egusphere-2023-321

Coastal retreat rates of high-Arctic rock cliffs on Brøgger peninsula, Svalbard, accelerate during the past decade

General comments:

The manuscript deals with a relevant and up-to-date topic concerning coastal retreat rates of Arctic rock cliffs located on the Brøgger peninsula in Svalbard, Norway. In times of ongoing and projected future climate change more detailed scientific results on coastal erosion particularly in permafrost regions are indeed necessary. The main aim of the study is to detect long-term trends in differently exposed coastlines and linking these changes to available climate date. Results are achieved by using high-resolution orthoimagery combined with dGNSS measurements. In addition a short term dataset of rock surface temperatures was acquired and topography-based downscaling of atmospheric reanalysis data was utilized. While the presented results elucidate the contrasting trends of retreat rates on the northeast and southwest facing coastlines very nicely, the given explanations for the detected differences as well for the potential main climatic drivers are too general and leave a lot of room for improvement. A one-year dataset of rock surface temperatures is certainly quite short for a more comprehensive temperature analysis.     

The manuscript is well presented and has a logical structure. All tables and figures including the appendix are well prepared. While the method and result sections are quite detailed, the introduction and discussion sections need to be clearly revised and improved. In its current state, the manuscript has the character of a regional case study with limited novelty for a broader and international readership. I recommend to widen the author’s perspective from Svalbard also to other Arctic coastal areas worldwide and to include references of similar or comparable studies from these regions too. A short compilation of published coastal retreat rates in the Arctic would certainly be interesting. I have listed some international references at the end that the authors might find useful. The study site description needs to be revised significantly. For readers unfamiliar with Svalbard, the given study site description is not helpful at all. Please see my detailed comments below. The discussion section needs to be significantly revised as well. At the moment, I can’t see a wider geoscientific relevance of the presented research. In addition, by placing your results achieved in an international context, you will definitely reach a larger readership. Under 5.2 it might be worth to discuss actual and potential implications of an increased susceptibility of the coastline. For instance what does enhanced coastal erosion mean for the population and infrastructure of Svalbard? I think the manuscript could benefit from illuminating these points in a more detailed and broader perspective as well as to discuss possible implications of the achieved findings in more detail and in an international context.

Specific comments:

Under point 2:

Some more details on the climatic setting of your study site are needed here, e.g.:

- geographical coordinates of the study sites

- mean annual air temperature for the period of recorded data

- any information about the wind regime

-  general information on snow cover and more specific information on the mentioned reduced snow cover

- permafrost distribution

- important denudational processes

I think it is necessary to explain the permafrost situation of your study site in more detail here. For readers who are not familiar with the study site or the special conditions of Svalbard, it is not clear how the permafrost distribution looks like in Svalbard and where and how it is measured. What is the specific permafrost situation along the coastal rock cliffs?    

Concerning the actual cliff study sites:

- information of the entire actual cliff height (if possible differentiated in bedrock height and the height of the unconsolidated marine deposits on top)

- information on crack and fracture density if available

Under point 3.4:

Line 169: Why wasn’t it possible to place a temperature logger at the northeast facing coastline? Is the study site not accessible?

How do you make sure that the iButtons are always attached closely to the surface of the rockwall, so that they don’t measure the air temperature inside the crack? Did you experience any problems with moisture or ice inside the cracks?

Under point 3:

The “accuracy and error analysis” is clearly structured and reasonable explained. It is obvious that the orthoimages from 1970 are the most critical ones. However, the shown example of the digitized coastline from 1970 is not clear to me. Why is the digitized line so far away from the “coastline” and how did you manage to recognize the notches?

Figure 4:

This is a very nice figure but it is a bit confusing that the retreat rates for all three time intervals are shown on the same orthoimage. I can understand the decision in order to have a good visibility. However, you could maybe mention the actual date of the orthoimage for clarification.

Under point 4.2:

It would certainly be interesting if you could elaborate on possible reasons for the different detected retreat rates on the NE and SW facing side in more detail, as this is one of your main findings.  

Under point 4.3:

Please see comment under point 2.

Technical comments:

Line 230: from 1 September 2020 to 31 August 2021

Line 342: September 2020 to August 2021?

Fig. 4: “however” is redundant here

References:

Barnhart, K.R., Overeem, I., Anderson, R.S., 2014. The effect of changing sea ice on the physical vulnerability of Arctic coasts. The Cryosphere. 8, 1777-1799.

Lantuit, H., Overduin, P.P., Couture, N., Wetterich, S., Aré, F., Atkinson, D., Brown, J., Cherkashov, G., Drozdov, D., Forbes, D.L., 2012. The Arctic coastal dynamics database: a new classification scheme and statistics on Arctic permafrost coastlines. Estuar. Coasts. 35, 383–400.

Overduin, P.P., Strzelecki, M.D., Grigoriev, M.N., Couture, N, Lantuit, H., St-Hilaire-Gravel, D., Gunther, F., Wetterich, S., 2014. Coastal changes in the Arctic. In: Martini, I.P. and Wanless, H.R. (eds). Sedimentary Coastal Zones from High to Low Latitudes: Similarities and Differences, Geological Society, London, Special Publications, 388, 103–129.

Reimnitz, E., Maurer, D.K., 1979. Effects of storm surges on the Beaufort Sea coast, northern Alaska. Arctic, 324, 329–344.

General comments:

The manuscript focuses on cliff erosion along 5.5 km coastline in NW Svalbard. Four aerial surveys were used to derive decadal-scale cliff retreat rates. The study is valuable given limited research on Arctic rock coasts, in particular at the timescales exceeding a few years.

I think that the paper is well-written with very clear methods. I really appreciate error estimation. The results are concrete. The figures are of good quality. There is no redundant information.

I agree with some conserns by Referee #1 and will not re-list them here. Definitely the reference to beyond-Svalbard Acrtic cliff studies is missing, so is a more detailed study area description including cliff morphology (cliff height, typical slope of bedrock wall and overlaying soft sediment). I would expect better justification of using top of the overlaying material as the proxy for coastline, given extensive discussion on this topic in rock coast literature. In terms of limited analysis on environmental conditions influencing coastal erosion raised by Refree #1, I would say you can go either way - perform more in-depth analyses or consider it out of scope of the study (but then I would move the temperature data to appendix and take the interpretation out of conclusions). Indeed, as of now there is quite a mismatch between measuring erosion and environmental conditions.

Specific comments:

Line 19: what about rising SST?

Line 21: state timespan of these observations

Line 46: specify that it is the coastal cliff erosion (there are decadal-scale coastal erosion studies in Svalbard such as Zagorski et al., 2015)

Line 65: Are there any estimates (in days) of the snow cover season shortening?

Line 69: do they provide values for the increase?

Technical comments:

Line 26: replace ‘comparatively’ with ‘relatively’

Line 264: replace ‘1970-2021’ with ‘51 years’

Line 293: perhaps replace ‘release of large blocks’ with ‘landsliding’ because the former term means something different in rock coast literature

Fig. 1. Add an arrow pointing to the study area on the map of Svalbard.

Fig. 1. Perhaps add names of water bodies on panel a (and inset?).

Fig. 1. Something is wrong with the scale bar given your study area is 5.5 km

Review of "Coastal retreat rates of high-Arctic rock cliffs on Brøgger peninsula, Svalbard, accelerate during the past decade" by Aga et al. [Preprint egusphere-2023-321], submitted to ESurf

The novelty of this work lies in measuring coastal cliff top retreat on Svalbard over 50 years on a high spatial resolution. Overall, most interpretations and conclusions are supported by the evidence, the assumptions are valid, the methodology is sound, the evidence is adequate, and the conclusions logically follow the results. However, the authors do not present any data to build a relationship between the measured retreat rates and potential environmental driving factors that can explain the increase in cliff top retreat and they cannot fulfill their third objective. Therefore, the final conclusion of the manuscript (retreat rates increase with climate change) remains highly speculative (although likely) and the manuscript lacks important scientific progress. Currently, the manuscript requires major improvements to make a substantial progress for the research field.

Overall, I agree with the comments from RC1. Therefore, I will focus on some suggestions to improve the manuscript. Although it takes an effort to incorporate all the suggested changes, I highly encourage the authors to go the extra mile and improve this manuscript. I think, if revised properly, this could become an important contribution for the field of Arctic coastal dynamics. I have six general suggestions to improve this study:

First, the authors mention coastline retreat throughout the manuscript, but they measure cliff top retreat. I agree with the authors that the cliff top can be detected more easily from historical aerial images, but I don’t agree with the statement: ‘we are confident that the shift of the top of the cliff is representative of the coastal retreat’. The authors do not provide a reference or any data to corroborate this statement. In contrast, a recent study has found that the cliff top is eroding with a higher magnitude, but a lower frequency compared to the cliff toe that is eroding with a lower magnitude but a higher frequency (Swirad et al. 2022, https://doi.org/10.1016/j.geomorph.2022.108318). Additionally, if the cliff is not a plunging cliff, which remains unclear in the authors description of the study site, the beach fronting the cliff plays an important role in the coastline retreat as well. The easiest solution would be to write cliff top retreat instead of coastline retreat throughout the manuscript. However, the most interesting solution would be to establish a relationship between coastline retreat and cliff toe as well as cliff top retreat. The latter is directly leading to the next point.

Second, the authors investigate cliff top retreat in 2D. Compared to low lying coastal environments like deltas and spits, coastal cliffs are unique because of their abrupt change in elevation. To fully understand the dynamics of coastal cliffs, including cliff toe, cliff face and cliff top retreat, the investigation requires 3D data. Based on cross-cliff elevation profiles derived from 3D elevation data, a relationship between cliff morphometrics (e.g. beach width and slope, cliff top and toe elevation, and cliff face slope) and cliff retreat rates can be built. Depending on those relationships, environmental driving factors (overland flow, ground temperature, wave run-up, sea-ice content, sea-level change etc.) can be detected and changes over time can be discussed. 3D data (point clouds, DEMs etc.) of coastal cliffs in the Arctic is rare, but necessary to advance the understanding of coastal cliff dynamics in the Arctic. Without 3D change observations, this study lacks progress in the field as 2D retreat rates of cliffs are investigated already at other places on Svalbard which is discussed by the authors in chapter 5.1.

Third, the current discussion chapter 5.2 presents no new insights because the authors present no data to discuss the potential environmental driving factors. In the current version of the chapter, the authors describe the conceptual model explaining how those cliffs erode which is already understood. What the scientific community is lacking is data showing if the theoretical understanding is accurate and applicable. For example, the authors discuss potential differences in exposure to waves and storms of the northeast and southwest facing coastlines. The topographically downscaled ERA5 climate data could be used to create a wave rose of significant wave height. This would base the assumption in the manuscript on actual data and increase the overall strength of the manuscript, providing that the relationship between cliff retreat rate and environmental driving factors are understood. The same goes for other potential environmental driving factors, e.g. provide actual data on sea-ice changes in the study area. The available data from the downscaled ERA5 data is very interesting, but the authors do not present the data in the main results. The discussion of these results is very short.

Fourth, the only reason to include the rock surface temperature data in the manuscript is to directly couple it to cliff retreat rates. It would be super interesting to show a relationship between seasonal cliff retreat and rock surface temperature (e.g. cliff retreat is faster when rock surface temperature is higher). Without this data, I would delete chapter 4.3.

Fifth, in chapter 5.2 the authors discuss ‘Coastal retreat rates under a warming climate’. Please rephrase to ‘Coastal cliff retreat rates…’. It is generally understood that the warming climate is leading to the melting of the icecaps and glaciers which in turn will lead to rising sea-levels. However, Svalbard is experiencing postglacial isostatic rebound which might outpace the regional sea-level rise. More importantly, the melting of the Greenland Ice Sheet is leading to a gravitational sea-level lowering around Svalbard because of their proximity (< 2.200 km). Lastly, due to the melting of the Greenland Ice Sheet, the rotational axis of the earth will move towards the area of mass loss which will also lead to a regional decrease of sea-levels around Svalbard. For more information see Slangen et al. 2022, e.g. figure 1 (https://doi.org/10.1175/JCLI-D-17-0110.1). Please include these considerations in your discussion on coastal cliff retreat rates under a warming climate as they are interconnected. You are welcome to get some inspiration from my recent paper discussing the same dynamic on a coastal cliff in Greenland (https://doi.org/10.1029/2022JF007026).

Sixth, the manuscript lacks appropriate referencing of relevant literature and is presented as a regional case study with little interest for a broader, pan-Arctic audience. I would suggest putting a greater effort into discussing the broader applicability of the results for all of Svalbard and compare the findings with studies from similar Arctic coastlines (e.g. Greenland and Nunavut). Below some suggestions for literature to consider:

Barnhart, K. R., Overeem, I., & Anderson, R. S. (2014). The effect of changing sea ice on the physical vulnerability of Arctic coasts. The Cryosphere, 8(5), 1777–1799. https://doi.org/10.5194/tc-8-1777-2014

Boisson, A., Allard, M., & Sarrazin, D. (2020). Permafrost aggradation along the emerging eastern coast of Hudson Bay, Nunavik (northern Québec, Canada). Permafrost and Periglacial Processes, 31(1), 128–140. https://doi.org/10.1002/ppp.2033

Bourriquen, M., Baltzer, A., Mercier, D., Fournier, J., Pérez, L., Haquin, S., Bernard, E., & Jensen, M. (2016). Coastal evolution and sedimentary mobility of Brøgger Peninsula, northwest Spitsbergen. Polar Biology, 39(10), 1689-1698. https://doi.org/10.1007/s00300-016-1930-1

Bourriquen, M., Mercier, D., Baltzer, A., Fournier, J., Costa, S., & Roussel, E. (2018). Paraglacial coasts responses to glacier retreat and associated shifts in river floodplains over decadal timescales (1966-2016), Kongsfjorden, Svalbard. Land Degradation & Development, 29(11), 4173-4185. https://doi.org/10.1002/ldr.3149

Casas-Prat, M., & Wang, X. L. (2020a). Projections of extreme ocean waves in the Arctic and potential implications for coastal inundation and erosion. Journal of Geophysical Research: Oceans, 125(8), e2019JC015745. https://doi.org/10.1029/2019JC015745

Casas-Prat, M., & Wang, X. L. (2020b). Sea ice retreat contributes to projected increases in extreme Arctic Ocean surface waves. Geophysical Research Letters, 47(15), e2020GL088100. https://doi.org/10.1029/2020GL088100

Jones, B. M., Irrgang, A., Farquharson, L. M., Lantuit, H., Whalen, D., Ogorodov, S., et al. (2020). Coastal permafrost erosion. https://doi.org/10.25923/e47w-dw52

Kavan, J., & Strzelecki, M. C. (2023). Glacier decay boosts the formation of new Arctic coastal environments—Perspectives from Svalbard. Land Degradation & Development. https://doi.org/10.1002/ldr.4695

McCrystall, M. R., Stroeve, J., Serreze, M., Forbes, B. C., & Screen, J. A. (2021). New climate models reveal faster and larger increases in Arctic precipitation than previously projected. Nature Communications, 12(1), 6765. https://doi.org/10.1038/s41467-021-27031-y

Nielsen, D. M., Pieper, P., Barkhordarian, A., Overduin, P., Ilyina, T., Brovkin, V., et al. (2022). Increase in Arctic coastal erosion and its sensitivity to warming in the twenty-first century. Nature Climate Change, 12(3), 263–270. https://doi.org/10.1038/s41558-022-01281-0

Sessford, E. G., Baeverford, M. G., & Hormes, A. (2015). Terrestrial processes affecting unlithified coastal erosion disparities in central fjords of Svalbard. Polar Research, 34(1), 24122. https://doi.org/10.3402/polar.v34.24122

St-Hilaire-Gravel, D., Bell, T. J., & Forbes, D. L. (2010). Raised gravel beaches as proxy indicators of past sea-ice and wave conditions, Lowther Island, Canadian Arctic archipelago. Arctic, 63(2), 213–226. https://doi.org/10.14430/arctic976

St-Hilaire-Gravel, D., Forbes, D. L., & Bell, T. (2012). Multitemporal analysis of a gravel-dominated coastline in the central Canadian Arctic archipelago. Journal of Coastal Research, 280, 421–441. https://doi.org/10.2112/jcoastres-d-11-00020.1

Zagórski, P., Rodzik, J., Moskalik, M., Strzelecki, M. C., Lim, M., Błaszczyk, M., et al. (2015). Multidecadal (1960–2011) shoreline changes in Isbjørnhamna (Hornsund, Svalbard). Polish Polar Research, 36(4), 369–390. https://doi.org/10.1515/popore-2015-0019

Detailed comments

L9 Mention the cliff top retreat increase in percent between the three periods of investigation instead of the absolute values in the abstract.

L51 You would also detect long-term trends in coastal retreat if you would analyze the two areas together. The second half of the sentence ‘by performing separate analyses for the northeast and southwest facing coastline’ does not fit to the first half ‘to detect long-term trends in coastal retreat along the Brøgger peninsula’.

L55 It is essential for the readers understanding to mention that the study area is located in an uplifted beach ridge system and that the topmost layer of the cliffs consists of uplifted marine terraces. Please describe the Holocene landscape evolution history of the study and mention the current day uplift rates that are measured in Ny-Ålesund.

L72 Did you try to use the aerial images from 1936 for cliff top mapping? Geyman et al. 2022 (https://doi.org/10.1038/s41586-021-04314-4) provides high resolution orthophotos in the supplement.  

L210 please just mention the retreat rates in the text and not the colors you assigned to the groups in fig 4.

L222 The last sentence of the paragraph should be moved to the discussion.

L249 In my opinion it doesn’t make sense to compare your very local retreat rate to an average pan-Arctic retreat rate. I would delete the sentence.

L260 reference the studies you are referring to.

L264 true, but you are also only looking at one geographically small, geologically homogenous area, and not several study sites all across the archipelago. I would rather investigate a smaller area in 3D than a larger area in 2D.

L312 Did you not observe an icefoot in spring in front of the cliff and what effect does this have for the cliff erosion dynamics?

L315 name the figure or chapter in the appendix you are referring to.

L358 where are those sites? Please include coordinates and a map.

Figure 1 please indicate the photo positions of panel b and c in panel a.

Figure 4 the color scheme is not very intuitive. I would assume green is stable and yellow moderate retreat rates. For the highest rates of erosion, I would assume red colors.