the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Aug 2023
Lake Ice Break-Up in Greenland: Timing and Spatio-Temporal Variability
Abstract. Synthetic aperture radar (SAR) data from the Sentinel-1 (S1) mission with its high temporal and spatial resolution allows for an automated detection of lake ice break-up timings from surface backscatter differences across South (S), Southwest (SW) and Northwest (NW) Greenland (< 71° N latitude) during the period 2017 to 2021. Median break-up dates of the 563 studied lakes range between 8 June and 10 July, being earliest in 2019 and latest in 2018. There is a strong correlation between break-up date and elevation, while no relationship with latitude and lake area could be observed. Lake-specific median break-up timings for 2017–2021 increase (i.e., are later) by 3 days per 100 m elevation gain. When assuming an earlier break- up timing of 8 days which corresponds to the observed median variability of ± 8 days, the introduced excess energy due to a changing surface albedo from ice to water translates to melting 0.5 m thick ice at the melting point or heating up a water depth down to 35 m by 1 K across the entire surface area of each respective lake. Upscaling the results to 100486 lakes across the regions S, SW and NW which correspond to 64.5 % of all lakes or 62.1 % of the overall lake area in Greenland yields an estimate of 1.8 * 106 TJ additional energy input. This translates to melting 5.8 Gt ice at the melting point or warming 432.3 Gt water by 1 K.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(4200 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(4200 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1762', Anonymous Referee #1, 02 Sep 2023
This study uses Sentinel-1 (S1) combined with surface backscatter differences to obtain disintegration dates for 563 lake ice in Greenland between 2017-2021, obtaining a relationship between disintegration date and altitude. Estimates of the additional radiative input due to the advancement of the disintegration dates are presented, and the possible consequences of the excess energy are discussed. The results reveal the uniqueness of Greenland's geographic location and topographic features in response to climatic conditions, and enhance understanding of the extent to which altitude, latitude and lake size correlate with disintegration dates.
However, more detailed descriptions should be added in some areas, e.g. the introduction clarifies the importance of lake ice phenology studies but lacks a description of climate studies or ice condition studies in Greenland. In addition, why did the authors only analyse the date of disintegration and ignore the date of freezing, were they limited by the detection method? It would be more convincing to quantify the additional heat input in terms of spatial and temporal variations in the length of the complete freeze-up. For the lack of credibility of the authors' assumption of a pre-melt lake ice albedo αi = 0.9, which is perhaps only achievable with fresh snow; the effect of cloudiness on incident solar radiation should also be considered.
The research work is interesting and covers a large number of lakes in Greenland, and the authors are advised to improve the manuscript in two ways. Ⅰ, emphasise the climatic characteristics of Greenland and the progress of ice condition research, suddenly the significance of this study; Ⅱ, determine empirical values of regional lake ice albedo and cloudiness to ensure that the estimation of additional heat input is accurate. In addition, it would have been better to include freezing dates in the analyses, or to explain the reasons for not considering freezing date identification. Below are some specific comments aimed to provide guidance on the revision.
L9-10: “There is a strong correlation between break-up date and elevation, while no relationship with latitude and lake area could be observed” 'no relationship' is too absolute, suggest 'weakly relationship'.
L18: “Introduction” the authors describe the relationship between lake ice and climate and the lack of break-up time studies in Greenland, I would also like to know more about Greenland's climate and ice conditions, e.g. what is the rate of warming? Is the lake ice surface dominated by bare ice or snow cover? What are the special features compared to other regions?
L21: “The duration of lake ice controls......” 'Duration of lake ice' might make more sense than 'break-up time'.
L34-35: “The seasonal changes in solar radiation, however, are the main influence for the overall energy availability to form and decay lake ice cover” It is for this reason that solar radiation may have a greater effect on lake ice melt than air temperature. Does it make sense to study the use of "Calculating Cumulative Positive Degree Days"?
L43: “Wang et al., 2018” missing information in References
L63-65: “While still being dependent on field measurements for validation, lake ice studies from remote sensing are no longer dependent solely on a small number of ground observations but can produce results and extrapolate measurements across landscapes and regions” I didn't understand the sentence.
L91-92: “…below 2 days for most of Greenland” This is a great idea, has anyone else also achieved 2 days accuracy with the help of this approach?
L117-118: “…we acquired incoming shortwave radiation and air temperature at 2 m data as climatological daily mean values for the period 1991-2020 from RACMO2” The study used "daily mean values". The credibility of Cumulative Positive Degree Days can be enhanced if there are measured daily mean values between 2017-2022.
L132-133: “…when most of the lake surface is…” "Most" is not a quantitative description; 80% or 90% would be a better standard.
L150: “Figure 1 (a)” could the sudden rise in November 𝜎0 also determine the freezing of the lake?
L150: “Figure 1 (c)” What does the pentagram represent?
L161-162: “The progressing melt on the lake surface leading to a rougher, wetter surface explains the 𝜎0 recovery before the major backscatter decline in summer indicating lake ice break-up” rougher and wetter ice is unlikely to result in αi = 0.9.
L164-165: “This is due to the nature of break-up processes being more complex due to melting on top and bottom or varying acquisition conditions” could it be a secondary freeze due to lower air temperatures?
L179: “Calculating Cumulative Positive Degree Days” whether climate averages for the period 1991-2020 are representative of today's ice-season environmental conditions?
L214: “…of lake ice αi (0.9) …” the value of 0.9 is too high, please revise or give a basis.
L223-224: Does the authors take into account the effect of cloud cover, which attenuates incident shortwave radiation?
L280: “Figure 4 (a)” I want to know r = 0.76, then p-value = ?
L284: “…we there is no relationship between break-up timing and lake size” it is necessary to control for equal elevation and latitude, and to analyse only the size of the lake in order to consider that there is no relationship between disintegration and lake size
L290-291: “Figure 5b shows that lakes with similar cumulative PDDs experience a later lake ice break-up at higher elevation” are the air temperature data corresponding to lakes at different elevations at the same latitude the same? Because altitude must bring about differences in air temperature, defaulting to the same value will lead to bias.
L332: “…down to 35 m by 1 K across…” 35 m is the average depth of regional lakes?
L349: “(between 40°N and 82.5°N)” perhaps because Greenland's latitudinal span is not large enough to become only weakly correlated.
Citation: https://doi.org/10.5194/egusphere-2023-1762-RC1 -
AC1: 'Reply on RC1', Christoph Posch, 31 Oct 2023
Dear editor, dear referees,
we are very grateful for two very constructive reviews and the editorial advice and appreciate the valuable time put into this. We believe by incorporating the reviews we will manage to achieve a much more mature manuscript.
In the following we mark bold the comments given by the referees and give our answers and comments in italic. We hope that the responses will satisfy both referees and editor, and we are looking forward to a decision and to revising the manuscript.
Once again, many thanks for the valuable input and all the best,
Christoph, on behalf of the author team.
This study uses Sentinel-1 (S1) combined with surface backscatter differences to obtain disintegration dates for 563 lake ice in Greenland between 2017-2021, obtaining a relationship between disintegration date and altitude. Estimates of the additional radiative input due to the advancement of the disintegration dates are presented, and the possible consequences of the excess energy are discussed. The results reveal the uniqueness of Greenland's geographic location and topographic features in response to climatic conditions, and enhance understanding of the extent to which altitude, latitude and lake size correlate with disintegration dates.
However, more detailed descriptions should be added in some areas, e.g. the introduction clarifies the importance of lake ice phenology studies but lacks a description of climate studies or ice condition studies in Greenland. In addition, why did the authors only analyse the date of disintegration and ignore the date of freezing, were they limited by the detection method? It would be more convincing to quantify the additional heat input in terms of spatial and temporal variations in the length of the complete freeze-up. For the lack of credibility of the authors' assumption of a pre-melt lake ice albedo αi = 0.9, which is perhaps only achievable with fresh snow; the effect of cloudiness on incident solar radiation should also be considered.
The research work is interesting and covers a large number of lakes in Greenland, and the authors are advised to improve the manuscript in two ways. ①, emphasise the climatic characteristics of Greenland and the progress of ice condition research, suddenly the significance of this study; ②, determine empirical values of regional lake ice albedo and cloudiness to ensure that the estimation of additional heat input is accurate. In addition (③), it would have been better to include freezing dates in the analyses, or to explain the reasons for not considering freezing date identification. Below are some specific comments aimed to provide guidance on the revision.
We greatly appreciate the constructive remarks, and we will seriously consider the incorporation of the raised points in our revised manuscript. Ad ①, we will add a discussion on climatic characteristics and trends of Greenland focusing on coastal climates and their variability (i.e., Taurisano et al. 2004, Hanna et al. 2021) and contextualize them with our results. Ad ②, cloudiness is already accounted for in our RACMO2 data and we will better explain the variables used from RACMO2. While it is true that the bare lake ice albedo might be as low as 0.2-0.5 in its latest stage before break-up (Mullen and Warren 1988, Heron and Woo 1994, Henneman and Stefan 1999, Grenfell and Perovich 2004, Jakilla et al. 2009, Semmler et al. 2012, Svacina et al. 2014a, Svacina et al. 2014b, Leppäranta 2015, Zdorovennova et al. 2018, and Robinson et al. 2021), we acknowledge that the hypothesized 8-day-earlier break-up must not only take the latest stage of the ice cover into account but also surface albedos along the melt season (dry snow, wet snow, disintegrating ice). Since we see in our data that the transition from snow covered lake ice to open water happens within days or few weeks (also, Abermann et al. 2019) and the development of the short-time albedo is highly variable, we generalized the energy input changes from albedo values αi = 0.9 for snow covered lake ice to αw = 0.1 for open water. We will expand on that matter as well as discuss and highlight the transition from snow covered lake ice to open water. Ad ③, when looking at the entire dataset and the 563 lakes for which we detected break-up timings, we found very different modes of freeze-up evolution in the SAR data which makes automated detection very difficult at best. Mostly we see a gradual increase of SAR backscatter and no clear timing when the lake surface is frozen - this is probably due to the gradual increase and development of the ice cover (thickness) which cannot be accounted for by SAR. Therefore, we focus on the break-up timing, as opposed to a comprehensive characterization of lake ice duration, which of course would be very relevant, too. We argue, that regarding the impact of lake ice presence on surface energy balance, the break-up timing of lake ice is more relevant than the freeze-up since the freeze-up typically occurs in October or November, where incoming solar radiation is already much less than during the timing of break-up, which coincides with closer to maximum solar radiation input (see Fig. 2).
L9-10: “There is a strong correlation between break-up date and elevation, while no relationship with latitude and lake area could be observed” 'no relationship' is too absolute, suggest ‘weakly relationship'.
We will correct the wording and also add the magnitude of the Pearson correlation coefficients as a table to the Appendix to show he weak statistical relationship.
Those are concerning latitude,
Relationship between lake-specific 2017-2021 median break-up DOY and latitude: -0.18
Relationship between lake-specific 2017-2021 median break-up DOY and latitude
(lakes grouped into elevation bands of 100 m): -0.07 to 0.45
Relationship between lake-specific annual break-up DOY and latitude: -0.41 to 0.16
Relationship between lake-specific annual break-up DOY and latitude
(lakes grouped into elevation bands of 100 m): -0.41 to 0.58
And concerning lake size,
Relationship between lake-specific 2017-2021 median break-up DOY and lake size: 0.03
Relationship between lake-specific 2017-2021 median break-up DOY and lake size
(lakes grouped into elevation bands of 100 m): 0.05 to 0.53
Relationship between lake-specific annual break-up DOY and lake size: 0.00 to 0.05
Relationship between lake-specific annual break-up DOY and lake size
(lakes grouped into elevation bands of 100 m): -0.40 to 0.65L18: “Introduction” the authors describe the relationship between lake ice and climate and the lack of break-up time studies in Greenland, I would also like to know more about Greenland's climate and ice conditions, e.g. what is the rate of warming? Is the lake ice surface dominated by bare ice or snow cover? What are the special features compared to other regions?
We will add a discussion on climatic characteristics and trends of Greenland focusing on coastal climates and their variability (i.e., Taurisano et al. 2004, Hanna et al. 2021) and contextualize them with our results.
L21: “The duration of lake ice controls......” 'Duration of lake ice' might make more sense than 'break-up time'.
The SAR backscatter data does not allow for identifying the timing of freeze-up, this is why we could not focus on lake ice duration. In the revised manuscript we will make this case clear and give an example in the appendix on the problem.
L34-35: “The seasonal changes in solar radiation, however, are the main influence for the overall energy availability to form and decay lake ice cover” It is for this reason that solar radiation may have a greater effect on lake ice melt than air temperature. Does it make sense to study the use of "Calculating Cumulative Positive Degree Days"?
We added air temperature (cumulative PDD) to our analysis since we assumed its significance based on several other studies which showed air temperature strongly determining the break-up timing (or duration). Studies by Magnuson et al. (2000), Weyhenmeyer et al. (2004), Williams et al. (2004), Duguay et al. (2006), Korhonen (2006), Williams and Stefan (2006), Brown and Duguay (2010), Jeffries et al. (2012), and Imrit and Sharma (2021) show both linear and non-linear relations between lake ice break-up and air temperature, as well as February-March and April-May air temperatures being determining lake ice break-up factors. However, we did not aim to quantify statistical relationships between air temperature and break-up timing but rather discuss the complexity regarding factors such as latitude, elevation, radiation and PDDs. Such as shown in Figure 5 and discussed in L289-296, we highlight the differences and influences of radiation and air temperature: “… Comparing two lakes at different elevation with a similar break-up timing, we see that a higher elevated lake with lower cumulative PDD values needs a comparingly higher energy input (or less energy output) to accommodate for the same break-up timing as the lower elevated lake with comparingly higher cumulative PDDs. This is provided by a location at lower latitudes with comparingly more incoming shortwave radiation.”
L43: “Wang et al., 2018” missing information in References
We will add the reference, which is:
Wang, J., Duguay, C. R., Clausi, D. A., Pinard, V., and Howell, S. E. L.: Semi-Automated Classification of Lake Ice Cover Using Dual Polarization RADARSAT-2 Imagery, Remote Sens., 10(11), 1727, 1-27, https://doi.org/10.3390/rs1011172, 2018.L63-65: “While still being dependent on field measurements for validation, lake ice studies from remote sensing are no longer dependent solely on a small number of ground observations but can produce results and extrapolate measurements across landscapes and regions” I didn’t understand the sentence.
We will rephrase the sentence to: “Lake ice studies from remote sensing can produce results and extrapolate field measurements across large spatial scales as opposed to field studies based on a small number of ground observations. However, remotely sensed lake ice studies still depend on field measurements for validation.”
L91-92: “…below 2 days for most of Greenland” This is a great idea, has anyone else also achieved 2 days accuracy with the help of this approach?
Thanks for the kind words! We are not aware of any studies having achieved that high temporal resolution. The studies on lake phenology from SAR (or satellite imagery in general) we found (and cited) are mostly lake-specific case studies in Canada or Europe which have a lower temporal resolution. The high latitudes and thus converging orbits leading to a high imagery overlap allow for an exceptionally high temporal resolution in Greenland. To our knowledge, our study is the first study focusing on lake phenology in Greenland on a large scale and achieving that high temporal resolution.
L117-118: “…we acquired incoming shortwave radiation and air temperature at 2 m data as climatological daily mean values for the period 1991-2020 from RACMO2” The study used “daily mean values”. The credibility of Cumulative Positive Degree Days can be enhanced if there are measured daily mean values between 2017-2022.
Since we attempt to cover a large spatial scale with high spatial variability in atmospheric variables, a seamless product from a high latitude adapted regional climate model (RCM) is an advantage. RACMO2 is a high-resolution RCM that has been shown to capture spatial and temporal variability and absolute values well (Noël et al., 2019). The current operational version RACMO2.3p2 which we used for the radiation and air temperature climatologies is validated against 37 AWSs on the Greenland Ice Sheet. Biases in daily mean 2-m temperatures and downward shortwave radiation amount to 0.14 °C and 4.8 W m-2 (latter corresponding to a bias of 2.7 %), respectively (Noël et al., 2019). In the revised version we will add this to the data description make this also clearer by discussing the discrepancy of observations vs. seamless RCM output.
L132-133: “…when most of the lake surface is…” “Most” is not a quantitative description; 80% or 90% would be a better standard.
We will rephrase the paragraph highlighting the nature of our method (averaging the backscatter of the 20 % of the central lake surface area to assess whether there is ice cover or not). However, we cannot state a definite quantitative measure since we do not perform a pixel-based ice cover classification (which would be beyond the scope and computing power at a that large sample we are working with).
L150: “Figure 1 (a)” could the sudden rise in November 𝜎 also determine the freezing of the lake?
Yes, we also assume that this period represents the freezing of the exemplary lake. However, due to the great local and inter-annual variability we were not able to establish an automated detection algorithm for freeze-up. Freeze-up is usually represented with a gradual increase in backscatter (the example given is easier to interpret than many others). In the revised version we will highlight the complexity of freeze-up better in order to better establish why we focus on break-up.
L150: “Figure 1 I” What does the pentagram represent?
We will add the pentagram to the legend, which represents the location of the exemplary lake shown in Figure 1 (a) and Figure 1 (b).
L161-162: “The progressing melt on the lake surface leading to a rougher, wetter surface explains the 𝜎0 recovery before the major backscatter decline in summer indicating lake ice break-up” rougher and wetter ice is unlikely to result in α = 0.9.
While it is true that the bare lake ice albedo might be as low as 0.2-0.5 in its latest stage before break-up (Mullen and Warren 1988, Heron and Woo 1994, Henneman and Stefan 1999, Grenfell and Perovich 2004, Jakilla et al. 2009, Semmler et al. 2012, Svacina et al. 2014a, Svacina et al. 2014b, Leppäranta 2015, Zdorovennova et al. 2018, and Robinson et al. 2021), we acknowledge that the hypothesized 8-day-earlier break-up must not only take the latest stage of the ice cover into account but surface albedos along the melt season (dry snow, wet snow, …). Since we see in our data that the transition from snow covered lake ice to open water happens within days or few weeks (also, Abermann et al. 2019) and the short-time albedo development is highly variable, we generalized the energy input changes from albedo values αi = 0.9 for snow covered lake ice to αw = 0.1 for open water. We will expand on that matter as well as discuss and highlight the transition from snow covered lake ice to open water.
L164-165: “This is due to the nature of break-up processes being more complex due to melting on top and bottom or varying acquisition conditions” could it be a secondary freeze due to lower air temperatures?
Assessing optical satellite imagery, we are very confident that in this period no secondary freeze happened. We will add this concern to the discussion.
L179: “Calculating Cumulative Positive Degree Days” whether climate averages for the period 1991-2020 are representative of today’s ice-season environmental conditions?
We acknowledge that the PDD from 2017-2021 is going to be higher compared to the 1991-2020 period due to the fact that the region is warming quickly with respect to the 1991-2020 period (Hanna et al., 2021). We utilized the 1991-2020 climatologies to get a robust comparison to the median break-up timings between 2017 and 2021. We expected that a yearly comparison to cum. PDDs or to average cum. PDDs for the period 2017-2021 are not representative of a general lake-specific generalization due to the extreme years of 2018 and 2019 (Hanna et al. 2021).
L214: “…of lake ice α (0.9) …” the value of 0.9 is too high, please revise or give a basis.
We generalized the energy input changes from albedo values αi = 0.9 for snow covered lake ice to αw = 0.1 for open water. We will expand on that matter as well as discuss and highlight the transition from snow covered lake ice to open water in contrast to bare lake ice in the late melt season to open water.
L223-224: Does the authors take into account the effect of cloud cover, which attenuates incident shortwave radiation?
The RACMO2 model output that we were working with describes incoming shortwave radiation at the surface – the attenuation effects by the atmosphere and clouds are accounted for. We will include this information in the Data section.
L280: “Figure 4 (a)” I want to know r = 0.76, then p-value = ?
We will include the probability value as well as the standard error for all reported statistical relationships.
L284: “…we there is no relationship between break-up timing and lake size” it is necessary to control for equal elevation and latitude, and to analyse only the size of the lake in order to consider that there is no relationship between disintegration and lake size
We analyzed the lake size by elevation and latitude bands, however, in order to keep the manuscript and figures concise and readable, we did not include all details in the figures. We will add the requested information to the Appendix and expand on this in the main text.
L290-291: “Figure 5b shows that lakes with similar cumulative PDDs experience a later lake ice break-up at higher elevation” are the air temperature data corresponding to lakes at different elevations at the same latitude the same? Because altitude must bring about differences in air temperature, defaulting to the same value will lead to bias.
Thank you, good point, we did not make this clear enough initially. Indeed, we derive a lake-specific time series for atmospheric input (both temperature and radiation). That way we clearly account for spatial gradients given the high resolution of RACMO. We assume by clarifying this in the text we can convince on that matter.
L332: “…down to 35 m by 1 K across…” 35 m is the average depth of regional lakes?
We greatly appreciate highlighting the need for extended explanations. From Eq. (6) and Eq. (7) we calculated the volume of ice melt at the melting point and water temperature increase from the excess energy input for each lake. Since we found that the lake surface areas explain more than 99 % of the variability in excess energy (Figure D2), we divided the volumes by the surface areas to get an estimate of the “ice column (thickness)” melted or “water column (depth)” increased by 1 K across the entire surface (ignoring bathymetry) – which is on average 35 m. The 35 m do not refer to a specific lake depth but rather indicate the average depth of water warmed due to the excess energy. We will add this calculation to the Methods section and add descriptive statistics to the Appendix.
L349: “(between 40°N and 82.5°N)” perhaps because Greenland’s latitudinal span is not large enough to become only weakly correlated.
Indeed, an expansion would be desirable – however, we stress, that the study area covered spans over approximately 1200 km latitudinal distance or approximately 11 latitudinal degrees (approximately 60-71° N), covering a very heterogeneous climate which is attributed to rugged coastal areas, with complex topography and fjord systems, covering a wide range of local climates effects. We will expand on this in the revised version.
References:
Abermann, J., Eckerstorfer, M., Malnes, E., and Hansen, B. U.: A large wet snow avalanche cycle in West Greenland quantified using remote sensing and in situ observations. Natural Hazards, 97, 517–534, https://doi.org/10.1007/s11069-019-03655-8, 2019.
Grenfell, T., C. and Perovich D., K.: Seasonal and spatial evolution of albedo in a snow-ice-land-ocean environment, Journal of Geophysical Research, 109, C1, 1-15, https://doi.org/10.1029/2003JC001866, 2004.
Henneman, H. E. and Stefan H. G.: Albedo models for snow and ice on a freshwater lake, Cold Regions Science and Technology, 29(1), 31-48, https://doi.org/10.1016/s0165-232x(99)00002-6, 1999.
Heron, R. and Woo, M.: Decay of a High Arctic lake-ice cover: observations and modelling, Journal of Glaciology, 40, 135, 1-10, https://doi.org/10.3189/S0022143000007371, 1994.
Imrit, M. A., Sharma, S.: Climate Change is Contributing to Faster Rates of Lake Ice Loss in Lakes Around the Northern Hemisphere, Journal of Geophysical Research: Biogeosciences, 126, e2020JG006134, https://doi.org/10.1029/2020JG006134, 2021.
Jakilla, J., Leppäranta, M., Kawamura, T., Shirasawa, K., and Salonen, K.: Radiation transfer and heat budget during the ice season in Lake Pääjärvi, Finland, Aquat. Ecol., 43(3), 681–692, https://doi.org/10.1007/s10452-009-9275-2, 2009.
Korhonen, J.: Long-term changes in lake ice cover in Finland, Nordic Hydrology, 37(4-5), 347-363, https://doi.org/10.2166/nh.2006.019, 2006.
Leppärantam, M. (Eds.): Freezing of Lakes and the Evolution of their Ice Cover, Springer-Verlag, Berlin Heidelberg, Germany, 309 pp., ISBN 978-3-642-29080-0, 2015.
Mullen P. C. and Warren S. G.: Theory of the Optical Properties of Lake Ice, Journal of Geophysical Research, 93, D7, 8403-8414, https://doi.org/10.1029/jd093id07p08403, 1988.
Noël, B., Van De Berg, W., J., Lhermitte, S., and Van Den Broeke, M.: Rapid ablation zone expansion amplifies north Greenland mass loss, Science Advances, 5, 9, 1-9, https://doi.org/10.1126/sciadv.aaw0123, 2019.
Robinson, A. L., Ariano, S., S., and Brown, L. C.: The Influence of Snow and Ice Albedo towards Improved Lake Ice Simulations, Hydrology 8(1), 11, 1-21, https://doi.org/10.3390/hydrology8010011, 2021.
Semmler, T., Cheng, B., Yang, Y., and Rontu, L.: Snow and ice on Bear Lake (Alaska) – sensitivity experiments with two lake ice models, Tellus A: Dynamic Meteorology and Oceanography, 64(1), 1-14, https://doi.org/10.3402/tellusa.v64i0.17339, 2012.
Svacina, N. A., Duguay, C. R., and Brown, L. C.: Modelled and satellite-derived surface albedo of lake ice - Part I: evaluation of the albedo parameterization scheme of the Canadian Lake Ice Model, Hydrological Processes, 28(16), 4550–4561, https://doi.org/10.1002/hyp.10253, 2014a.
Svacina, N. A., Duguay, C. R., and Brown, L. C.: Modelled and satellite-derived surface albedo of lake ice - Part II: evaluation of MODIS albedo products, Hydrological Processes, 28(16), 4562–4572, https://doi.org/10.1002/hyp.10257, 2014b.
Taurisano, A., Bøggild, C. E., and Karlsen H. G.: A century of climate variability and climate gradients from coast to ice sheet in West Greenland, Geogr. Ann., 86 A (2), 217–224, https://doi.org/10.1111/j.0435-3676.2004.00226.x, 2004.
Wang, J., Duguay, C. R., Clausi, D. A., Pinard, V., and Howell, S. E. L.: Semi-Automated Classification of Lake Ice Cover Using Dual Polarization RADARSAT-2 Imagery, Remote Sens., 10(11), 1727, 1-27, https://doi.org/10.3390/rs1011172, 2018.
Zdorovennova, G., Palshin, N., Efremova, T., Zdorovennov, R., Gavrilenko, G., Volkov, S., Bodanov, S., and Terzhevik, A.: Albedo of a Small Ice-Covered Boreal Lake: Daily, Meso-Scale and Interannual Variability on the Background of Regional Climate, Geosciences, 8(6), 206, 1-17, https://doi.org/10.3390/geosciences8060206, 2018.
Citation: https://doi.org/10.5194/egusphere-2023-1762-AC1
-
AC1: 'Reply on RC1', Christoph Posch, 31 Oct 2023
-
RC2: 'Comment on egusphere-2023-1762', Anonymous Referee #2, 25 Sep 2023
General comment:
I enjoyed reading the paper. The text is clear. The quality of the plots are very good. I think it is a nice paper and contributes to the field.I have one general (major) comment:
I think the discussion is somewhat lacking in putting the work in the broader context. It is not well motivated why this study matters from a global perspective and what it adds to the general understanding of the climate change impacts. More specifically:- US is a wide continent covering various Köppen-Geiger climatic zone but Greenland is specifically located in one climatic zone and is not spanning across multiple latitudes with different climatic features. Therefore, I am not too surprised that for US there was a strong spatial correlation while such correlation could not be observed in Greenland. Please consider adding more to this discussion and make the comparison a bit stronger.
- Put the whole study in the broader context: Maybe mention implication for hydropower. You can use snow-dominated locations for the sake of comparison. I tried to find some examples and the discussion section in paper https://www.sciencedirect.com/science/article/pii/S0022169423007497 and https://hess.copernicus.org/articles/24/3815/2020/ might work. Please try to find other papers to add to this point.
- Compare the finding with other boreal countries: Note https://tc.copernicus.org/articles/16/2493/2022/tc-16-2493-2022-discussion.html where they studied ice break-up patterns in Sweden. Please try to find other papers to add to this point.
- Mention what is expected to happen to Greenland under climate change. You can use latest IPCC report for this.
Specific comments:
- L280: please remove we.
- Please report the significance of the correlations whenever you report the correlation strength.
- The paper lacks climatic description of Greenland.
Citation: https://doi.org/10.5194/egusphere-2023-1762-RC2 -
AC2: 'Reply on RC2', Christoph Posch, 31 Oct 2023
Dear editor, dear referees,
we are very grateful for two very constructive reviews and the editorial advice and appreciate the valuable time put into this. We believe by incorporating the reviews we will manage to achieve a much more mature manuscript.
In the following we mark bold the comments given by the referees and give our answers and comments in italic. We hope that the responses will satisfy both referees and editor, and we are looking forward to a decision and to revising the manuscript.
Once again, many thanks for the valuable input and all the best,
Christoph, on behalf of the author team.
General comment:
I enjoyed reading the paper. The text is clear. The quality of the plots are very good. I think it is a nice paper and contributes to the field.
Thank you and we greatly appreciate the constructive remarks. We will seriously consider the incorporation of the raised points them in our revised manuscript.
I have one general (major) comment:
I think the discussion is somewhat lacking in putting the work in the broader context. It is not well motivated why this study matters from a global perspective and what it adds to the general understanding of the climate change impacts. More specifically:
- US is a wide continent covering various Köppen-Geiger climatic zone but Greenland is specifically located in one climatic zone and is not spanning across multiple latitudes with different climatic features. Therefore, I am not too surprised that for US there was a strong spatial correlation while such correlation could not be observed in Greenland. Please consider adding more to this discussion and make the comparison a bit stronger.
- Thank you for pointing out that there is a need for more explanation of the comparison and to make it stronger. We will add a discussion on climatic characteristics and trends of Greenland focusing on coastal climates and their variability (i.e., Taurisano et al. 2004, Hanna et al. 2021) and contextualize them with our results. The spatial climate variability of Greenland is well established and while it is true that it lies within one Köppen-Geiger climatic zone, we stress its heterogeneity. This is true for meteorological input (Hanna et al. 2020, Box et al. 2023) as well as local surface mass balance variability (Mankoff et al. 2021, Slater et al. 2021). Our study in fact shows regional differences of the response of lake ice break-up. We will elaborate on that in the discussion and stress the difference of the spatial gradients between the US study and Greenland.
- Put the whole study in the broader context: Maybe mention implication for hydropower. You can use snow-dominated locations for the sake of comparison. I tried to find some examples and the discussion section in paper https://www.sciencedirect.com/science/article/pii/S0022169423007497 and https://hess.copernicus.org/articles/24/3815/2020/ might work. Please try to find other papers to add to this point.
- This is a good point, thank you for the literature suggestions. Indeed, the timing of lake break-up particularly impacts challenges regarding hydropower and we will embed this larger framing in the discussion, referring to further studies such as Prowse et al. (2011), Gebre et al. (2013) and Cherry et al. (2017).
- Compare the finding with other boreal countries: Note https://tc.copernicus.org/articles/16/2493/2022/tc-16-2493-2022-discussion.html where they studied ice break-up patterns in Sweden. Please try to find other papers to add to this point.
- We will discuss and compare our findings with other high latitude and boreal countries. Thank you for the literature suggestion, we will refer to further studies such as Korhonen (2006), L’Abée-Lund et al. (2021), Zhang et al. (2021) and Cai et al. (2022).
- Mention what is expected to happen to Greenland under climate change. You can use latest IPCC report for this.
- We will add this to our discussion and draw an outlook on expected future changes referring to the projections of the IPCC report. Furthermore, we will discuss them in context with studies on lake ice phenology projections such as Imrit and Sharma (2021), and Huang et al. (2022).
Specific comments:
- L280: please remove we.
- We will remove “we”.
- Please report the significance of the correlations whenever you report the correlation strength.
- We will include all probability values and standard errors of our reported statistical relationships.
- The paper lacks climatic description of Greenland.
- We will add a discussion on climatic characteristics and trends of Greenland focusing on coastal climates and their variability (i.e., Taurisano et al. 2004, Hanna et al. 2021) and contextualize them with our results.
References:
Box, J. E., Nielsen, K. P., Yang, X., Niwano, M., Wehrlé, A., van As, D., Fettweis, X., Køltzow, M. A. Ø., Palmason, B., Fausto, R. S., van den Broeke, M. R., Huai, B., Ahlstrøm, A. P., Langley, K., Dachauer, A., and Noël, B.: Greenland ice sheet rainfall climatology, extremes and atmospheric river rapids, Meteorological Applications, 30, 4, 1-24, https://doi.org/10.1002/met.2134, 2023.
Cai, Y., Duguay, C. R., Ke, C.-Q.: A 41-year (1979–2019) passive-microwave-derived lake ice phenology data record of the Northern Hemisphere, Earth Syst. Sci. Data, 14, 3329–3347, https://doi.org/10.5194/essd-14-3329-2022, 2022.
Cherry, J. E., Knapp, C., Trainor, S., Ray, A. J., Tedesche, M., Walker, S.: Planning for climate change impacts on hydropower in the Far North, Hydrol. Earth Syst. Sci., 21, 133–151, https:// doi:10.5194/hess-21-133-2017, 2017.
Gebre, S., Alfredsen, K., Lia, L., Stickler, M., Tesaker, E.: Review of Ice Effects on Hydropower Systems, Journal of Cold Regions Engineering, 27(4), 196–222, https://doi.org/10.1061/(ASCE)CR.1943-5495.0000059, 2013.
Hanna, E., Capellen, J., Fettweis, X., Mernild, S. H., Mote, T. L., Mottram, R., Steffen, K., Ballinger, T. J., and Hall, R. J.: Greenland surface air temperature changes from 1981 to 2019 and implications for ice-sheet melt and mass-balance change, Int J. Climatol., 41, 1336–1352, https://doi.org/10.1002/joc.6771, 2021.
Huang, L., Timmermann, A., Lee, S.-S., Rodgers, K. B., Yamaguchi, R., and Chung, E.-S.: Emerging unprecedented lake ice loss in climate change projections, Nature Communications, 13(5798), 1-12, https://doi.org/10.1038/s41467-022-33495-3, 2022.
Imrit, M. A., Sharma, S.: Climate Change is Contributing to Faster Rates of Lake Ice Loss in Lakes Around the Northern Hemisphere, Journal of Geophysical Research: Biogeosciences, 126, e2020JG006134, https://doi.org/10.1029/2020JG006134, 2021.
Korhonen, J.: Long-term changes in lake ice cover in Finland, Nordic Hydrology, 37(4-5), 347-363, https://doi.org/10.2166/nh.2006.019, 2006.
L’Abée-Lund, J. H., Vøllestad, L. A., Brittain, J. E, Kvambekk, Å. S., and Solvang, T.: Geographic variation and temporal trends in ice phenology in Norwegian lakes during the period 1890–2020, The Cryosphere, 15, 2333–2356, https://doi.org/10.5194/tc-15-2333-2021, 2021.
Mankoff, K. D., Fettweis, X., Langen, P. L., Stendel, M., Kjeldsen, K. K., Karlsson, N. B., Noël, B., van den Broeke, M. R., Solgaard, A., Colgan, W., Box, J. E., Simonsen, S. B., King, M. D., Ahlstrøm, A. P., Andersen, S. B., and Fausto, R. S.: Greenland ice sheet mass balance from 1840 through next week, Earth Syst. Sci. Data, 13, 5001–5025, https://doi.org/10.5194/essd-13-5001-2021, 2021.
Prowse, T., Alfredsen, K., Beltaos, S., Bonsal, B. R., Bowden, W. B., Duguay, C. R., Korhola, A., McNamara, J., Vincent, W. F., Vuglinsky, V., Walter Anthony, K. M., and Weyhenmeyer, G. A.: Effects of Changes in Arctic Lake and River Ice, AMBIO 2011, 40, 63–74, https://doi.org/10.1007/s13280-011-0217-6, 2011.
Slater, T., Shepherd, A., McMillan, M., Leeson, A., Gilbert, L., Muir, A., Munneke, P. K., Noël, B., Fettweis, X., van den Broeke, M., and Briggs, K.: Increased variability in Greenland Ice Sheet runoff from satellite observations, Nature Communications, 12, 6069, 1-9, https://doi.org/10.1038/s41467-021-26229-4, 2021.
Taurisano, A., Bøggild, C. E., and Karlsen H. G.: A century of climate variability and climate gradients from coast to ice sheet in West Greenland, Geogr. Ann., 86 A (2), 217–224, https://doi.org/10.1111/j.0435-3676.2004.00226.x, 2004.
Zhang, S., Pavelsky, T. M., Arp, C. D., and Yang X.: Remote sensing of lake ice phenology in Alaska, Environ. Res. Lett. 16, 1-12, https://doi.org/10.1088/1748-9326/abf965, 2021.
Citation: https://doi.org/10.5194/egusphere-2023-1762-AC2
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1762', Anonymous Referee #1, 02 Sep 2023
This study uses Sentinel-1 (S1) combined with surface backscatter differences to obtain disintegration dates for 563 lake ice in Greenland between 2017-2021, obtaining a relationship between disintegration date and altitude. Estimates of the additional radiative input due to the advancement of the disintegration dates are presented, and the possible consequences of the excess energy are discussed. The results reveal the uniqueness of Greenland's geographic location and topographic features in response to climatic conditions, and enhance understanding of the extent to which altitude, latitude and lake size correlate with disintegration dates.
However, more detailed descriptions should be added in some areas, e.g. the introduction clarifies the importance of lake ice phenology studies but lacks a description of climate studies or ice condition studies in Greenland. In addition, why did the authors only analyse the date of disintegration and ignore the date of freezing, were they limited by the detection method? It would be more convincing to quantify the additional heat input in terms of spatial and temporal variations in the length of the complete freeze-up. For the lack of credibility of the authors' assumption of a pre-melt lake ice albedo αi = 0.9, which is perhaps only achievable with fresh snow; the effect of cloudiness on incident solar radiation should also be considered.
The research work is interesting and covers a large number of lakes in Greenland, and the authors are advised to improve the manuscript in two ways. Ⅰ, emphasise the climatic characteristics of Greenland and the progress of ice condition research, suddenly the significance of this study; Ⅱ, determine empirical values of regional lake ice albedo and cloudiness to ensure that the estimation of additional heat input is accurate. In addition, it would have been better to include freezing dates in the analyses, or to explain the reasons for not considering freezing date identification. Below are some specific comments aimed to provide guidance on the revision.
L9-10: “There is a strong correlation between break-up date and elevation, while no relationship with latitude and lake area could be observed” 'no relationship' is too absolute, suggest 'weakly relationship'.
L18: “Introduction” the authors describe the relationship between lake ice and climate and the lack of break-up time studies in Greenland, I would also like to know more about Greenland's climate and ice conditions, e.g. what is the rate of warming? Is the lake ice surface dominated by bare ice or snow cover? What are the special features compared to other regions?
L21: “The duration of lake ice controls......” 'Duration of lake ice' might make more sense than 'break-up time'.
L34-35: “The seasonal changes in solar radiation, however, are the main influence for the overall energy availability to form and decay lake ice cover” It is for this reason that solar radiation may have a greater effect on lake ice melt than air temperature. Does it make sense to study the use of "Calculating Cumulative Positive Degree Days"?
L43: “Wang et al., 2018” missing information in References
L63-65: “While still being dependent on field measurements for validation, lake ice studies from remote sensing are no longer dependent solely on a small number of ground observations but can produce results and extrapolate measurements across landscapes and regions” I didn't understand the sentence.
L91-92: “…below 2 days for most of Greenland” This is a great idea, has anyone else also achieved 2 days accuracy with the help of this approach?
L117-118: “…we acquired incoming shortwave radiation and air temperature at 2 m data as climatological daily mean values for the period 1991-2020 from RACMO2” The study used "daily mean values". The credibility of Cumulative Positive Degree Days can be enhanced if there are measured daily mean values between 2017-2022.
L132-133: “…when most of the lake surface is…” "Most" is not a quantitative description; 80% or 90% would be a better standard.
L150: “Figure 1 (a)” could the sudden rise in November 𝜎0 also determine the freezing of the lake?
L150: “Figure 1 (c)” What does the pentagram represent?
L161-162: “The progressing melt on the lake surface leading to a rougher, wetter surface explains the 𝜎0 recovery before the major backscatter decline in summer indicating lake ice break-up” rougher and wetter ice is unlikely to result in αi = 0.9.
L164-165: “This is due to the nature of break-up processes being more complex due to melting on top and bottom or varying acquisition conditions” could it be a secondary freeze due to lower air temperatures?
L179: “Calculating Cumulative Positive Degree Days” whether climate averages for the period 1991-2020 are representative of today's ice-season environmental conditions?
L214: “…of lake ice αi (0.9) …” the value of 0.9 is too high, please revise or give a basis.
L223-224: Does the authors take into account the effect of cloud cover, which attenuates incident shortwave radiation?
L280: “Figure 4 (a)” I want to know r = 0.76, then p-value = ?
L284: “…we there is no relationship between break-up timing and lake size” it is necessary to control for equal elevation and latitude, and to analyse only the size of the lake in order to consider that there is no relationship between disintegration and lake size
L290-291: “Figure 5b shows that lakes with similar cumulative PDDs experience a later lake ice break-up at higher elevation” are the air temperature data corresponding to lakes at different elevations at the same latitude the same? Because altitude must bring about differences in air temperature, defaulting to the same value will lead to bias.
L332: “…down to 35 m by 1 K across…” 35 m is the average depth of regional lakes?
L349: “(between 40°N and 82.5°N)” perhaps because Greenland's latitudinal span is not large enough to become only weakly correlated.
Citation: https://doi.org/10.5194/egusphere-2023-1762-RC1 -
AC1: 'Reply on RC1', Christoph Posch, 31 Oct 2023
Dear editor, dear referees,
we are very grateful for two very constructive reviews and the editorial advice and appreciate the valuable time put into this. We believe by incorporating the reviews we will manage to achieve a much more mature manuscript.
In the following we mark bold the comments given by the referees and give our answers and comments in italic. We hope that the responses will satisfy both referees and editor, and we are looking forward to a decision and to revising the manuscript.
Once again, many thanks for the valuable input and all the best,
Christoph, on behalf of the author team.
This study uses Sentinel-1 (S1) combined with surface backscatter differences to obtain disintegration dates for 563 lake ice in Greenland between 2017-2021, obtaining a relationship between disintegration date and altitude. Estimates of the additional radiative input due to the advancement of the disintegration dates are presented, and the possible consequences of the excess energy are discussed. The results reveal the uniqueness of Greenland's geographic location and topographic features in response to climatic conditions, and enhance understanding of the extent to which altitude, latitude and lake size correlate with disintegration dates.
However, more detailed descriptions should be added in some areas, e.g. the introduction clarifies the importance of lake ice phenology studies but lacks a description of climate studies or ice condition studies in Greenland. In addition, why did the authors only analyse the date of disintegration and ignore the date of freezing, were they limited by the detection method? It would be more convincing to quantify the additional heat input in terms of spatial and temporal variations in the length of the complete freeze-up. For the lack of credibility of the authors' assumption of a pre-melt lake ice albedo αi = 0.9, which is perhaps only achievable with fresh snow; the effect of cloudiness on incident solar radiation should also be considered.
The research work is interesting and covers a large number of lakes in Greenland, and the authors are advised to improve the manuscript in two ways. ①, emphasise the climatic characteristics of Greenland and the progress of ice condition research, suddenly the significance of this study; ②, determine empirical values of regional lake ice albedo and cloudiness to ensure that the estimation of additional heat input is accurate. In addition (③), it would have been better to include freezing dates in the analyses, or to explain the reasons for not considering freezing date identification. Below are some specific comments aimed to provide guidance on the revision.
We greatly appreciate the constructive remarks, and we will seriously consider the incorporation of the raised points in our revised manuscript. Ad ①, we will add a discussion on climatic characteristics and trends of Greenland focusing on coastal climates and their variability (i.e., Taurisano et al. 2004, Hanna et al. 2021) and contextualize them with our results. Ad ②, cloudiness is already accounted for in our RACMO2 data and we will better explain the variables used from RACMO2. While it is true that the bare lake ice albedo might be as low as 0.2-0.5 in its latest stage before break-up (Mullen and Warren 1988, Heron and Woo 1994, Henneman and Stefan 1999, Grenfell and Perovich 2004, Jakilla et al. 2009, Semmler et al. 2012, Svacina et al. 2014a, Svacina et al. 2014b, Leppäranta 2015, Zdorovennova et al. 2018, and Robinson et al. 2021), we acknowledge that the hypothesized 8-day-earlier break-up must not only take the latest stage of the ice cover into account but also surface albedos along the melt season (dry snow, wet snow, disintegrating ice). Since we see in our data that the transition from snow covered lake ice to open water happens within days or few weeks (also, Abermann et al. 2019) and the development of the short-time albedo is highly variable, we generalized the energy input changes from albedo values αi = 0.9 for snow covered lake ice to αw = 0.1 for open water. We will expand on that matter as well as discuss and highlight the transition from snow covered lake ice to open water. Ad ③, when looking at the entire dataset and the 563 lakes for which we detected break-up timings, we found very different modes of freeze-up evolution in the SAR data which makes automated detection very difficult at best. Mostly we see a gradual increase of SAR backscatter and no clear timing when the lake surface is frozen - this is probably due to the gradual increase and development of the ice cover (thickness) which cannot be accounted for by SAR. Therefore, we focus on the break-up timing, as opposed to a comprehensive characterization of lake ice duration, which of course would be very relevant, too. We argue, that regarding the impact of lake ice presence on surface energy balance, the break-up timing of lake ice is more relevant than the freeze-up since the freeze-up typically occurs in October or November, where incoming solar radiation is already much less than during the timing of break-up, which coincides with closer to maximum solar radiation input (see Fig. 2).
L9-10: “There is a strong correlation between break-up date and elevation, while no relationship with latitude and lake area could be observed” 'no relationship' is too absolute, suggest ‘weakly relationship'.
We will correct the wording and also add the magnitude of the Pearson correlation coefficients as a table to the Appendix to show he weak statistical relationship.
Those are concerning latitude,
Relationship between lake-specific 2017-2021 median break-up DOY and latitude: -0.18
Relationship between lake-specific 2017-2021 median break-up DOY and latitude
(lakes grouped into elevation bands of 100 m): -0.07 to 0.45
Relationship between lake-specific annual break-up DOY and latitude: -0.41 to 0.16
Relationship between lake-specific annual break-up DOY and latitude
(lakes grouped into elevation bands of 100 m): -0.41 to 0.58
And concerning lake size,
Relationship between lake-specific 2017-2021 median break-up DOY and lake size: 0.03
Relationship between lake-specific 2017-2021 median break-up DOY and lake size
(lakes grouped into elevation bands of 100 m): 0.05 to 0.53
Relationship between lake-specific annual break-up DOY and lake size: 0.00 to 0.05
Relationship between lake-specific annual break-up DOY and lake size
(lakes grouped into elevation bands of 100 m): -0.40 to 0.65L18: “Introduction” the authors describe the relationship between lake ice and climate and the lack of break-up time studies in Greenland, I would also like to know more about Greenland's climate and ice conditions, e.g. what is the rate of warming? Is the lake ice surface dominated by bare ice or snow cover? What are the special features compared to other regions?
We will add a discussion on climatic characteristics and trends of Greenland focusing on coastal climates and their variability (i.e., Taurisano et al. 2004, Hanna et al. 2021) and contextualize them with our results.
L21: “The duration of lake ice controls......” 'Duration of lake ice' might make more sense than 'break-up time'.
The SAR backscatter data does not allow for identifying the timing of freeze-up, this is why we could not focus on lake ice duration. In the revised manuscript we will make this case clear and give an example in the appendix on the problem.
L34-35: “The seasonal changes in solar radiation, however, are the main influence for the overall energy availability to form and decay lake ice cover” It is for this reason that solar radiation may have a greater effect on lake ice melt than air temperature. Does it make sense to study the use of "Calculating Cumulative Positive Degree Days"?
We added air temperature (cumulative PDD) to our analysis since we assumed its significance based on several other studies which showed air temperature strongly determining the break-up timing (or duration). Studies by Magnuson et al. (2000), Weyhenmeyer et al. (2004), Williams et al. (2004), Duguay et al. (2006), Korhonen (2006), Williams and Stefan (2006), Brown and Duguay (2010), Jeffries et al. (2012), and Imrit and Sharma (2021) show both linear and non-linear relations between lake ice break-up and air temperature, as well as February-March and April-May air temperatures being determining lake ice break-up factors. However, we did not aim to quantify statistical relationships between air temperature and break-up timing but rather discuss the complexity regarding factors such as latitude, elevation, radiation and PDDs. Such as shown in Figure 5 and discussed in L289-296, we highlight the differences and influences of radiation and air temperature: “… Comparing two lakes at different elevation with a similar break-up timing, we see that a higher elevated lake with lower cumulative PDD values needs a comparingly higher energy input (or less energy output) to accommodate for the same break-up timing as the lower elevated lake with comparingly higher cumulative PDDs. This is provided by a location at lower latitudes with comparingly more incoming shortwave radiation.”
L43: “Wang et al., 2018” missing information in References
We will add the reference, which is:
Wang, J., Duguay, C. R., Clausi, D. A., Pinard, V., and Howell, S. E. L.: Semi-Automated Classification of Lake Ice Cover Using Dual Polarization RADARSAT-2 Imagery, Remote Sens., 10(11), 1727, 1-27, https://doi.org/10.3390/rs1011172, 2018.L63-65: “While still being dependent on field measurements for validation, lake ice studies from remote sensing are no longer dependent solely on a small number of ground observations but can produce results and extrapolate measurements across landscapes and regions” I didn’t understand the sentence.
We will rephrase the sentence to: “Lake ice studies from remote sensing can produce results and extrapolate field measurements across large spatial scales as opposed to field studies based on a small number of ground observations. However, remotely sensed lake ice studies still depend on field measurements for validation.”
L91-92: “…below 2 days for most of Greenland” This is a great idea, has anyone else also achieved 2 days accuracy with the help of this approach?
Thanks for the kind words! We are not aware of any studies having achieved that high temporal resolution. The studies on lake phenology from SAR (or satellite imagery in general) we found (and cited) are mostly lake-specific case studies in Canada or Europe which have a lower temporal resolution. The high latitudes and thus converging orbits leading to a high imagery overlap allow for an exceptionally high temporal resolution in Greenland. To our knowledge, our study is the first study focusing on lake phenology in Greenland on a large scale and achieving that high temporal resolution.
L117-118: “…we acquired incoming shortwave radiation and air temperature at 2 m data as climatological daily mean values for the period 1991-2020 from RACMO2” The study used “daily mean values”. The credibility of Cumulative Positive Degree Days can be enhanced if there are measured daily mean values between 2017-2022.
Since we attempt to cover a large spatial scale with high spatial variability in atmospheric variables, a seamless product from a high latitude adapted regional climate model (RCM) is an advantage. RACMO2 is a high-resolution RCM that has been shown to capture spatial and temporal variability and absolute values well (Noël et al., 2019). The current operational version RACMO2.3p2 which we used for the radiation and air temperature climatologies is validated against 37 AWSs on the Greenland Ice Sheet. Biases in daily mean 2-m temperatures and downward shortwave radiation amount to 0.14 °C and 4.8 W m-2 (latter corresponding to a bias of 2.7 %), respectively (Noël et al., 2019). In the revised version we will add this to the data description make this also clearer by discussing the discrepancy of observations vs. seamless RCM output.
L132-133: “…when most of the lake surface is…” “Most” is not a quantitative description; 80% or 90% would be a better standard.
We will rephrase the paragraph highlighting the nature of our method (averaging the backscatter of the 20 % of the central lake surface area to assess whether there is ice cover or not). However, we cannot state a definite quantitative measure since we do not perform a pixel-based ice cover classification (which would be beyond the scope and computing power at a that large sample we are working with).
L150: “Figure 1 (a)” could the sudden rise in November 𝜎 also determine the freezing of the lake?
Yes, we also assume that this period represents the freezing of the exemplary lake. However, due to the great local and inter-annual variability we were not able to establish an automated detection algorithm for freeze-up. Freeze-up is usually represented with a gradual increase in backscatter (the example given is easier to interpret than many others). In the revised version we will highlight the complexity of freeze-up better in order to better establish why we focus on break-up.
L150: “Figure 1 I” What does the pentagram represent?
We will add the pentagram to the legend, which represents the location of the exemplary lake shown in Figure 1 (a) and Figure 1 (b).
L161-162: “The progressing melt on the lake surface leading to a rougher, wetter surface explains the 𝜎0 recovery before the major backscatter decline in summer indicating lake ice break-up” rougher and wetter ice is unlikely to result in α = 0.9.
While it is true that the bare lake ice albedo might be as low as 0.2-0.5 in its latest stage before break-up (Mullen and Warren 1988, Heron and Woo 1994, Henneman and Stefan 1999, Grenfell and Perovich 2004, Jakilla et al. 2009, Semmler et al. 2012, Svacina et al. 2014a, Svacina et al. 2014b, Leppäranta 2015, Zdorovennova et al. 2018, and Robinson et al. 2021), we acknowledge that the hypothesized 8-day-earlier break-up must not only take the latest stage of the ice cover into account but surface albedos along the melt season (dry snow, wet snow, …). Since we see in our data that the transition from snow covered lake ice to open water happens within days or few weeks (also, Abermann et al. 2019) and the short-time albedo development is highly variable, we generalized the energy input changes from albedo values αi = 0.9 for snow covered lake ice to αw = 0.1 for open water. We will expand on that matter as well as discuss and highlight the transition from snow covered lake ice to open water.
L164-165: “This is due to the nature of break-up processes being more complex due to melting on top and bottom or varying acquisition conditions” could it be a secondary freeze due to lower air temperatures?
Assessing optical satellite imagery, we are very confident that in this period no secondary freeze happened. We will add this concern to the discussion.
L179: “Calculating Cumulative Positive Degree Days” whether climate averages for the period 1991-2020 are representative of today’s ice-season environmental conditions?
We acknowledge that the PDD from 2017-2021 is going to be higher compared to the 1991-2020 period due to the fact that the region is warming quickly with respect to the 1991-2020 period (Hanna et al., 2021). We utilized the 1991-2020 climatologies to get a robust comparison to the median break-up timings between 2017 and 2021. We expected that a yearly comparison to cum. PDDs or to average cum. PDDs for the period 2017-2021 are not representative of a general lake-specific generalization due to the extreme years of 2018 and 2019 (Hanna et al. 2021).
L214: “…of lake ice α (0.9) …” the value of 0.9 is too high, please revise or give a basis.
We generalized the energy input changes from albedo values αi = 0.9 for snow covered lake ice to αw = 0.1 for open water. We will expand on that matter as well as discuss and highlight the transition from snow covered lake ice to open water in contrast to bare lake ice in the late melt season to open water.
L223-224: Does the authors take into account the effect of cloud cover, which attenuates incident shortwave radiation?
The RACMO2 model output that we were working with describes incoming shortwave radiation at the surface – the attenuation effects by the atmosphere and clouds are accounted for. We will include this information in the Data section.
L280: “Figure 4 (a)” I want to know r = 0.76, then p-value = ?
We will include the probability value as well as the standard error for all reported statistical relationships.
L284: “…we there is no relationship between break-up timing and lake size” it is necessary to control for equal elevation and latitude, and to analyse only the size of the lake in order to consider that there is no relationship between disintegration and lake size
We analyzed the lake size by elevation and latitude bands, however, in order to keep the manuscript and figures concise and readable, we did not include all details in the figures. We will add the requested information to the Appendix and expand on this in the main text.
L290-291: “Figure 5b shows that lakes with similar cumulative PDDs experience a later lake ice break-up at higher elevation” are the air temperature data corresponding to lakes at different elevations at the same latitude the same? Because altitude must bring about differences in air temperature, defaulting to the same value will lead to bias.
Thank you, good point, we did not make this clear enough initially. Indeed, we derive a lake-specific time series for atmospheric input (both temperature and radiation). That way we clearly account for spatial gradients given the high resolution of RACMO. We assume by clarifying this in the text we can convince on that matter.
L332: “…down to 35 m by 1 K across…” 35 m is the average depth of regional lakes?
We greatly appreciate highlighting the need for extended explanations. From Eq. (6) and Eq. (7) we calculated the volume of ice melt at the melting point and water temperature increase from the excess energy input for each lake. Since we found that the lake surface areas explain more than 99 % of the variability in excess energy (Figure D2), we divided the volumes by the surface areas to get an estimate of the “ice column (thickness)” melted or “water column (depth)” increased by 1 K across the entire surface (ignoring bathymetry) – which is on average 35 m. The 35 m do not refer to a specific lake depth but rather indicate the average depth of water warmed due to the excess energy. We will add this calculation to the Methods section and add descriptive statistics to the Appendix.
L349: “(between 40°N and 82.5°N)” perhaps because Greenland’s latitudinal span is not large enough to become only weakly correlated.
Indeed, an expansion would be desirable – however, we stress, that the study area covered spans over approximately 1200 km latitudinal distance or approximately 11 latitudinal degrees (approximately 60-71° N), covering a very heterogeneous climate which is attributed to rugged coastal areas, with complex topography and fjord systems, covering a wide range of local climates effects. We will expand on this in the revised version.
References:
Abermann, J., Eckerstorfer, M., Malnes, E., and Hansen, B. U.: A large wet snow avalanche cycle in West Greenland quantified using remote sensing and in situ observations. Natural Hazards, 97, 517–534, https://doi.org/10.1007/s11069-019-03655-8, 2019.
Grenfell, T., C. and Perovich D., K.: Seasonal and spatial evolution of albedo in a snow-ice-land-ocean environment, Journal of Geophysical Research, 109, C1, 1-15, https://doi.org/10.1029/2003JC001866, 2004.
Henneman, H. E. and Stefan H. G.: Albedo models for snow and ice on a freshwater lake, Cold Regions Science and Technology, 29(1), 31-48, https://doi.org/10.1016/s0165-232x(99)00002-6, 1999.
Heron, R. and Woo, M.: Decay of a High Arctic lake-ice cover: observations and modelling, Journal of Glaciology, 40, 135, 1-10, https://doi.org/10.3189/S0022143000007371, 1994.
Imrit, M. A., Sharma, S.: Climate Change is Contributing to Faster Rates of Lake Ice Loss in Lakes Around the Northern Hemisphere, Journal of Geophysical Research: Biogeosciences, 126, e2020JG006134, https://doi.org/10.1029/2020JG006134, 2021.
Jakilla, J., Leppäranta, M., Kawamura, T., Shirasawa, K., and Salonen, K.: Radiation transfer and heat budget during the ice season in Lake Pääjärvi, Finland, Aquat. Ecol., 43(3), 681–692, https://doi.org/10.1007/s10452-009-9275-2, 2009.
Korhonen, J.: Long-term changes in lake ice cover in Finland, Nordic Hydrology, 37(4-5), 347-363, https://doi.org/10.2166/nh.2006.019, 2006.
Leppärantam, M. (Eds.): Freezing of Lakes and the Evolution of their Ice Cover, Springer-Verlag, Berlin Heidelberg, Germany, 309 pp., ISBN 978-3-642-29080-0, 2015.
Mullen P. C. and Warren S. G.: Theory of the Optical Properties of Lake Ice, Journal of Geophysical Research, 93, D7, 8403-8414, https://doi.org/10.1029/jd093id07p08403, 1988.
Noël, B., Van De Berg, W., J., Lhermitte, S., and Van Den Broeke, M.: Rapid ablation zone expansion amplifies north Greenland mass loss, Science Advances, 5, 9, 1-9, https://doi.org/10.1126/sciadv.aaw0123, 2019.
Robinson, A. L., Ariano, S., S., and Brown, L. C.: The Influence of Snow and Ice Albedo towards Improved Lake Ice Simulations, Hydrology 8(1), 11, 1-21, https://doi.org/10.3390/hydrology8010011, 2021.
Semmler, T., Cheng, B., Yang, Y., and Rontu, L.: Snow and ice on Bear Lake (Alaska) – sensitivity experiments with two lake ice models, Tellus A: Dynamic Meteorology and Oceanography, 64(1), 1-14, https://doi.org/10.3402/tellusa.v64i0.17339, 2012.
Svacina, N. A., Duguay, C. R., and Brown, L. C.: Modelled and satellite-derived surface albedo of lake ice - Part I: evaluation of the albedo parameterization scheme of the Canadian Lake Ice Model, Hydrological Processes, 28(16), 4550–4561, https://doi.org/10.1002/hyp.10253, 2014a.
Svacina, N. A., Duguay, C. R., and Brown, L. C.: Modelled and satellite-derived surface albedo of lake ice - Part II: evaluation of MODIS albedo products, Hydrological Processes, 28(16), 4562–4572, https://doi.org/10.1002/hyp.10257, 2014b.
Taurisano, A., Bøggild, C. E., and Karlsen H. G.: A century of climate variability and climate gradients from coast to ice sheet in West Greenland, Geogr. Ann., 86 A (2), 217–224, https://doi.org/10.1111/j.0435-3676.2004.00226.x, 2004.
Wang, J., Duguay, C. R., Clausi, D. A., Pinard, V., and Howell, S. E. L.: Semi-Automated Classification of Lake Ice Cover Using Dual Polarization RADARSAT-2 Imagery, Remote Sens., 10(11), 1727, 1-27, https://doi.org/10.3390/rs1011172, 2018.
Zdorovennova, G., Palshin, N., Efremova, T., Zdorovennov, R., Gavrilenko, G., Volkov, S., Bodanov, S., and Terzhevik, A.: Albedo of a Small Ice-Covered Boreal Lake: Daily, Meso-Scale and Interannual Variability on the Background of Regional Climate, Geosciences, 8(6), 206, 1-17, https://doi.org/10.3390/geosciences8060206, 2018.
Citation: https://doi.org/10.5194/egusphere-2023-1762-AC1
-
AC1: 'Reply on RC1', Christoph Posch, 31 Oct 2023
-
RC2: 'Comment on egusphere-2023-1762', Anonymous Referee #2, 25 Sep 2023
General comment:
I enjoyed reading the paper. The text is clear. The quality of the plots are very good. I think it is a nice paper and contributes to the field.I have one general (major) comment:
I think the discussion is somewhat lacking in putting the work in the broader context. It is not well motivated why this study matters from a global perspective and what it adds to the general understanding of the climate change impacts. More specifically:- US is a wide continent covering various Köppen-Geiger climatic zone but Greenland is specifically located in one climatic zone and is not spanning across multiple latitudes with different climatic features. Therefore, I am not too surprised that for US there was a strong spatial correlation while such correlation could not be observed in Greenland. Please consider adding more to this discussion and make the comparison a bit stronger.
- Put the whole study in the broader context: Maybe mention implication for hydropower. You can use snow-dominated locations for the sake of comparison. I tried to find some examples and the discussion section in paper https://www.sciencedirect.com/science/article/pii/S0022169423007497 and https://hess.copernicus.org/articles/24/3815/2020/ might work. Please try to find other papers to add to this point.
- Compare the finding with other boreal countries: Note https://tc.copernicus.org/articles/16/2493/2022/tc-16-2493-2022-discussion.html where they studied ice break-up patterns in Sweden. Please try to find other papers to add to this point.
- Mention what is expected to happen to Greenland under climate change. You can use latest IPCC report for this.
Specific comments:
- L280: please remove we.
- Please report the significance of the correlations whenever you report the correlation strength.
- The paper lacks climatic description of Greenland.
Citation: https://doi.org/10.5194/egusphere-2023-1762-RC2 -
AC2: 'Reply on RC2', Christoph Posch, 31 Oct 2023
Dear editor, dear referees,
we are very grateful for two very constructive reviews and the editorial advice and appreciate the valuable time put into this. We believe by incorporating the reviews we will manage to achieve a much more mature manuscript.
In the following we mark bold the comments given by the referees and give our answers and comments in italic. We hope that the responses will satisfy both referees and editor, and we are looking forward to a decision and to revising the manuscript.
Once again, many thanks for the valuable input and all the best,
Christoph, on behalf of the author team.
General comment:
I enjoyed reading the paper. The text is clear. The quality of the plots are very good. I think it is a nice paper and contributes to the field.
Thank you and we greatly appreciate the constructive remarks. We will seriously consider the incorporation of the raised points them in our revised manuscript.
I have one general (major) comment:
I think the discussion is somewhat lacking in putting the work in the broader context. It is not well motivated why this study matters from a global perspective and what it adds to the general understanding of the climate change impacts. More specifically:
- US is a wide continent covering various Köppen-Geiger climatic zone but Greenland is specifically located in one climatic zone and is not spanning across multiple latitudes with different climatic features. Therefore, I am not too surprised that for US there was a strong spatial correlation while such correlation could not be observed in Greenland. Please consider adding more to this discussion and make the comparison a bit stronger.
- Thank you for pointing out that there is a need for more explanation of the comparison and to make it stronger. We will add a discussion on climatic characteristics and trends of Greenland focusing on coastal climates and their variability (i.e., Taurisano et al. 2004, Hanna et al. 2021) and contextualize them with our results. The spatial climate variability of Greenland is well established and while it is true that it lies within one Köppen-Geiger climatic zone, we stress its heterogeneity. This is true for meteorological input (Hanna et al. 2020, Box et al. 2023) as well as local surface mass balance variability (Mankoff et al. 2021, Slater et al. 2021). Our study in fact shows regional differences of the response of lake ice break-up. We will elaborate on that in the discussion and stress the difference of the spatial gradients between the US study and Greenland.
- Put the whole study in the broader context: Maybe mention implication for hydropower. You can use snow-dominated locations for the sake of comparison. I tried to find some examples and the discussion section in paper https://www.sciencedirect.com/science/article/pii/S0022169423007497 and https://hess.copernicus.org/articles/24/3815/2020/ might work. Please try to find other papers to add to this point.
- This is a good point, thank you for the literature suggestions. Indeed, the timing of lake break-up particularly impacts challenges regarding hydropower and we will embed this larger framing in the discussion, referring to further studies such as Prowse et al. (2011), Gebre et al. (2013) and Cherry et al. (2017).
- Compare the finding with other boreal countries: Note https://tc.copernicus.org/articles/16/2493/2022/tc-16-2493-2022-discussion.html where they studied ice break-up patterns in Sweden. Please try to find other papers to add to this point.
- We will discuss and compare our findings with other high latitude and boreal countries. Thank you for the literature suggestion, we will refer to further studies such as Korhonen (2006), L’Abée-Lund et al. (2021), Zhang et al. (2021) and Cai et al. (2022).
- Mention what is expected to happen to Greenland under climate change. You can use latest IPCC report for this.
- We will add this to our discussion and draw an outlook on expected future changes referring to the projections of the IPCC report. Furthermore, we will discuss them in context with studies on lake ice phenology projections such as Imrit and Sharma (2021), and Huang et al. (2022).
Specific comments:
- L280: please remove we.
- We will remove “we”.
- Please report the significance of the correlations whenever you report the correlation strength.
- We will include all probability values and standard errors of our reported statistical relationships.
- The paper lacks climatic description of Greenland.
- We will add a discussion on climatic characteristics and trends of Greenland focusing on coastal climates and their variability (i.e., Taurisano et al. 2004, Hanna et al. 2021) and contextualize them with our results.
References:
Box, J. E., Nielsen, K. P., Yang, X., Niwano, M., Wehrlé, A., van As, D., Fettweis, X., Køltzow, M. A. Ø., Palmason, B., Fausto, R. S., van den Broeke, M. R., Huai, B., Ahlstrøm, A. P., Langley, K., Dachauer, A., and Noël, B.: Greenland ice sheet rainfall climatology, extremes and atmospheric river rapids, Meteorological Applications, 30, 4, 1-24, https://doi.org/10.1002/met.2134, 2023.
Cai, Y., Duguay, C. R., Ke, C.-Q.: A 41-year (1979–2019) passive-microwave-derived lake ice phenology data record of the Northern Hemisphere, Earth Syst. Sci. Data, 14, 3329–3347, https://doi.org/10.5194/essd-14-3329-2022, 2022.
Cherry, J. E., Knapp, C., Trainor, S., Ray, A. J., Tedesche, M., Walker, S.: Planning for climate change impacts on hydropower in the Far North, Hydrol. Earth Syst. Sci., 21, 133–151, https:// doi:10.5194/hess-21-133-2017, 2017.
Gebre, S., Alfredsen, K., Lia, L., Stickler, M., Tesaker, E.: Review of Ice Effects on Hydropower Systems, Journal of Cold Regions Engineering, 27(4), 196–222, https://doi.org/10.1061/(ASCE)CR.1943-5495.0000059, 2013.
Hanna, E., Capellen, J., Fettweis, X., Mernild, S. H., Mote, T. L., Mottram, R., Steffen, K., Ballinger, T. J., and Hall, R. J.: Greenland surface air temperature changes from 1981 to 2019 and implications for ice-sheet melt and mass-balance change, Int J. Climatol., 41, 1336–1352, https://doi.org/10.1002/joc.6771, 2021.
Huang, L., Timmermann, A., Lee, S.-S., Rodgers, K. B., Yamaguchi, R., and Chung, E.-S.: Emerging unprecedented lake ice loss in climate change projections, Nature Communications, 13(5798), 1-12, https://doi.org/10.1038/s41467-022-33495-3, 2022.
Imrit, M. A., Sharma, S.: Climate Change is Contributing to Faster Rates of Lake Ice Loss in Lakes Around the Northern Hemisphere, Journal of Geophysical Research: Biogeosciences, 126, e2020JG006134, https://doi.org/10.1029/2020JG006134, 2021.
Korhonen, J.: Long-term changes in lake ice cover in Finland, Nordic Hydrology, 37(4-5), 347-363, https://doi.org/10.2166/nh.2006.019, 2006.
L’Abée-Lund, J. H., Vøllestad, L. A., Brittain, J. E, Kvambekk, Å. S., and Solvang, T.: Geographic variation and temporal trends in ice phenology in Norwegian lakes during the period 1890–2020, The Cryosphere, 15, 2333–2356, https://doi.org/10.5194/tc-15-2333-2021, 2021.
Mankoff, K. D., Fettweis, X., Langen, P. L., Stendel, M., Kjeldsen, K. K., Karlsson, N. B., Noël, B., van den Broeke, M. R., Solgaard, A., Colgan, W., Box, J. E., Simonsen, S. B., King, M. D., Ahlstrøm, A. P., Andersen, S. B., and Fausto, R. S.: Greenland ice sheet mass balance from 1840 through next week, Earth Syst. Sci. Data, 13, 5001–5025, https://doi.org/10.5194/essd-13-5001-2021, 2021.
Prowse, T., Alfredsen, K., Beltaos, S., Bonsal, B. R., Bowden, W. B., Duguay, C. R., Korhola, A., McNamara, J., Vincent, W. F., Vuglinsky, V., Walter Anthony, K. M., and Weyhenmeyer, G. A.: Effects of Changes in Arctic Lake and River Ice, AMBIO 2011, 40, 63–74, https://doi.org/10.1007/s13280-011-0217-6, 2011.
Slater, T., Shepherd, A., McMillan, M., Leeson, A., Gilbert, L., Muir, A., Munneke, P. K., Noël, B., Fettweis, X., van den Broeke, M., and Briggs, K.: Increased variability in Greenland Ice Sheet runoff from satellite observations, Nature Communications, 12, 6069, 1-9, https://doi.org/10.1038/s41467-021-26229-4, 2021.
Taurisano, A., Bøggild, C. E., and Karlsen H. G.: A century of climate variability and climate gradients from coast to ice sheet in West Greenland, Geogr. Ann., 86 A (2), 217–224, https://doi.org/10.1111/j.0435-3676.2004.00226.x, 2004.
Zhang, S., Pavelsky, T. M., Arp, C. D., and Yang X.: Remote sensing of lake ice phenology in Alaska, Environ. Res. Lett. 16, 1-12, https://doi.org/10.1088/1748-9326/abf965, 2021.
Citation: https://doi.org/10.5194/egusphere-2023-1762-AC2
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 329 | 143 | 30 | 502 | 15 | 21 |
- HTML: 329
- PDF: 143
- XML: 30
- Total: 502
- BibTeX: 15
- EndNote: 21
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Christoph Posch
Jakob Abermann
Tiago Manuel Ferreira da Silva
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(4200 KB) - Metadata XML