Dudek & Pętlicki digitize and georeference three topographic map sheets covering the western portion of Sørkapp Land, Svalbard. They clearly outline the steps they took to digitize the contour lines, generate a raster digital elevation model (DEM), use triangulation points to co-register the historical map to a more modern (1990) DEM of the area, and assess the uncertainty in the final historical DEM using elevation differences to the reference DEM over non-glaciated terrain. A notable contribution of this paper is that it establishes that the contour lines in the IGF PAN map sheets represent the 1961 glacier elevations, rather than the elevations from the 1980s, which is when the maps were updated and published.

In my view, there are two categories of shortcomings in this work. The first pertains to the question of whether this paper presents a substantial new dataset to the community—a question that the Topic Editor raised some skepticism about. In lines 97-100, the authors explain that they digitized three of the ten sheets in the IGF PAN topographic map. But they don’t provide an explanation for why they didn’t digitize the other seven. Since Dudek & Pętlicki reconstruct the historical (1961) geometry of just a handful of the ~1,600 glaciers in Svalbard (König et al., 2014), this manuscript reads more like a nice recipe for digitizing historical topographic maps for analysis of geodetic glacier change, rather than presenting a large new dataset.

If this paper is going to be a template to help others digitize and georeference historical topographic maps, it becomes especially important that the methodology in the paper is careful and robust. That brings me to the second potential shortcoming of this work. In my view, there are three ways that the others could generate improved 1961 elevation reconstructions from the data already available to them. First, the authors use a TIN interpolation of the contour lines to create a DEM directly from the contour map. They could get more accurate glacier change observations by differencing just the contour lines to the reference DEM, and then interpolating the difference map. The reason is that glacier elevation change between two time points (dh) varies in space more smoothly than the topography at any one time point. This kind of problem is discussed tangentially in McNabb et al. (2019). Second, the final co-registered map (Fig. 14) still has large regions of positive dz and large regions of negative dz over ice-free land areas. Figure 14 suggests that the authors could still improve the georeferencing of the map by correcting for regional warping (see comment #3 below for some ideas how to do this). Finally, the authors cite Mertes et al. (2017) and Midgley and Tonkin (2017), two papers that use historical aerial imagery and Structure-from-Motion (SfM) photogrammetry to create historical DEMs. Lines 135-139 of the manuscript make it sound as though the authors already have the aerial imagery over their study area. Why not create a DEM from those photographs, following the methods of Mertes et al. (2017) and Midgley and Tonkin (2017) and many others (see references), and compare those results to the hand-drawn topographic maps?

In summary, although I think that the digitization of historical datasets is an important and worthwhile task, the small coverage of the digitized maps presented here, and the fact that the digitization could be done in better ways given the methods already described in published literature, makes me question whether the manuscript warrants publication in ESSD. However, I think that digitizing more of the map sheets, and following some or all of the 3 suggestions mentioned in the paragraph above, would substantially improve this contribution.

Specific comments:

1. If I recall correctly, the 1990 NP DEM for Sørkapp Land has a spatial resolution of 20 m. That relatively low resolution means that the DEM elevation at the pixel location of mountain peaks will consistently underestimate the peak elevation. How do your georeferencing results compare when you only georeference to steep mountain peaks vs. only georeference to flatter terrain? Also, now that Norsk Polarinstitutt (NP) has released a 5 m regional DEM (2010) for Sørkapp Land (geodata.npolar.no), you might want to experiment with the georeferencing results when you use the 5 m reference. Finally, for your discussion, you might consider evaluating the two time periods (1961-1990 vs. 1990-2010). Were thinning rates and retreat rates similar over those two periods? Have they increased/decreased?

2. You say in lines 135-139 that you analyze the original 1961 aerial photographs from the Norwegian Polar Institute. Can you use structure-from-motion to make a 3D model from those images? This has been done successfully for Svalbard glaciers numerous times. See, for example, Mertes et al. (2017), Midgley & Tonkin (2017), Girod et al. (2018), Kavan (2020), Holmlund (2021).

3. You do the co-registration based on the few dozen “triangulation points” listed in Tables 2-4. How would the co-registration differ if you instead applied the Nuth & Kaab (2011) procedure using all DEM pixels in stable ice-free areas? The reason I ask is that, in Figure 14, there remains a considerable amount of spatial structure in the dz map over land areas (i.e., large regions that are consistently blue, transitioning to large regions that are consistently red). That suggests that you might want to change the way you do the co-registration. For example, you could compute the dx, dy, dz offsets using the Nuth & Kaab (2011) method for individual patches of ~2 km x ~2 km (and you should explore the sensitivity to that window size), then unwarp the map using the vector field you generate, where the “vector field” is the <dx, dy, dz> vectors at the grid of locations where you extracted co-registration chips).

Below are some line-by-line comments, which mostly focus on clarity, language, and style. Feel free to ignore any suggestions that you disagree with.

Line comments:

Line 2: you might consider changing “reliable comparative” to “quantitative”

Line 9: change “from” to “on”

Lines 13-15: this first sentence needs a citation. Also, the word “dynamic” connotes ice dynamics (e.g., the flow of ice, surge behavior, etc.). But it seems like you are trying to say that melting is happening more rapidly in Svalbard than other places in the world. Perhaps you could start by saying that Svalbard is warming more rapidly than elsewhere (Nordli 2020), and then relate that warming trend to negative mass balance (e.g., Nuth et al., 2010; Morris et al., 2020). Or, you could skip straight to the negative mass balance.

Line 26: What is “this scientific field” referring to? Glaciology or remote sensing?

Lines 54-55: Clarify what you mean about aerial images being “more competitive” than terrestrial photographs, and why.

Table 1: Should you clarify in the caption that these mapping campaigns were done by the Norwegian Polar Institute, since, in the main text, you just discussed the Polish mapping efforts?

Table 1 (a more general comment): I wonder if the information in this table would be conveyed more effectively by having a series of 8 simple line maps of Sørkapp Land, lined up side by side and colored to show the regions covered by each photogrammetric mapping campaign?

Line 60: replace “first decade of the 21st century” with “2000s”

Lines 65-74: These two paragraphs would benefit from a little more clarity. Is the IGF PAN topographic map for Sørkapp Land the result of analyzing the Norwegian 1960/1961 photos in a stereoplanograph?

Line 76: Missing a space at the end of the sentence.

Lines 76-77: Perhaps you should cite Ziaja & Ostafin (2015) here? Is it correct to call Sørkapp Land an island if it is still connected to Spitsbergen via the isthmus?

Line 79: why do you say “as many as 14 land-terminating glaciers,” rather than simply “14 land-terminating glaciers”?

Line 97: clarify which topographic map.

Lines 97-100: where are the other 7 sheets? Is there a reason you chose not to digitize those ones?

Line 103: cite Fig. 2 again at the end of this sentence for clarity.

Line 112: what does “made in desk research” mean?

Line 121-122: tell the reader here what the answer is, rather than saying “the question was answered.” Do the contour lines represent the glacier elevations in 1960s or the 1980s? You get to that in lines 131-134, but you might improve clarity by telling the reader up front that the glacier contour lines represent the 1961 surface, and then go into the paragraph of how you determined that (lines 123-130).

Lines 141-144: I suppose you began this research before NPI had released the 2010 DEM for Sørkapp Land. Would that be a better reference dataset, since it is 5 m resolution rather than 20 m resolution?

Figure 4: Readers will see that this dz map resembles a hillshade map, due to delta-x and delta-y offsets between the two DEM datasets (i.e., poor co-registration--Nuth & Kaab, 2011). You should explain that in the caption. You already explain it nicely in the main text (lines 174-176), so just add a brief explanation in the caption, too.

Lines 172-173: You don’t need to say how you subtracted one raster layer from another (i.e., which GIS module you used).

Figure 5: Reference Figure 2 so people know where sheet 8 comes from. Also, in Figure 2, you might consider adding (a), (b), and (c) and clearly labeling which sheet is which.

Lines 177-179: You don’t need to tell the reader which of the 3 sheets you worked on first.

Lines 180: Add “elevation points” after “195”

In Figures 5, 6, and 8 (the maps showing triangulation points for the 3 sheets), it would be helpful if you plotted vector arrows pointing in the direction of the dx, dy offset to the NPI reference map. That would let readers see if there are consistent patterns of warping across the map sheets.

Lines 214-216: Rather than referring to “visual assessment” of the model accuracy, can you give a quantitative metric of the elevation accuracy. For example, the root mean square error (RMSE) between the 2 DEMs on ice-free land?

Lines 224-226: This sentence will be more compelling if you provide some stats. For example, use the NPI DEM to say XX% of the glaciated area in Sørkapp Land has slopes < 20 degrees.”

Line 230: remove “for obvious reasons”

Line 242: to improve clarity, you could say “The mean elevation difference (the bias) between…”

Line 291: Missing space after “zone”

Line 308: You might want to cite Nuth et al. (2013) here, since it investigates similar datasets across Svalbard.

Lines 310, 312: replace “overflight” with “survey”

Line 313: Be specific: say “structure-from-motion (SfM) photogrammetry” or

structure-from-motion (SfM)-multi-view stereo (MVS)” rather than “modern methods”

Line 341: remove “very”

Line 343: what does it mean for an air mass “to be in effect”?

Lines 367-369: Rephrase this sentence for clarity and remove the word “ignorance.” Make it clear that you are saying that, even though the map sheet is labeled as 1984, the glacier contour lines reflect the 1961 elevations.

Lines 384-386: You could remove this final sentence of the manuscript.

References

Girod, L., Nielsen, N. I., Couderette, F., Nuth, C. & K ̈aa ̈b, A. Precise DEM extraction from Svalbard using 1936 high oblique imagery. Geoscientific Instrumentation, Methods and Data Systems 7, 277 (2018).

Holmlund, E. S. Aldegondabreen glacier change since 1910 from structure-from-motion photogrammetry of archived terrestrial and aerial photographs: utility of a historic archive to obtain century-scale Svalbard glacier mass losses. Journal of Glaciology 67, 107–116 (2021).

Kavan, J. Early twentieth century evolution of Ferdinand glacier, Svalbard, based on historic photographs and structure- from-motion technique. Geografiska Annaler: Series A, Physical Geography 102, 57–67 (2020).

Nordli, Ø. et al. Revisiting the extended Svalbard Airport monthly temperature series, and the compiled corresponding daily series 1898–2018. Polar Research (2020).

Nuth, C., Moholdt, G., Kohler, J., Hagen, J. O. & Kaab, A. Svalbard glacier elevation changes and contribution to sea level rise. Journal of Geophysical Research: Earth Surface 115 (2010).

Nuth, C. & Kaab, A. Co-registration and bias corrections of satellite elevation data sets for quantifying glacier thickness change. The Cryosphere 5, 271–290 (2011).

Nuth, C. et al. Decadal changes from a multi-temporal glacier inventory of Svalbard. The Cryosphere 7, 1603–1621 (2013).

McNabb, R., Nuth, C., Kaab, A. & Girod, L. Sensitivity of glacier volume change estimation to DEM void interpolation. The Cryosphere 13, 895–910 (2019).

Mertes, J. R., Gulley, J. D., Benn, D. I., Thompson, S. S. & Nicholson, L. I. Using structure-from-motion to create glacier DEMs and orthoimagery from historical terrestrial and oblique aerial imagery. Earth Surface Processes and Landforms 42, 2350–2364 (2017).

Midgley, N. & Tonkin, T. Reconstruction of former glacier surface topography from archive oblique aerial images. Geomorphology 282, 18 – 26 (2017).

Morris, A., G. Moholdt, and L. Gray. "Spread of Svalbard Glacier Mass Loss to Barents Sea Margins Revealed by CryoSat‐2." Journal of Geophysical Research: Earth Surface 125.8 (2020): e2019JF005357.

Ziaja, Wieslaw, and Krzysztof Ostafin. "Landscape–seascape dynamics in the isthmus between Sørkapp Land and the rest of Spitsbergen: Will a new big Arctic island form?" Ambio 44.4 (2015): 332-342.

This manuscript presents a method for how to digitize and georeference archival maps, and finally evaluates their potential for quantifying changes in glacier geometry on Sørkapp Land, Svalbard. For this exercise, three topographic map sheets from the Institute of Geophysics of the Polish Academy of Sciences (IGF PAN) published in 1987 were used together with a reference dataset from 1990. The 1987 map sheets have contour lines based on aerial photos from 1961 (from the Norwegian Polar Institute; NPI), and the 1990 dataset consists of a 20-m-resolution digital elevation model (DEM) and a glacier outline vector layer (also NPI).

Overall, this is a nice study with clearly described methods, walking us through the different steps of source data processing, verification, and data fitting, as well as presents the final DEM and examines the glacier elevation and areal change. However, there are some issues that need to be addressed, mainly considering the dataset size and the potential for improving the quality of the digitization/georeferencing. I have listed my main concerns and suggestions for improvements in the specific comments below. If these questions are resolved, I think the paper could be a good contribution to the journal.  

Specific comments  

1. Why did you not digitize all ten IGF PAN map sheets? Digitizing these map sheets is a valuable contribution, but it would be most useful to have all of them compiled. Especially considering the initial confusion with the contour lines (i.e., that they represented the 1961 elevations, without being corrected based on field observation before publication in the 1980’s). I strongly recommend digitizing the rest of the map sheets to provide a complete updated dataset with greater areal coverage. If choosing not to, please provide an explanation for why you decided to process only three (you describe this nicely in your response to the editor), as this is not obvious to the reader. To help illustrate that these are the most valuable map sheets (for glaciological research), show all of them in Figure 2 and mark the three you present in the paper.
2. The method section would benefit from having a figure showing the workflow, describing all the steps. That would provide a good overview of the methodology and make it more user friendly. Also, I am a bit skeptical to whether the method is innovative enough, and whether the final map (Figure 14) is really satisfactory? There are still areas with both large positive and negative dz, suggesting that the georeferencing could still be improved. Have you considered supplementing with other methods? You refer to studies using structure-from-motion to create historical DEMs from aerial imagery (Mertes et al., 2017; Midgley et al., 2017). Why not try that, using the NPI aerial photographs from 1961, to compare to the topographic maps? Another option could be DEM production by digital stereophotogrammetry on the 1961 images. For methods, see e.g., Korsgaard et al. (2016) and references therein.
3. Why not compare the 1961 and 1990 data to the 2010 data from NPI? The spatial resolution is higher (5 m for the 2010 DEM vs. 20 m for the 1990 DEM), and it would allow you to do a two-step comparison (1961 vs. 1990, and 1990 vs. 2010), to see if the retreat and/or thinning rates have varied between these periods, and whether they have accelerated or not. Further, you could even use the 1936 oblique photos, since from what I can tell from Table 1, the entire peninsula was covered also during that survey? Several studies compare the 1936/38 maps to the 1990 DEM (e.g., Nuth et al., 2007; Girod et al., 2018), so I strongly suggest that you compare more than two years (i.e., by adding 2010 and/or 1936), since that would add something extra to this paper.

Overall syntax and structure

The structure is generally good, but the readability would benefit from having a native English speaker reading though this manuscript. Sometimes the word choices are not optimal, the sentence structure not correct, or the sentences too long (e.g., line 80-84). This makes it somewhat difficult to understand the message, without re-reading some of the sentences. In the introduction, the paragraphs are very short, often only two sentences each. Merge some of them to get a clearer structure and give the text a better flow. I would remove the word ‘glacier’ after all of the glacier names, since ‘breen’ in the end of all names already indicates that those are glaciers (in Norwegian). Abbreviate Norwegian Polar Institute to NPI after the first use.

Study area

Start by introducing Svalbard, the influence of the West Spitsbergen Current, strong climate gradients etc., and how the ice masses vary between different parts of the archipelago. Then move on to the Hornsund area and the glaciers there. You could also show or describe some climate data from Hornsund, to see how the temperature/precipitation have changed from 1961 to 1990. Also introduce the concept of surging glaciers.

Source material

You mix present and past tense. Decide on one (I recommend past tense) and stick to it.

Line by line comments

Line 3: materials -> data

Line 8: glaciers of -> glaciers on (this is reoccurring in several places)

Line 9-10: Remove sentence about dataset availability. Enough to have this information in the ‘Data Availability’ section

Line 13-15: This sentence needs to be supported by 2-3 references. Also, do you with ‘more dynamic’ in the first part of the sentence mean faster? Or in terms of ice dynamics? This needs to be clarified. Reference suggestions: Nordli et al. (2014), Isaksen et al. (2016), Schuler et al. (2020)

Line 19: huge -> crucial

Line 20: Give some examples of ‘traditional research methods’, e.g., in situ stake mass balance measurements

Line 23: Support this statement with a couple more references, e.g., Jacob et al. (2012) and Nuth et al. (2013)

Line 24: Change to ‘The use of remote-sensing has a number of advantages’

Line 44: Change to ‘the University of Wrocław were held there (Zyszkowski, 1982)’

Line 47: has -> have

Line 51-52: Change to ‘including primarily the Gåsbreen area’

Line 57: Norwegian Polar Institute (NPI)

Line 59-61: Do you mean that the next map covering the entire peninsula was not published until in the first decade of the 21^st century? Also, better to write out the year (2010) instead of ‘the first decade of the 21^st century’. Add reference

Line 64: 49 years later -> in 2010. Additional comment: if there are data with the same spatial extent from 2010, why don’t you compare the 1961 and 1990 glacier extents to the 2010 extents as well? It would be interesting to compare the glacier areas, and see if the rate of retreat and thinning have increased or decreased etc.

Line 68: add the spatial coverage (XXX km²) of this map

Line 69: Enough to use the abbreviation (IGF PAN) here

Line 73: attempts -> aims

Line 73-73: Change to ‘geometry of glaciers on the western Sørkapp Land peninsula’

Line 78-79: change to ‘It hosts 14 land-terminating and 4 tidewater glaciers’

Line 85: Remove ‘s’ in the end of ‘Körberbreen glaciers’

Line 88: 'outflow glaciers' should be 'outlet glaciers'?

Line 93-94: Change to ‘They formerly served as tributary glaciers to Samarinbreen, but as its snout receded, they split from it and today constitute separate calving glaciers’

Line 97: The IGF PAN topographic map

Line 98-100: Remove ‘These were the following sheets’ and put ‘(No. 5 –Hornsund, No. 8 – Gåsbreen, and No. 10 – Bungbreen; Fig. 2)’ in brackets.

Line 102: Change to ‘The topographic map sheets presented the relief, permanent and periodic watercourses, …’

Line 106: colour from orange (land) to blue (glacier). Additional comment: is it blue or purple?

Line 111-112: Explain what you mean by ‘somewhat non-standard mean’ – what was done and in comparison to what (the standard way)? And what do you mean by ‘in desk research’?

Line 116: Change he first ‘extent’ to 'degree'?

Line 117: Is “completed in the field” a quote from the map? Otherwise rephrase.

Line 123: Remove line break

Line 124-127: Change to ‘These studies were based on photos from NPI’s photogrammetric overflight over the west of Sørkapp Land in the summer of 1960 (Table 1) and resulted in a publication that included a map 125 showing the hypsometric variation of Gåsbreen and a hillshade that was valid for 1960 (Schöner and Schöner, 1996).’

Line 127: Another -> A (since referring to other glaciers)

Line 136: Add reference to the NPI photos

Line 142: Add reference to the NPI data after first sentence

Line 144: ‘slightly less’ – Can this be quantified?

Line 145: Divide into 4 Methods, and 4.1 Source data processing and evaluation of output data accuracy

Line 146: stages -> steps

Line 153: Does a newer version of R2V allow data to be saved as shapefiles and to be georeferenced? In that case, why did you not use the newer version?

Line 154: resultant -> resulting

Line 161: Remove ‘(northern hemisphere, zone 33)’

Line 162: Remove ‘later in this work’

Line 165: GRID -> grid

Line 168: Remove ‘with the working name DEM IGF 1961’

Line 171: Add reference

Line 174-186: Make this into one paragraph, and remove the sentence starting ‘Work began with the correction of sheet 8…’

Line 180: Change to ‘the locations of the elevation points were assessed’

Line 180: Remove ‘as many as’ and add ‘elevation points’ after 95

Line 181: about 50 elevation points? Not exactly 50, which are the same as 50 of the points in the map sheet?

Line 187-188: shifted southeastwards by how much?

Line 193: (Barna and Warchol, 1987)

Line 195: Remove ‘because much of it was covered by the Greenland Sea’

Line 202: fragment -> portion

Line 205: adding a few points

Line 228: Remove ‘natural’ before ‘processes’

Line 230: Remove ‘for obvious reasons’

Line 231: Change to ‘After considering the aforementioned criteria, the part of the IGF PAN model selected’

Line 247: Change to ‘The measure for examining the extent and pattern of glacier retreat’

Line 248: The research -> This analysis

Line 251: Change to ‘During the study period’

Line 255: most intense -> greatest

Line 256: Change to ‘For these glaciers’

Line 257-258: Rephrase sentence to follow better after the previous one

Line 258-261: This comparison to the LIA does not belong in Results, but rather Discussion

Line 261-262: Repetition of narrowing of the lowest/lower parts of the glaciers. Rephrase

Line 268: shrank -> decreased

Line 268: amounted to -> was

Line 273: situation -> setting

Line 281: Change to ‘at their termini’

Line 286-287 Change to ‘the area of Nordfallbreen decreased by only 0.03 km² (3.6%), which is among the lowest values in the entire region’

Line 289: Remove one of the ‘However’

Line 290: Space missing after ‘zone’

Line 292: ‘receded by only 120 m’ - Is not 120 m in 30 years a lot?

Line 297: shrinkage -> areal decrease

Line 298: Remove ‘though this did vary between glaciers’. Change to ‘For the region’s largest glaciers’

Line 299: shrinkage -> decrease

Line 299: leading into -> calving into

Line 300: for the glaciers instead of in the glaciers

Line 309: Could add Nuth et al. (2007, 2013) and Holmlund (2021)

Line 313: Specify which modern methods. Could also cite Girod et al. (2018) and Holmlund (2021)

Line 315-316 Change to ‘The accuracy of simulations prognosing changes in glacier volumes based on dynamics models depends largely on that those models have been initialised correctly’

Line 323-324: How did the temperature change during this period? Relate this to meteorological data to explain the changes

Line 325-326: For the same time period (1961-90), or something else?

Line 331-332: extending upwards to -> reaching. Additional comment:  what is the equilibrium line altitude in this area?

Line 354-355: Change to ‘Nordfallbreen is adjacent to Nordfallet (824 m a.s.l.) to the south, which shades it against the sun while also providing it additional supply by avalanches’

Line 359: What’s the elevation of the equilibrium line?

Line 367-373: Remove ‘Ignorance of the principles on which they were compiled may lead to conclusions drawn as to the glacier recession rate being erroneous and, consequently, recession being overestimated for the years 1984–90, as the apparent status in 1984 would be contrary to reality. Specifically, the misapprehension lies in the fact that, a’. Instead, start the second sentence ‘Although the IGF PAN field campaign was conducted in the early 1980s, the maps published after the expedition were based on elevation data taken from aerial photos from 1961, upon which only glacier extents were updated (with a change in colour of contours). Crucially, contour lines were not updated in this 1984 edition, and continued to represent the elevations of 1961.’

Line 384-385: Remove last sentence

Tables

Table 1: Add a third column with references to the published works from these overflights (or surveys?). Could also add a map next to the table, showing the coverage for each of the flights?

Tables 2-5: Add references to IGF PAN and NPI data. In Table 2: Vestre Zdanovfjellet with a capital V

Table 7: It says land-terminating glaciers instead of calving glaciers in the caption

Figures

Figure 1: Why the red frame in (c)? Make (c) only cover the area of interest or make the text ‘Area of research’ in red and put inside red box. Figure could also benefit from having a map showing Svalbard’s location in the North Atlantic instead of panel (b)? Add all names mentioned in the text.

Figure 2: Label the maps (a), (b), and (c), and provide information on which sheet is which. Consider showing all ten map sheets, to help illustrate why these three are the most valuable (supported by an explanation in the text).  

Figure 3: Mark the extent of this figure in Figure 2 or present them (Figures 2 and 3) together.

Figures 5-8. Merge these figures into one (or at least Figures 5, 6 and 8), and increase the letter size. Add info on contour line interval.

Figure 7: Stupprygen -> Stupryggen. This typo reoccurs in several figures and in the text.

Figures 10-12: Merge into one figure, potentially also adding the plot from Figure 13.

Figure 9 and 14: Present these figures together, to make them easier to compare. Could even present them together with Figure 4.

References

Girod, L., Nielsen, N. I., Couderette, F., Nuth, C., and Kääb, A. (2018) Precise DEM extraction from Svalbard using 1936 high oblique imagery. Geoscientific Instrumentation, Methods and Data Systems 7(4), 277–288. https://doi.org/10.5194/gi-7-277-2018.

Holmlund, E. S. (2021) Aldegondabreen glacier change since 1910 from structure-from-motion photogrammetry of archived terrestrial and aerial photographs: utility of a historic archive to obtain century-scale Svalbard glacier mass losses. Journal of Glaciology 67, 107–116. https://doi.org/10.1017/jog.2020.89

Isaksen, K., Nordli, Ø., Førland, E. J., Łupikasza, E., Eastwood, S., & Niedźwiedź, T. (2016). Recent warming on Spitsbergen—Influence of atmospheric circulation and sea ice cover. Journal of Geophysical Research: Atmospheres, 121(20), 11-913. https://doi.org/10.1002/2016JD025606

Jacob, T., Wahr, J., Pfeffer, W. T., & Swenson, S. (2012). Recent contributions of glaciers and ice caps to sea level rise. Nature, 482(7386), 514-518. https://doi.org/10.1038/nature10847

Korsgaard, N. J., Nuth, C., Khan, S. A., Kjeldsen, K. K., Bjørk, A. A., Schomacker, A., & Kjær, K. H. (2016). Digital elevation model and orthophotographs of Greenland based on aerial photographs from 1978–1987. Scientific Data, 3(1), 1-15. https://doi.org/10.1038/sdata.2016.32

Mertes, J. R., Gulley, J. D., Benn, D. I., Thompson, S. S., & Nicholson, L. I. (2017). Using structure‐from‐motion to create glacier DEMs and orthoimagery from historical terrestrial and oblique aerial imagery. Earth Surface Processes and Landforms, 42(14), 2350-2364. https://doi.org/10.1002/esp.4188

Midgley, N. G., & Tonkin, T. N. (2017). Reconstruction of former glacier surface topography from archive oblique aerial images. Geomorphology, 282, 18-26. https://doi.org/10.1016/j.geomorph.2017.01.008

Nordli, Ø., Przybylak, R., Ogilvie, A. E., & Isaksen, K. (2014). Long-term temperature trends and variability on Spitsbergen: the extended Svalbard Airport temperature series, 1898–2012. Polar research, 33(1), 21349. https://doi.org/10.3402/polar.v33.21349

Nuth, C., Kohler, J., Aas, H. F., Brandt, O., & Hagen, J. O. (2007). Glacier geometry and elevation changes on Svalbard (1936–90): a baseline dataset. Annals of Glaciology, 46, 106-116. https://doi.org/10.3189/172756407782871440

Nuth, C., Kohler, J., König, M., Deschwanden, A. V., Hagen, J. O., Kääb, A., ... & Pettersson, R. (2013). Decadal changes from a multi-temporal glacier inventory of Svalbard. The Cryosphere, 7(5), 1603-1621. https://doi.org/10.5194/tc-7-1603-2013

Schuler, T. V., Kohler, J., Elagina, N., Hagen, J. O. M., Hodson, A. J., Jania, J. A., Käbb, A. M., Luks, Bartlomiej, Malecki, J., Moholdt, G., Pohjola, V. A., Sobota, I., & Van Pelt, W. J. (2020). Reconciling Svalbard glacier mass balance. Frontiers in Earth Science, 8, 156. https://doi.org/10.3389/feart.2020.00156