the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Feb 2026
Detecting long-term canopy change and vegetation shifts in northern peatlands using LAI and climate data
Abstract. Peatlands store vast amounts of carbon, yet their canopies are changing under northern warming. We assessed recent vegetation trajectories by analysing greening and browning across northern peatlands using a gap-filled, sensor-independent climate data record of leaf area index (LAI) for 2001–2023. To our knowledge, this provides the first multi-decadal, peatland-specific assessment of canopy trends based strictly on mapped peatlands. Although greening was widespread at the pixel level (77 % of peatlands; greening-to-browning ratio 3.5:1), the area-weighted LAI trend at the map scale was not significant. LAI anomalies were weakly positively correlated with temperature and weakly negatively correlated with precipitation. Higher tree cover was associated with less greening, with a smaller effect observed in areas with deciduous needleleaf forests and under higher precipitation. Decadal variability left a regional, non-linear imprint: most pixels showed no breakpoints, but where present they often were temporally aligned with phase shifts in the Pacific Decadal Oscillation (PDO). Cup-shaped trends were concentrated in West Siberia, whereas hat-shaped trends were widespread across Europe, northeastern Asia, and Canada. Protected peatlands did not show different LAI trends when differences in climate and canopy were taken into account. Overall, recent peatland canopy change was not a uniform increase in greenness but reflected moisture-sensitive, composition-dependent responses modulated by decadal climate variability. Together, these results provide a circumpolar, peatland-specific baseline that clarifies where and why LAI is changing and enables evaluation of how moisture conditions, decadal variability, canopy composition, and protection status relate to recent canopy trajectories in northern peatlands.
- Preprint
(1313 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 09 Jun 2026)
-
CC1: 'Comment on egusphere-2026-109', Kai Yan, 09 Mar 2026
reply
-
CC2: 'Reply on CC1', Iuliia Burdun, 09 Mar 2026
reply
Thank you for the positive review! We will take your suggestion into account when preparing the final version of the manuscript.
Citation: https://doi.org/10.5194/egusphere-2026-109-CC2
-
CC2: 'Reply on CC1', Iuliia Burdun, 09 Mar 2026
reply
-
RC1: 'Comment on egusphere-2026-109', Anonymous Referee #1, 14 May 2026
reply
The article titled “Detecting long-term canopy change and vegetation shifts in northern peatlands using LAI and climate data” explores the vegetation trajectory, namely greening and browning trends, across northern peatlands using satellite-based measurements. The study found that while greening was more common than browning across northern peatlands at the local scale, overall canopy change across the circumpolar region was not statistically significant. Instead, vegetation responses appeared to be shaped by complex interactions among temperature, moisture availability, existing canopy composition, and longer-term climate oscillations, highlighting that peatland responses to climate change are spatially variable rather than uniformly increasing in productivity.
I think additional attention should be paid to two main areas:
- Expansion and clarification of the statistical methods
- Improving the readability and overall digestibility of the study
Generally, this work appears to make a meaningful contribution to the field, and I appreciated the use of mostly accessible and publicly available datasets. Overall, I believe the authors present a relatively complete and thoughtful study, but I recommend revisions prior to publication.
While there are several areas where this manuscript may benefit from minor revisions, I believe the greatest payoff would come from improving the readability of the paper, which from my perspective may require more substantial attention. The findings and methods presented here have relevance across multiple disciplines and have the potential to be of interest beyond a strictly scientific audience, yet some portions of the manuscript are challenging to follow. I believe the paper would be strengthened by reducing reliance on nonessential acronyms, uncommon or highly field-specific terminology, and assumptions that readers already possess the necessary background knowledge, some of which does not feel fully established in the introduction.
In addition, aspects of the statistical approach would benefit from further explanation. While the analyses themselves appear appropriate and seem to be robust, portions of the statistical methodology are described at a level that can feel somewhat opaque, making it difficult for readers to fully understand the rationale behind certain analytical choices or how specific models were implemented. Similarly, some of the statistical results are presented in a way that can feel somewhat "blackbox," where readers are asked to accept the outputs without sufficient interpretation of what those results mean in a practical or ecological context. Providing clearer explanations of method selection, assumptions, and interpretation of key statistical outputs would make the findings more accessible and strengthen confidence in the conclusions.
The results themselves are interesting and have the potential to be widely applicable but would benefit from revisions to the written text to make this work more digestible and improve the communication of what are ultimately compelling findings.
In addition to the above general comments I have the following specific feedback:
- Line 8: “Peatlands store vast amounts of carbon, yet their canopies are changing under northern warming.” The use of “yet” makes it seem like these are directly in contrast or in spite of, which I’m not sure is the strongest way to begin the abstract.
- Line 16-17: cup-shaped trends and hat-shaped trends, would consider more layperson language in abstract since I think these results are interesting and tell a story, but this language may lose some readers. In the main text you have opportunity to define these terms for the reader but there isn’t room for that context in the abstract.
- Line 24: suggest leaving Carbon as Carbon rather than C for readability, doesn’t seem like there’s any benefit to shortening
- Line 25-30: expand on background context, especially status quo vegetation, shifts to vegetation, and carbon sink vs source
- Line 33: either in previous paragraph or here further explain the idea of vegetation self-engineering, also worth expanding/tying back into this in the discussion section more explicitly
- Line 37: either in previous paragraph or here, explain the relationship between specific vegetation types and carbon sink vs source. I think general relationships are known but I think intricacies are important to the story being told through the rest of this paper
- Line 41: New paragraph to break things up but start with something other than “Because”
- Line 54: suggest briefly but directly explaining how LAI will get you to these greening and browning trends in very plain language
- Line 73: wonder if it would be useful to include these datasets in a table with summary information for each
- Line 77-83: expand on how each of the datasets came together to create final dataset and overcome the limitations you identified
- Figure 1: Better explanation needed (and/or different language in legend) in both caption and main text for reader to understand “number of pixels within hexagons”
- Figure 1: might be worth putting lakes/water bodies on as a contextual layer so readers who have a specific area of interest can orient themselves and quickly look at trends in that area. This would be especially helpful in the large contiguous landmasses of Canada/Alaska and Siberia.
- Figure 1: somewhat surprised to see areas that appear to not have peatlands based on this map, even at this resolution, especially in NW Canada and CE Siberia. I know your data sources are described but I wonder if more explanation might be worthwhile or even a description of any limitations by using the methods you did
- Figure A1: this symbology (colour scheme) doesn’t seem intuitive
- Line 113: expansion on these correlation methods? Is PDO with no time lag expected? Is there any concern of the possibility of the multiple comparisons or multiplicity problem emerging? How was this dealt with?
- Line 142: “continent information to account for continental differences in vegetation sensitivity to climatic parameters” such as?
- Line 160: I found this a bit confusing maybe expand or clarify
- Line 164 (methods section as a whole): I think further detail throughout the methods would improve the possible replicability of this study. I feel further detail would be particularly beneficial in 3.1, 3.3, and 3.4
- Line 207: similar comment to line 113
- Line 207-222: I think there should be expansion into the statistical methods used, either here or elsewhere. These methods feel a bit “blackbox” to me and I think further detail is needed
- Line 224: expand and include some of this basic mapping approach alongside figure 1, which is introduced before any discussion on pixels or hexagons
- Figure 2: more contrast and/or larger figure panels may be beneficial here
- Figure 2: also might be worth putting lakes/water bodies on as a contextual layer so readers who have a specific area of interest can orient themselves and quickly look at trends in that area. This would be especially helpful in the large contiguous landmasses of Canada/Alaska and Siberia.
- Line 246: strongly recommend further description of “cup-shaped” and “hat-shaped” and even “linear” to better describe the patterns you have observed as cup/hat are not mainstream enough descriptors to be used standalone. I think this is particularly important given that the plots in figure 3 don’t appear as the classic, textbook examples (often financial) you would see in statistics texts. Improving the description and communication of statistics in this paper would be beneficial
- Figure 5: improve tie-in with main text and expand on description of results seen in figure
- Line 294 (long-term trends in LAI): any consideration for the reasons behind this? Some discussion on vegetation in following sections but wonder if it’s worth tying in works by Loranty et al. (2018), Wang and Friedl (2019), Dearborn et al. (2021), Walker et al. (2021), and Carpino et al. (2025) who discuss specific changes to forest cover in this section of the discussion to strengthen the applicability? Are there vegetation successional changes associated with permafrost thaw? Are there species shifts related to the transition from more forest-dominated to more-wetland dominated peatlands in some of these areas? The discussions are much stronger for the following sections than for this section so it is worth adding to this section
Citation: https://doi.org/10.5194/egusphere-2026-109-RC1 -
CC3: 'Reply on RC1', Iuliia Burdun, 22 May 2026
reply
Thank you for the positive review! We will take your suggestions into account when preparing the final version of the manuscript.
Citation: https://doi.org/10.5194/egusphere-2026-109-CC3
-
RC2: 'Comment on egusphere-2026-109', Anonymous Referee #2, 30 May 2026
reply
The authors around Iuliia Burdun present a study on LAI trends and anomalies on northern peatlands in relation to climate variables for the period 2001-2023. The manuscript is well-written, addresses a relevant topic and is based on a well-executed analysis. It fits the scope of the journal and the quality of the study is generally suited for publication. However, in the present form the manuscript is not as convincing as it could/should be, due to some minor and -from my perspective- one major deficiency, which the authors should clarify in a revised version. Especially the relevance may be further improved. I look forward to such new version and believe it can contribute to an increased understanding on the global trends in these relevant ecosystems.
- My main concern is related to the masking of study area and a potential bias towards greening trends: the masking approach requires at least one “recent” product to show peatland area as defined from visible land cover, i.e. a related vegetation class and, hence, some harder to detect peatlands may be excluded (as long as we define peatlands from the depth of organic soil). This would be different with the global peatland map alone. I could imagine that the masking will exclude more pixels that underwent browning, which may be lost in the recent vegetation signals. Masking may have been different at the beginning of the analysed time period. Likewise, areas affected by fires are excluded, which may also affect a mainly pixels under settings more likely to show browning (higher risk of burning). Ideally, the authors try to take a detailed look on excluded areas and show for some examples if such bias does or does not exist. The minimum requirement -for me- would be a convincing discussion how this may affect the results. (I am aware of the greening-to-browning-ratio argument in line 308f. However, if this is the only argument for the estimates, then the justification of the paper suffers, that we need a closer look on peatlands, only.)
Below some recommendations of minor urgency which may further improve the manuscript:
- Line 27 – also the changes in hydrology as such alter the C dynamics, esp. emissions. Here it sounds as if the reorganisation of plant communities is the driver of C dynamics, which only accounts for sequestration.
- Line 28 – with a focus on peatlands > 45 North you should also mention the active drainage of peatlands together with the fact that the masking excludes peatlands under different management from this study (may come in data section, but should be clear)
- Line 49 – mention those “other ecosystem types”, ideally in relation to their greening-browning trends detected by others.
- The final paragraphs of the introduction may be re-organized: Move sentence “This provides…” (line 54) to the very end of he introduction; state the hypothesis before the research questions; formulate the research questions as objectives (so far they are pretty weakly formulated research questions, especially #1)
- Introduction in general – it reads a little that the “gap” is the main motivation for the analysis, please try to further strengthen the arguments why this study is needed and what may be learned from this focus on peatlands.
- Section 2.1 – please be clearer, how you define peatlands and how this is related to wetlands and masking classes and what your mask in the end includes from the greater entity “peatland”, what it does not include and what this means for the presented results (this may fit in the discussion!). See my main concern.
- 325ff – would this argumentation maybe ask for a split analysis where e.g. the last 10-15 years are looked at individually?
- The discussion should clearer show, why the separate analysis of peatlands was relevant and needed in addition to exiting studies, which leads over to the conclusion…
- … which is in the end a bullet summary and not a conclusion. Conclude by highlighting the value of the outcomes of the paper.
I do agree to the comments of referee #1. I do not share the concern in the discussion that NDVI or similar should be included, but a short paragraph on LAI as compared to the mentioned indices may be added in the discussion
(Disclaimer: I am not qualified to review the formula in section 3.2 and I hope for other referees or the open discussion process to check on this.)
Citation: https://doi.org/10.5194/egusphere-2026-109-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 984 | 567 | 105 | 1,656 | 109 | 129 |
- HTML: 984
- PDF: 567
- XML: 105
- Total: 1,656
- BibTeX: 109
- EndNote: 129
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This study detected long-term canopy change and vegetation shifts in northern peatlands using LAI and climate data. The innovation of the article is now reflected in the peatland map and conclusions about there is no significant trends in peatlands. The study is consistent with the journal’s scope, supported by rich datasets and thorough analyses, and the topic holds clear scientific significance. Overall, this research has certain significance in long-term vegetation shifts in northern peatlands, but it still needs to be modified to improve.
Abstract
Q1: Lins 16-17: What is the means of Cup-shaped trends? And hat-shaped trends? These are not mentioned before. Pls delete is or explain more.
1 Introduction
Q2: Lines 54-56: “This provides, to our knowledge, the first multi-decadal, peatland-specific assessment of canopy trends based strictly on mapped peatlands, rather than generic boreal or Arctic ecosystems.” Could the author elaborate on this point? It's the most innovative aspect of the article and understanding it would be crucial for readers.
2 Data
Q3: Section 2.1: Authors listed some datasets, could you also list the link for these datasets?
3 Methods
Q4: Can you provide a flowchart to help highlight the understanding of the entire study?
Q5: Section 3.1: Can you add some formulas to help explain this part?
4 Results
Q6: Fig. 5: Numbers at bar ends give the count of overlapping pixels. However, this is quite confusing and should be deleted. Additionally, this figure should be optimized, and it looks significantly different from the other figures.
5 Discussions
Q7: The discussion is well designed. However, only LAI discussed here, can you discuss other metrics also (e.g., NDVI, EVI, and GPP)?