needed to the challenges of applying EC data to narrow features like ditches, which contribute less than 2% of the EC footprint. While it is interesting to explore spatial partitioning and the variability across different surface types, it is important to note that such narrow features, like ditches, are often difficult to assess with EC systems. The fluxes from these areas are likely to be diluted by the surrounding landscape, which could result in an underestimation of the actual fluxes. While this is an interesting approach, it should be mentioned that the results may not fully capture the fluxes from these narrow features.
Site Characterization: The manuscript lacks sufficient demonstration of site-specific conditions, particularly regarding the fertility and hydrological status of the study area. Information on these conditions is essential to contextualize the observed fluxes and to understand the underlying environmental drivers. Providing more detailed background on soil fertility and the hydrological conditions at the site would strengthen the manuscript and help interpret the results in a more meaningful way.
Conclusion: Overall, this manuscript offers valuable insights into greenhouse gas emissions in boreal peatland forests post-clearcutting. The methodology is innovative and provides a strong foundation for further research. However, addressing the points outlined above—particularly the estimation of uncertainty, spatial partitioning challenges, and more detailed site characterization—would significantly improve the robustness of the manuscript.
1. Line 34: ” Greenhouse gas (GHG) fluxes have been quantified (Ojanen et al., 2010)”. The citation appears a bit strange to me as there should be many GHG related papers before and after Ojanen et al 2010.
2. Line 47: Rotation forestry (The term R is capitalized)
3. Line 51: What is the “duration” of GHG fluxes? Please explain or use a clearer term for that.
4. The paragraph #41-#56: I think you did a clear illustration on introducing rotation forestry. However, I think this paragraph could be shortened or even combined with other paragraphs. This manuscript does not study the entire rotation period, but just one year after clearcutting. Of course clearcutting is a part of rotation forestry but it should not be a main theme that worths 16 lines to introduce. Specifically, the line regarding DOC and biodiversity of rotation forestry is irrelevant and so could be removed.
5. Line 126-129: Is there any reference to the climate data? Did you do the vegetation inventory and peat depth measurements yourselves?
6. Line 139: “It was completed in June 2021 in the north-western section of the CC area” What is “it”? So the harvesting was primarily done 18-Mar to 1-Apr but was completed in June?
7. Section 2.1: You mentioned that “the site is a fertile and well-drained”, but provided no information about that. It would be beneficial to provide more information, including CN ratio, mean WTD prior to (if available) and after clear cut.
8. Line 155: You mentioned that the EC tower is 3.1 m tall. I assume that the canopy is about zero, then using the 1:100 rule of thumb, the flux footprint 90% should be about 300m. But based on figure 2, the footprint is less than 200m. Is there any reason to this?
9. Line 226-228: So the N2O gapfilling model has the best performance among the three GHGs in terms of R2? I am actually quite surprised of that given the small magnitude of N2O fluxes and the complexity of the gas. In the same paragraph, you mentioned only the input variables for CH4 gapfilling but what variables did you use to gapfill N2O so its performance is so good?
10. Line 323-324: If I understand correctly, you developed the models separated for each land surface type? If so, then how did you derive the soil moisture? Did you use a single value for all land surface types, or did you consider the spatial variation of soil moisture (also soil temperature) as I assume the ditches can behave very differently ?
11. Line 407: Consider using a more updated GWP based on the recent IPCC reports.
12. Figure 2A: Consider removing the green colour for NEE, and showing only as a line. It looks a bit misleading now, for instance, that there is only NEE in winter but no Reco.
13. Figure 2D: You mentioned that you have 3 soil moisture monitoring points in Figure 1. You can consider showing the variation of Tsoil by shades also.
14. Figure 2: You have mentioned in the introduction that a fluctuating WTD could be a hot spot for N2O emissions. Have you considered also showing the time series of WTD?
15. Figure 3: Did you include only measured flux data in the correlation analysis , or also the gapfilled data? If you include also the gapfilled flux data, do you think that this will interfere the correlation results?
16. Figure 3: Why only showing absolute value is higher than 0.25, instead of the statistically significant correlation? Also, it makes sense that P does not show any correlation with any variables if using 30-min data, but there should be some lag effects, so if you present also daily sum/means then the correlation result could be different. Indeed, correlation maps are symmetric so you can remove one side of the diagonal, and consider showing correlation result of daily means or other meaningful things you want to make use of the space.
17. Figure 6: Is it that 𝛾 denotes the strength of gas emissions, and 𝛿 denotes it dependence on the temperature? It seems that the information is not very clear neither in the figure or in the text.
18. Figure 7: If 𝛾 in figure 6 denotes the strength of gas emissions, then Figure 7 seems a bit repetitive. At least I see the relative difference across the land types are the same between 𝛾 in figure 6 and figure 7.
19. Line 526: ”Finally, we calculated the total emissions for CH4 and N2O for the snow free period using the best models”. I assume you mean gapfillig
20. Line 630-631: If you separate the surface types in categories and calculate the 𝛿 and 𝛾 separately for each surface type, then I do not think that adding C:N ratio would improve the model as it is just a constant through the year. But C:N ratio is definitely a very important number for you to explain the spatial variations.
21. Line 721: Please be consistent when writing the year in units (ie either a-1 or yr-1) |
