Summary:

               This is useful study, documenting distribution of peatland permafrost along the Labrador coast of northeastern Canada. The author’s undertake a point-based survey of potential peatland permafrost from satellite imagery and then validate survey points with field and aerial surveys to determine areas of “probable, possible, and unlikely” peatland permafrost. The study is worthy of publication with some revisions. Primarily, authors should address: i) the boundaries of the actual study area; ii) rational for using a point-based survey rather than outlining peatland terrain, along with related issues of scale and differentiation between peatland areas; iii) potential to extend and portray distribution of sporadic permafrost based on results of this study. Other discussion points are raised that may be assessed at the discretion of the authors, including: i) perceived abundance of peatland permafrost distribution in Labrador; extent of past permafrost and loss of permafrost relative to present; consideration of defining point-based features as “small, medium, and large” based on a catagorization; age relations of coastal peatlands and permafrost initiation based upon elevation above present sea level; issue of “mis-identification” of permafrost peatlands along the northern section of the study area; correlation between permafrost peatland and terrain other than palsa bogs mapped by ELC; and, discussion of climatic effects related to formation/preservation of peatland permafrost along the Labrador coast and objectives for further study.

Major questions and comments:

Ln 48-49 – “… suggest that peatland permafrost is more abundant along the coast than in the interior”. This seems to be an important point. Does this study confirm this suggestion? Does the study adequately cover the interior peatlands, or focus primarily on the coast? Despite an underestimation of the coastline peatlands, does this study conclude that permafrost is more abundant along the coast than within the interior?

Ln 65 Another important point. To be clear, the study is a point-based inventory. Does this mean that peatland areas are not outlined, and that no coverage of their extent presently exists? In this case, we do not know the individual area or total area of these peatlands.

Figure 4d – does the distribution of peatland permafrost landforms by MAAT say anything about past or present conditions in terms of temperature, for their development? That is to say, why does frequency decrease with cooler MAATs? What is the optimal MAAT for their formation?

Do unlikely peatland permafrost landform areas say anything about past or recent loss of permafrost? Did unlikely areas have permafrost in the past or did they develop without permafrost?

Ln 71 Study Area. What is the actual study area – being that area that is encompassed by this study? This section suggests that the study area is all of Labrador – suggesting that the inventory of wetlands of interest using satellite imagery is to cover all of Labrador. If this is not the case, then a specific section that defines the actual study area is necessary. On a map of Labrador, authors should show the actual area covered by their study, otherwise this is rather misleading as it seems that all of Labrador is the study area and has been examined. Suggest the inclusion of a Section 2.4 entitled “limits of study area” which clearly shows and defines the spatial limit that this survey encompasses. At the same time, some statement on what this implies is important as it appears that the study only identifies and attempts to validate peatland permafrost along the coast of Labrador, but not inland.

Ln 94 Permafrost distribution. This figure should include the outline of the study area within which the surveys were conducted. In this way, readers will be aware of the area in which the study may attempt to validate permafrost distribution. It seems, in fact, that the results of this study should be sufficient, based on observations, to redefine the distribution of permafrost zones along the coastline based on its findings. This could be an added objective and it seems reasonable that if the surveys found permafrost peatlands along the coastline but not inland – that the extent of sporadic permafrost could be extended along the coastline and shown as an additional result in this study. If the authors feel they do not have enough evidence in their study to extend the sporadic zone at present, then they should suggest what else is needed to do so either in the discussion or the conclusion.

Ln 115 Methods. Again, it is important to define the area along the Labrador Sea and Gulf of St. Lawrence coastline that is actually covered by this study. In essence, the study only identifies and attempts to validate peatland permafrost within these areas – not within all of Labrador. Figure 1 can be used to show contiguous survey areas along coast and can also indicate that inland point features outside of these areas were also investigated.

Ln 115 Methods. The methods section needs to discuss issues of scale. Specifically, how large / how small an area was identified on satellite imagery. Not only the resolution of the imagery, but what is the minimum size of a permafrost peatland that was counted as a peatland complex and, similarly, how large. It seems that this study did not outline peatland permafrost complexes, but simply identified them as point-based features. Does this mean that each feature was contiguous, or does this include multiple features close together. Similarly, how far away does another feature need to be to be counted as a separate feature? As these are indicated only as point features, it is important to provide some methodological constraints on how a feature was included (minimum size) and how it was differentiated from a separate feature (minimum separation distance). It would be very useful it there were also some insight into the size range of these features – even if they were mapped only as point features.

It is not generally clear why a point-based inventory was approached, rather than outlining the potential peatland permafrost terrain units. Perhaps, at least, it could be stated why point-based mapping was undertaking rather than defining polygons and areas.

As a note, it would have been beneficial for the authors to have perhaps differentiated the sizes of the peatland permafrost terrain into a least “small”, “medium” and “large” peatland units with some type of catagorization. For example, in Figure S9 it becomes clear that permafrost peatlands are of different sizes, and may benefit from differentiation. In Figure S10 it is not really clear how one peatland unit is differentiate from another as they are shown only as point features and the boundaries of each a not easily to distinguish. Again, a simple differentiation of the size of each in catagorization would have been beneficial.

Ln 208-212 Even though areas were identified only as point features, something about their size should be included. What was minimum size, what was maximum size? Even point features have separation distances, so what was the minimum separation distance between features?

Ln 280-285. Discussion regarding distribution of permafrost peatland complexes is intriguing, and also opens up additional discussion. Where are data showing which peatland complexes lie below marine limit, and which are above? This is alluded to but not shown. The issue of deglacial history and marine recession history are relevant here, in terms of defining the oldest terrestrial age surface in the study area and, thus, oldest peatlands. It appears that deglaciation of the region was from as early as 11 ka BP, along the coastline and then younger moving inland to about 7 ka BP. At the same time, marine recession was occurring in the southern areas along the coastline. Presumably, along the coastline at certain elevations deglaciation and marine recession were the earliest, and these are the oldest peatlands. So – are the oldest peatlands generally also the ones with likely permafrost? Are they thickest, do they have the most syngenetic ground ice? It would be useful to tie the history of marine recession and deglaciation into this discussion a bot more. At present, this is portion of the discussion very limited and is worthy of further consideration.

Ln 287-299. Again, there seems to be more to say here when speculating on the history of peatland intiation ages within the study area – which most of these products/datasets do not take into consideration (and presently, the authors do not either). Admittedly, few peatland initiation ages exist in the region, though theoretically the youngest may be constrained to near the coast. The authors might consider referring to the following articles as a starting points on understanding peatland ages in the region and their possible influence on permafrost peatland distribution:

Gorham, E., Lehman, C., Dyke, A., Janssens, J. and Dyke, L., 2007. Temporal and spatial aspects of peatland initiation following deglaciation in North America. Quaternary Science Reviews, 26(3-4), pp.300-311.

And:

Dyke, A.S., Giroux, D. and Robertson, L., 2004. Paleovegetation Maps of Northern North America, 18 000 to 1 000 BP. Geological Survey of Canada.

Ln312-313: It seems that this study could go a step further by outlining the proposed extension of sporadic permafrost based on their results. Providing an additional Figure 7 with proposed areas of sporadic permafrost would be a useful addition and seems reasonable based on the extent of the study and the results.

Ln330-333: This may warrant an additional sentence or two for clarification. What is the basis for mis-identification based on? For example, most maps in Fig S5 show greater abundance of wetland or peatland areas in the south than in the north. Is it the absence of mapped peatlands along the coastline in these inventories that leads author’s to suggest that their identified areas here may not be peatland permafrost, but instead lithalsa’s? Or did field visits (Fig. 2) along the northern coastline confirm that these were lithalsa’s or in fact peatland permafrost? In general, the absence of peatlands shown in Fig. S5 suggests that either there are few peatlands here, or they are too small to be mapped at that scale.

Figure 6. Reference source for this map seems odd “audio tape?”. Whereas it is interesting to show palsa bogs mapped by ELC here, were there other terrain types related to peatlands that were mapped too? There seems to be a good agreement between the mapped palsa bogs and peatland permafrost, but what were other areas mapped as? Were these peatland areas that did not contain permafrost or other terrain types? Could be discussed in text if not in figure itself.

This study seems almost purposefully vague about existing weather and climatic conditions occurring within the areas of identified permafrost peatland terrain. Given the adherence of these areas to the Labrador coastline, it is indeed interesting to speculate to what extent a maritime climate influences the distribution of permafrost across the study area. The authors allude to conditions of fog, cloud cover, snowpack and wind being potential factors in their distribution. Presumably, these factors are being examined in site-specific studies. The authors could elaborate somewhat further, in the discussion, and most certainly in the conclusion, for the need to investigate local climatic conditions that may support the presence of permafrost in these areas. In a way, this is similar to the examination of the role of inversions in some mountainous environments for sustaining permafrost. It would be suitable for the authors to provide some insight into the intent and value of local studies to understand the distribution of contemporary permafrost further. In addition, such work could aid in more accurately determining extent of sporadic permafrost along this maritime area.

Figure S3. Not sure that depicting only locations of non-peatland permafrost locations is useful. Perhaps better to include both those that did as well as those that did not.

Minor Edits:

Suggest adding “northeastern Canada” to the end of the title

Ln 12 Change “maps” to “depictions”

Ln 21 Ditto

Ln 27 consider replacing “perennially frozen ground” with “permafrost”

Ln 41 delete “they”

Ln 43 consider replacing “have suggested that peatland permafrost is present” with “have depicted peatland permafrost as present”

Ln 46 change “is” to “are”

Ln 58 change “have  been” to “are”

Ln 60 change “and no” to “with no”

Ln 60 change “efforts have been completed”   to “effort completed”

Ln 75 provide location of coldest MAAT (-11.9C) and  warmest MAAT (+1.5C) for context and, if possible, so locations on Figure 1.

Ln 73-78. Unless provided elsewhere, indicate proportion of snowfall versus rainfall and range in total precipitation.

Ln 87 How can glacial till be deposited following retreat of the Laurentide Ice Sheet, except by another glacial/glaciation? Explain, rephrase or delete.

Ln 96 Try to keep spelling of words like “archaeological” and “paleogeographic” consistent. Decide on preferred spelling and use it throughout.

Ln 141 change “that exceeded” to “exceeding”

Ln 177-178. Change “wetland complex by wetland complex” to “WOI” if appropriate.

Ln 188 Change “was” to “were”.

Ln 189 Delete “of WOIs”

Ln 191 Delete “that was”

Ln 251 95 % - remove space.

Ln 262 Delete “In this, study, we demonstrated that”. Start sentence with “Peatland permafrost …”. Reference Figure 4b at end of sentence.

Ln 265 Provide reference to a figure as supporting evidence.

Ln 535 Reference seems incomplete. Nordicana D?

Ln 538-539 Reference incomplete.

General Comments

The manuscript submitted by Wang et al. presents an interesting approach for compiling an inventory of peatland permafrost occurrences along the Labrador coast. The consensus based approach integrates multiple methods including imagery interpretation and field observations. The overall approach is unique compared to other permafrost mapping studies with its utilisation of extensive field validation and the consensus based approach. It could therefore be a model for similar mapping projects elsewhere.

The inventory generated is compared to other studies that have produced maps of permafrost peatlands. The importance of climate and geomorphological influences are discussed and the MS highlights the importance of local-regional scale mapping as well as the integration of field observations. The MS is generally well written and interpretations and conclusions are sound. The paper would be of interest to a wide audience. I have no major concerns and it should be published following some minor revisions. I do have a number of comments and suggestions for the authors’ consideration in preparation of the revised MS.

Some consideration of the scale of existing maps of permafrost and peatland distribution compared to the scale of the authors’ study is required in the analysis and formulation of the main conclusions regarding adjustment of existing maps for southern Labrador. The maps used for comparison are at a smaller scale (national and circumpolar) than the more local to sub-regional scale mapping presented in the MS. Many of the maps used (or the ones used to develop them) will have minimal mapping units so that the characteristics of smaller units will not be shown on the map. This would be the case for example, with the Heginbottom (1995) which is at a scale of 1:7 million, and to some extent O’Neill et al. (2019) which utilizes similar scale maps in its development. It is therefore not surprising that your results would be a bit different. At a national or circumpolar scale, the 15 km that the authors’ suggest the southern permafrost boundary should be extended, is within the precision of these maps. One of the points that could be made is that the application of national and circumpolar scale maps is not really appropriate for addressing sub-regional to local scale issues including those related to plant and animal habitat or infrastructure scale integrity as has been done in a number of other studies. Although the authors do seem to hint at issues of scale, this aspect could be strengthened in the paper.

The inventory would appear to consist of point observation of frozen peatlands. It is not clear if the area of these features has also been determined. This would be useful for the comparison to existing permafrost and peatland maps which show distribution in terms of areal coverage rather than location of specific occurrence of features. Although the density of peatland complexes likely containing permafrost (number) per 400 km² is shown in figure 3b this is not the same as % areal coverage as shown on other existing maps. This makes it difficult to determine whether the results indicate greater occurrence of frozen peatlands than the maps that are used for comparison in the MS (i.e. comparing apples to oranges). Many of the likely or possible occurrences of peatland permafrost complexes are for example within the sporadic or isolated patches zones shown on the Heginbottom et al. (1995) map which means permafrost is more likely than not to be absent and limited to organic terrain in the case of isolated patches. It is difficult to determine from the results presented whether the map presented in the MS indicates a permafrost distribution that is different from the Heginbottom et al. map. Some further discussion is probably required regarding area of the features identified in the inventory.

I am somewhat curious as to how the maps for comparison in the main paper (figure 5) were chosen especially the circumpolar maps (Hugleius et al. 2020; Olefeldt et al. 2021) rather than some of those included in the supplementary information. Would the larger scale map of Tarnocai et al. (2011) for example (which I believe also includes information on whether peatlands are frozen), be more suitable for comparison in the main paper.

Some clarification on the study area is required. It would seem that the focus is on Labrador (coastal Labrador?) but the authors should clarify if the imagery analysis was done for all of Labrador or only specific areas. Also there appear to be observations outside of Labrador and it is unclear which areas outside of Labrador were included in the imagery analysis. A map clearly showing the area for which imagery analysis was done would therefore be useful. For field-based observations, some information on how sites were chosen beyond accessibility is probably required for the reader to understand whether there is any bias in the site selection and validation.

Additional minor comments are provided below keyed to line number

L2 – Title – would it be better to refer to the “Labrador coast”?

L30-31 – insert “in temperature” between “offset” and “between” (i.e. be clear that the offset is referring to a difference in temperature). You could also add that it is the difference between the frozen and unfrozen thermal properties that is an important factor.

L34 – “assessment of thermokarst….” Is probably better and more inclusive.

L50 – O’Neill et al is a national scale map and is based on integration of a national scale surficial map which will not show local scale distribution of peatlands or other organic terrain.

L60 – There is the peatland map and database which I believe is at least partly based on air photo interpretation of Tarnocai at al. (cited in Supplemental Information).

L275-277 – Way and Lewkowicz (2018) includes ground temperature measurements in Labrador and the thermal offsets for various terrain types. Could you be more quantitative and use these results to strengthen the point you are trying to make regarding importance of thermal offset. James et al. 2013 ERL also discusses the importance of thermal offset in persistence of permafrost in organic terrain.

L278-285 (also figure 4) – With respect to associations with elevation, it might be more important to consider whether the area is above or below the marine limit rather than the elevation itself. Given the marine limit varies with latitude, as described in section 2.2, it would make sense to consider the location with respect to the marine limit. For sites below the marine limit, wouldn’t the time since emergence be a factor as it would influence age of peatland and also length of time over which ground freezing occurs.

L287-299 – Reference is made to model predictions. It might be better to refer to simulations which would be more inclusive as the various studies mentioned use various approaches including compilation/synthesis of existing information.

L298 – The surficial deposits are a key factor influencing drainage and accumulation of organic matter as well as formation of segregated ice. You might consider association of peatland permafrost with surficial deposits as has been done for other parameters in figure 4.

L309-310 - Obu et al. (2019) map represents equilibrium conditions so it doesn’t adequately consider past climate history which is important as you have mentioned in the discussion. Permafrost occurrence will be underestimated, especially in the southern portion of the permafrost zone.

L319 – You need to consider the scale of the maps to which you are comparing your results. Heginbottom et al. is a national scale map and is much at a much smaller scale than your study – 15 km on the national scale mapping is likely within the precision of the map.

References

A number of the citations are incomplete and missing information should be added.

L400-401 – Is this a conference presentation with abstract? Provide the conference details and abstract if that is the case

L404-405 – Incomplete citation. Is this an unpublished report?

L406 – Unpublished report, conference presentation? Provide details.

L413-414 – This is NRC Internal Report No. 82 with 1956 publication date.

L415-416 – Incomplete. This is NRC Technical Paper 449

L432 – Is this correct. Seems like an odd reference for a land survey

L434 – van Everdingen is the editor. Also, you should indicate this is an International Permafrost Association publication of the Terminology Working Group

L441 – Is this from the Quaternary Geology of Canada and Greenland. Add missing citation info.

L442 – This is Map 1880A and it should have a doi number (check GEOSCAN https://geoscan.nrcan.gc.ca/ )

L453 Missing information. This is from the National Atlas (5^th Edition) Geomatics Canada series number MCR 4177. It also has a doi number (check GEOSCAN https://geoscan.nrcan.gc.ca/ )

L535 – Is this the database for the inventory (at Nordicana D?) – There should be additional information including doi number.

L538-539 – Is this a conference presentation/abstract, unpublished report? Provide additional information.

Supplemental Information

Figure S3 – Why only show where permafrost is not present based on 2013-17 study? It would be more useful to also include where permafrost was present during the 2013-17 study.

Figure S5 – I believe Tarnocai et al. (2011) also indicates whether peatland is frozen or unfrozen. Wouldn’t it be useful to show this on the map?

Figure S8 – How useful is this comparison given Obu et al. map is based on an equilibrium permafrost distribution and past climate conditions are not considered? Since permafrost aggradation in this region likely occurred under a colder climate than present, the Obu et al. map will underestimate the permafrost occurrence.

Supplemental References

Some citations are incomplete

L71-72 – Heginbottom et al. – see earlier comment

L86-88 – More information about these publications should be provided. Is the NRCan Land cover map the one described below (it might also be from National Atlas 6^th Edition reference outline series 6409).

Canada's land cover; Latifovic, R. Natural Resources Canada, General Information Product 119e, (ed. version 2015), 2019, 1 sheet, https://doi.org/10.4095/315659

L100-101 – Missing info for Tarnocai et al. This is Geological Survey of Canada Open File 6561 and has a doi number – check GEOSCAN

Summary

In this manuscript, “Significant underestimation of peatland permafrost along the Labrador Sea coastline”, the authors have initiated an interesting observational dataset on peatland permafrost distribution for a region with limited data on the nature and extent of permafrost. As such the inventory method and data set provide a useful contribution to the knowledge of permafrost in the study region. The inventory method provides a straightforward point dataset that identifies occurrences of likely peatland permafrost, explores some environmental factors associated with peatland permafrost occurrence, and compares this fine-scale dataset with broadscale depictions of organic terrain, thermokarst, and ground ice distribution. The dataset provides an empirical basis to guide future investigations or to inform potential revisions of broader-scale depictions of thaw-sensitive permafrost and future modeling efforts. The data presented is straightforward and relatively easy to understand and the analyses are useful. There are a number of items that the Authors should consider prior to publication. These amount to minor to moderate revisions and would help improve the clarity and impact of this study. These general comments are provided below followed by a series of detailed items which I hope the Authors will find useful to guide their revisions.   

General comments

The Introduction is reasonably effective at framing the study but should be further strengthened by better linking the state of knowledge with clearly articulated research questions or hypotheses. This will help to better frame the content of the paper, and provide clear logic behind the methods and analyses that are implemented. In relation to this point, there is a fair bit of data shown in the Supplementary materials, some of which seem central to the paper, while other figures in the main manuscript host relatively small amounts of information (F1, 2a). Some figure content could be better organized to make more economic use of figure space while highlighting the data that best supports key arguments.  

Some minor editorial adjustments and additions to the figures would be helpful to more clearly define the spatial scope of the study. Early in the manuscript, it seemed that the paper developed Labrador-wide datasets, but only later in the manuscript did it become clear that the manuscript was focused on the coastal region as indicated in the title. Also, it would be useful to express whether the inventory was aimed to be exhaustive or whether it is thought to represent a subsample of the total population of the features within the focal area of study.

I think that the paper would also benefit significantly if the point data could be more effectively linked to some spatial characteristics of the peatlands. In this regard, I suggest three points to consider. First, it would be useful to clearly express the rule-base for decisions of how and where researchers dropped points to indicate the presence of a (permafrost) peatland complex. In Figure S9 the points seem to represent discrete features, however, it is less apparent why multiple points are dropped in peatland areas in Figure S10. In relation to this point, I think it would add significant value to the paper if the points could be attributed by a size index describing the peatland. This could be through establishing categories based on the area (discrete/small, basin/medium, landscape/large). Alternatively, or in addition, it would be useful to digitize a random subsample of peatlands to show the size distribution of a sample population. This would better contextualize the point dataset giving the inventory more “depth” and providing a better picture of the areal coverage of permafrost peatlands. This data would also provide the Authors with a solid platform for future analyses. It would be useful to include a table showing the data model describing attributes that were collected by the inventory.

I generally like the comparisons between the data generated by this project and the broad-scale spatial products. I think the comparisons are made in a reasonable manner, despite the difficulty of direct comparison with most of these broad-scale datasets because what they represent can be unclear. Some straightforward quantitative comparisons that show the degree of agreement between empirical permafrost peatland observations and grid cell classifications for the datasets portrayed in Figure 5 or S5 would be useful and should be added to the results section. The implications of these results can remain in the discussion.

Specific Comments.

P1 L10-13. Consider making a clear statement of the general distribution of peatlands across Labrador early in the paper to help frame this study. This added context would help a reader not familiar with the region.

L15 – I think it would be useful to briefly explain what is meant by a wetland and peatland permafrost complex. Does the area of the landform matter?

L21 – It is not clear why the presence of “frost susceptible sediments” is important for peatland permafrost to form. Is it that peatlands typically develop in flat, poorly drained environments often characterized by lacustrine or glaciolacustrine deposits, which also happen to be frost susceptible?

Consider that total peatland counts are not the best way to highlight the relative importance of the phenomenon over a geographical area. While the totals have value in comparing permafrost vs non-permafrost peatlands, reporting the data as a frequency density (count/unit area) is more useful to understand the relative importance of the phenomenon, it can be portrayed spatially, and it can be compared more readily with data from other regions.     

L28. Suggested modification. Add “in the form of” palsas (peat mounds…)

L28-29. Suggested modification for the definition of peat plateau. “variable-sized fields of frozen peat elevated above the general surface of the peatland”

P2L44-51. This narrative is good, but it would also be useful to describe the distribution of peatlands in Labrador (and the coast) to better contextualize the study. There are some nice maps in the supplement but those don’t get introduced until much later in the paper. If peatland distribution was integrated into a map earlier in the main manuscript it would help contextualize the discussion from L44-51.

P2-3L64-65. It would be useful to more clearly indicate the spatial scope of the study. It is implied in P2 L63-65 but should be clarified and shown in Figure 1. 

P3L66. Overall, the introduction is well-constructed and the need for research into peatland permafrost is apparent. Still, the final paragraph could be improved by clarifying the research questions or main hypotheses.

2.2 Physical environment

L86-93. To support this text it would be useful to show the relative proportion of different terrain types in one of the maps.

2.3 Permafrost distribution

P4L95-109.

I find this section to be well-written and informative. It highlights data gaps and provides a nice context for your study. Some of this narrative could be situated in the introduction section to help establish the relevance of your work and to frame clear research questions.

3. Methods

It would be useful to define the study area up front and show it in a figure early in the main manuscript.  

It would be useful to elaborate on the description of Peatland permafrost complexes in the study area with reference to figures early on. On P7 L149-154 you could clarify that variation in elevation was used to assess permafrost presence.

P7. Upon inspecting some of the supplementary materials the Authors should clarify what comprises a WOI, or a point. Was there a rule base that indicates how a researcher identified a discrete “complex”, and when one vs. two points were dropped? For example, the identification of discrete wetlands seems clear on FS9, but the distinction is less obvious on FS10.

P8. With respect to utilizing the DJI Mini 2 as explained in the methods, I would caution promoting a “best practice” since Canadian regulations require maintaining a visual line of sight.  

 Source: https://www.gazette.gc.ca/rp-pr/p2/2019/2019-01-09/html/sor-dors11-eng.html

Visual line of sight see 901.11; also see definitions of VLOS.

Figure S3. It would be useful to show all of the survey points and the flight line.

3.3 Validation: It would be useful to describe the data model that guided the collection of the inventory information. 

Figure 2. Please indicate the study area that bounded the extent of the inventory. Also, please adjust the contrast of the “Not Permafrost Peatland” symbol to improve their visibility.

4. Results

P9 L208-211. Section 4.1 is very brief without much supporting analyses or graphics. Consider integrating this section with the next section.

Supplement Sect. S3. Can the Authors indicate all of the points showing the different WOI categories?

The data in Figure 4 is good and the descriptions are clear. Consider paired plots that normalize the distribution against available terrain within that class. Also, it would be interesting to see a plot of the distribution of peatlands without evidence of permafrost.  

5. Discussion

P14 L278. Permafrost peatlands can also develop in flat sandy areas so that while ice segregation is commonly associated with peatland permafrost it is not a prerequisite. Here I would also suggest referencing the primary literature to support this point rather than a national-scale rule-based model.  

P15-17. Figure 5 and S5 host a large amount of spatial data and the Discussion narrative compares and contrasts this study with modeled outputs of related variables. Systematic comparisons of these data sets should be presented as results and the implications can then be addressed more qualitatively in the Discussion. The comparisons are interesting and should be expressed as a study objective given that Figures 5, and S5 present 13 maps with significant amounts of data aimed at comparing new results from this study with existing mapping data.

6. Conclusions

To reiterate a previous point, I think it is also helpful to present results as a count per area because reporting total numbers of peatland occurrences does not provide a great sense of their spatial coverage or regional importance. Furthermore, counts that are not contextualized by total study area are difficult to compare with other datasets.