Pb-210 Dating of Ice Scour in the Kara Sea
, Irina Usyagina 3, Nikita Meshcheriakov 1,3, Roman Ananiev 4, Vasiliy Arkhipov 1,2, Aino Kirillova 1, Stepan Maznev 1,2, Sergey Nikiforov 4 and Nikolay Sorokhtin 4Round 1
Reviewer 1 Report
This paper describes a method to date an ice scour, estimating the sedimentation rate inside the scour via 210Pb method. The aim of the paper is interesting but there are some issues that should be discussed about the dating of the core taken inside the scour before considering for publication. In fact, since the sediment inside the scour is accumulating in a constrained environment (the shape of the scour itself) some basic assumption of dating models are not verified. It is necessary to adapt the basic equation to the specific geometry of this system. The basic assumption of CF model is that the inventory under a given layer is the same as the total inventory if corrected for decay. But inside the scour, every layer corresponds to a different wideness, this effect being more and more important moving near the bottom of the scour. Moreover, to apply CF model, even weighing differently each layer for its wideness, it would be necessary to reach the equilibrium depth, otherwise it would be possible that the whole core is “younger” than 110 years. But this is not the case in the sampled core.
Line 170 Please, add some more detail about cores sampling: Did you use the same device to sample both cores? Which is the diameter of your corer? Fig 2 B clearly explains why you took as external reference a longer core, but how did you manage to control the core length? According to your opinion, is it possible that due to its longer penetration in the ANS-52-17 core the surficial layers has been compressed, squeezing away a fraction of porewater (see line 351)? Did you observe any shortening of this core? Did you measure any short-lived radionuclide (7Be, 234Th) in the surficial layer to check if you had correctly sampled the surficial layer?
Line 225 How did you calibrate the efficiency of your detector? Did you perform any quality check?
Line 229 Why did you use 1120 keV line instead of 609 keV which has a higher branching ratio and a higher detector efficiency?
Line 256, in an open, unconstrained environment, to get an exponential profile, you need also a constant sedimentation.
Line 267 I do not think that those formulae may be used inside the scour. Please, consider that as you have said in line 446, the scour works as a sedimentological trap: it collects the whole vertical flux falling in its width (about 35 m) and focus it on its bottom. But since its shape is roughly triangular (see fig 3c), the focusing factor (nowadays about 3, comparing the flat region at the bottom with the width) was previously much higher (since the flat region was shorter). In such a constrained environment, the 210Pb profile cannot be stationary because in each layer, the incoming flux is distributed in a surface wider than the previous one. So, the same inputs in different layers should bring to different sedimentation rates.
Lines 380-386 Fig 4 a suggests that you have missed the inventory underlying the 43 cm layer, since this profile does not reach the equilibrium depth. Even in an unconstrained environment, you should estimate this missing inventory (see for example par. 3.5 in Sanchez-Cabeza and Ruiz-Fernandez 2012).
Line 391-397. All those remarks are affected by the problems in core ans-52-16 dating.
Fig 5 Can you explain the alternate trend in SAR during years 1965-2000 in core ans-52-16?
Line 401 Identifying the lowest layer with 137Cs as 1954 is not a robust hypothesis: the 137Cs has a low Kd (IAEA 2004) and can migrate through porewater to reach deeper layer (see for example Abril et al. 1991, or Zaborska et al 2008, where the sedimentation rates estimated by 137Cs are sometime twice the 210Pb ones)
Line 404 To correlate this 137Cs peak to 1962 explosions in Novaya Zemlya is a more robust hypothesis. Such a sharp peak suggests a local input: is it possible that the main sources of sediments that have accumulated in the following years has been less affected by the local fallout. On the contrary, the global fallout should bring a more homogeneous signal (and a step distribution instead of a peaked one).
Line 438 Why do you say that higher hydrodynamics should affect the coarser sediments (sand and gravel)? Moreover, the 137Cs is a tracer of the finer sediments and you find missing 137Cs too.
Par 5.3 The whole discussion is based on a flawed chronology in core 16.
Abril J, García-León M, García-Tenorio R, Sánchez C, El-Daoushy F (1991) Dating of marine sediments by an incomplete mixing model. J Environ Radioact 15(2): 135 – 151
IAEA (2004) Sediment distribution coefficients and concentration factors for biota in the marine environment. International Atomic Energy Agency Technical report series 422, Vienna, Austria
Author Response
Dear Reviewer, thank you for reading our work carefully and helping to improve it, this was very appreciated. Below you can find a point-by-point responses in red to your comments and in the manuscript, you can find all the corrections and integrations.
Point 1: This paper describes a method to date an ice scour, estimating the sedimentation rate inside the scour via 210Pb method. The aim of the paper is interesting but there are some issues that should be discussed about the dating of the core taken inside the scour before considering for publication. In fact, since the sediment inside the scour is accumulating in a constrained environment (the shape of the scour itself) some basic assumption of dating models are not verified. It is necessary to adapt the basic equation to the specific geometry of this system. The basic assumption of CF model is that the inventory under a given layer is the same as the total inventory if corrected for decay. But inside the scour, every layer corresponds to a different wideness, this effect being more and more important moving near the bottom of the scour. Moreover, to apply CF model, even weighing differently each layer for its wideness, it would be necessary to reach the equilibrium depth, otherwise it would be possible that the whole core is “younger” than 110 years. But this is not the case in the sampled core.
Response 1: We agree with you that the application of the CF model in the ice scour conditions may affect the results correctness. We have added an additional section to Discussion (5.1) in which we give a detailed estimation of the possible uncertainty in the results due to the changing width of the ice scour bottom during sedimentation. In our case, this uncertainty is insignificant (about 2%), which allows us to apply this model for our “ice scour” core.
Point 2: Line 170 Please, add some more detail about cores sampling: Did you use the same device to sample both cores? Which is the diameter of your corer? Fig 2 B clearly explains why you took as external reference a longer core, but how did you manage to control the core length? According to your opinion, is it possible that due to its longer penetration in the ANS-52-17 core the surficial layers has been compressed, squeezing away a fraction of porewater (see line 351)? Did you observe any shortening of this core? Did you measure any short-lived radionuclide (7Be, 234Th) in the surficial layer to check if you had correctly sampled the surficial layer?
Response 2: Yes, we used the same device to sample both cores (Table 1). Corer diameter is 147 mm (specified in the text of section 3.3 and in Table 1). The core length was not controlled by us. In section 5.3, we have added an explanation of the core lengths difference. In our opinion, the penetration depth does not affect the compaction of the upper layers, that is, the upper layers of both cores experienced the same load, regardless of what the penetration depth was. We haven't observed any core shortening. We did not have the goal of short-lived radionuclides specific activity measurements. However, we established the results of 7Be measurements from the stored spectra of core sediments. Appropriate additions were made in Sections 3.5 and 4.3.
Point 3: Line 225 How did you calibrate the efficiency of your detector? Did you perform any quality check?
Response 3: We check the quality of the detector efficiency on a regular basis. Information about this has been added to section 3.5.
Point 4: Line 229 Why did you use 1120 keV line instead of 609 keV which has a higher branching ratio and a higher detector efficiency?
Response 4: During measurements, we used the certified "Methodology for measuring the activity (specific activity) of gamma-emitting radionuclides in countable samples using the CANBERRA gamma-ray energy spectrometer with GENIE 2000 software", which recommends the use of this line. The methodology is included in the register of certified methods of the Russian Federation, which is in the public domain, https://fgis.gost.ru/fundmetrology/registry/16/items/302615 (works only with Russian IP addresses).
Point 5: Line 256, in an open, unconstrained environment, to get an exponential profile, you need also a constant sedimentation.
Response 5: The phrase "there are no processes leading to its mixing or redistribution" suggests "constant sedimentation". Clarification added in brackets.
Point 6: Line 267 I do not think that those formulae may be used inside the scour. Please, consider that as you have said in line 446, the scour works as a sedimentological trap: it collects the whole vertical flux falling in its width (about 35 m) and focus it on its bottom. But since its shape is roughly triangular (see fig 3c), the focusing factor (nowadays about 3, comparing the flat region at the bottom with the width) was previously much higher (since the flat region was shorter). In such a constrained environment, the 210Pb profile cannot be stationary because in each layer, the incoming flux is distributed in a surface wider than the previous one. So, the same inputs in different layers should bring to different sedimentation rates.
Response 6: In the new section 5.1, we have shown that CF model can be used in our case with insignificant uncertainty.
Point 7: Lines 380-386 Fig 4 a suggests that you have missed the inventory underlying the 43 cm layer, since this profile does not reach the equilibrium depth. Even in an unconstrained environment, you should estimate this missing inventory (see for example par. 3.5 in Sanchez-Cabeza and Ruiz-Fernandez 2012).
Response 7: The estimation of lead reserves in layers below 42 cm has been added to section 4.4.
Point 8: Line 391-397. All those remarks are affected by the problems in core ans-52-16 dating.
Response 8: See Response 1.
Point 9: Fig 5 Can you explain the alternate trend in SAR during years 1965-2000 in core ans-52-16?
Response 9: This alternate trend is the result of a systematic error in the calculations, which was corrected along with the Figure.
Point 10: Line 401 Identifying the lowest layer with 137Cs as 1954 is not a robust hypothesis: the 137Cs has a low Kd (IAEA 2004) and can migrate through porewater to reach deeper layer (see for example Abril et al. 1991, or Zaborska et al 2008, where the sedimentation rates estimated by 137Cs are sometime twice the 210Pb ones).
Line 404 To correlate this 137Cs peak to 1962 explosions in Novaya Zemlya is a more robust hypothesis. Such a sharp peak suggests a local input: is it possible that the main sources of sediments that have accumulated in the following years has been less affected by the local fallout. On the contrary, the global fallout should bring a more homogeneous signal (and a step distribution instead of a peaked one).
Response 10: In our case, the first appearance of 137Cs occurs in the layer 30-32, and just 2 cm higher (layer 34-36 cm) a sharp peak is observed, exceeding the first appearance by 23 times. We associate this peak with 1963, when the peak of 137Cs atmospheric concentration was recorded. If we accept 1962, then this will not fundamentally affect the verification. This peak could not be formed due to the 137Cs migration, since the specific activity in it significantly exceeds the specific activity of adjacent layers. If we assume that 137Cs accumulated in the 2 underlying layers due to migration, then we do not have a signal of the first thermonuclear explosion. In this case, only one peak verifies our 210Pb results. However, our assumption about 1954 in layer 30-32 is in good agreement with the 210Pb results and only complements the verification by the 137Cs peak in 1963. Therefore, it is unlikely that such a coincidence is associated only with 137Cs migration. We believe that the key role in this case is played by the high increase in the content of cesium in the atmosphere in 1954, which is widely known in the literature.
Point 11: Line 438 Why do you say that higher hydrodynamics should affect the coarser sediments (sand and gravel)? Moreover, the 137Cs is a tracer of the finer sediments and you find missing 137Cs too.
Response 11: Only higher hydrodynamics can lead to the removal of coarse particles from the background surface. Under conditions of low hydrodynamics in the ice scour, coarse particles originating from ice rafting, slope processes and saltation are not able to move laterally and remain where they settled to the bottom.
Point 12: Par 5.3 The whole discussion is based on a flawed chronology in core 16.
Response 12: We do not agree with this. We admit that our chronology in the ice scour may have a higher uncertainty than on the background surface, but this does not mean that it is flawed. We have added an additional section to Discussion (5.1) in which we give a detailed assessment of the possible uncertainty of the results in the case of the core 16.
Reviewer 2 Report
Dear Autors,
I read your article entitled 'Pb-210 Dating of Ice Scour in the Kara Sea' with a great interest . Please find attached pdf with my review where you can find my corrections and suggestions that I have proposed to improve a presentation of your work.
Sincerely
Comments for author File: 
Comments.pdf
Author Response
Dear Reviewer, thank you for reading our work carefully and helping to improve it, this was very appreciated. We did almost all the corrections that you proposed and followed your suggestions. In the manuscript you can find all the corrections and integrations. Below you can find a point-by-point responses in red to your comments with which we do not agree.
Point 1: Lines 235, 452. Space.
Response 1: According to the rules of the journal, in these cases [31–33,40–44] and [58,59], there should be no space.
Point 2: Lines 411, 427, 501, 510, 513. What doeas it mean? January 1967? – 1967.1 AD. Such an information may be unclear to readers. – 1965.6 AD. Please make these date clear.
Response 2: The obtained calendar ages have the form of decimal fractions since they are calculated mathematically relative to the date of the core selections – November 2, 2021 (2021.8 AD). Clarification has been added to the section 4.4.
Point 3: Lines 433, 435, 445. "Ice scour".
Response 3: We use "ice scour" in quotes when referring to a core name. If we are talking about the landform, then we do not use quotes. In these cases, this term is used in the sense of the landform.
Point 4: Lines 435, 532. I recommend to add the information after this sentence: Fine sediments carries particularly sensitive data, therefore, automated methods are reccomended to their grain size analysis, as the automated mineralogy (Pszonka et al., 2021; Pszonka and Schulz, 2022).
The grain size analysis of such fine and thus tough sediments is tricky and appropriate for automated methods such as automated mineralogy.
Response 4: For grain size analysis, we used the classical method, which has proven itself for a long time. Our method is considered the most accurate for fine sediments, as it is based on the hydraulic settling of particles according to the Stokes formula, the accuracy of settling at each stage is controlled by a microscope. The objectives of this work did not include a detailed study of mineralogy. Therefore, the recommended sentence does not seem appropriate to us. Nevertheless, we are grateful for the provided references. We will definitely read them and, perhaps, apply the proposed method in our future work.
Reviewer 3 Report
This paper estimated date the ice scour in Kara Sea based on measurements of the specific activity of the natural radionuclide 210Pb and the concentration of atmospheric nuclear test-derived 137Cs. As a result, the formation date is estimated to be 1810 ± 30 AD, with subsequent estimates of the rate of deposition. I believe that the methodology is reasonable and it presents important findings. However, some figures are not legible and should be rewritten. I believe that this paper is suitable for the publication in the Journal of Marine Science and Engineering after minor revisions.
Specific comments.
L. 98; Fig. 1 A larger map should be shown at the same time to show the location of the Kara Sea.
L. 142; Fig. 2 The graph shown in B is not legible. Please make it a little bigger. Also, the Gouge is difficult to read, please describe it in more detail.
L. 289; Please describe Ice Scour clearly in Fig. 2.
L. 320; Fig. 3 Please describe the spatial scale of the horizontal axis.
Author Response
Dear Reviewer, thank you for reading our work carefully and helping to improve it, this was very appreciated. Below you can find a point-by-point responses in red to your comments. In the manuscript you can find all the corrections and integrations.
Point 1: L. 98; Fig. 1 A larger map should be shown at the same time to show the location of the Kara Sea.
Response 1: A larger map showing the location of the Kara Sea has been included in Figure 1.
Point 2: L. 142; Fig. 2 The graph shown in B is not legible. Please make it a little bigger. Also, the Gouge is difficult to read, please describe it in more detail.
289; Please describe Ice Scour clearly in Fig. 2.
Response 2: Figures 2A and 2B were made larger. A detailed description of the ice scour is given in the section 4.1.
Point 3: L. 320; Fig. 3 Please describe the spatial scale of the horizontal axis.
Response 3: The horizontal scale has been added in Figure 3.
Round 2
Reviewer 1 Report
I have appreciated the very accurate answers of the authors to many of my questions.
I think however that there are still a couple of issues to be discussed. The basis of the CF model is that, if one assumes a constant 210Pbex flux towards the sediment, the total inventory (Bq m-2) in a fixed area remains constant. Hence, it is possible to estimate the age of one “cut” dividing the total inventory by the inventory underlying that cut, using the third formula in supplementary materials. This calculus holds, for example, in core ANS-52-17 with no lateral constraint.
But the ANS-52-16 core has some lateral boundaries, so it requires an approach outside the canon. The boundary of the ice scour should be a layer of sediment compressed and mixed by the interaction with the iceberg and, in the assumption that the trench is old enough, with no 210Pbex content. If one assumes that the 210Pbex flux incoming into a trench section is constant in time the inventory in the whole section of trench remains constant. So, if one assumes that specific activities are spatially homogeneous across the trench, to get the correct date of one “cut” one should get the total inventory in the trench under that depth, multiplying the inventory of each slice of the core DAi times the width of the trench at that depth (let’s call it W(i)) and summing on slices to calculate new A’(i) values. To get r(i) you should calculate r(i)= lA’(i)/(c(i)W(i)). s(i) calculation does not change.
If you compare the SAR between the two cores, you will observe a similar feature: a sharp increase occurred between 2002-2004 in core 16 and between 1988-1994 in core 17. If this reflects a significant environmental change, it should have occurred in the same year in both cores. The new calculation may reduce the time gap between the two cores.
Another critical point is the estimate of the missing inventory: figure 4 shows the differences between a pure exponential decay and the data experimentally observed, so I do not think that the exponential decay is a good estimate. My suggestion is to use a time marker to estimate the missing inventory (par 3.5.2 in SC and RF 2012 reference) and the 1963 peak to check if the estimate is accurate enough. As time marker you can use for example the abrupt change in sedimentation rate or the 1954 signal of 137Cs input.
Line 491 as I have tried to explain before, the difference in 1 cm in width in each layer is correctly not significant. But if you use the usual CF formulae, the same mass flux at the surface and around 40 cm depth should have produced the same sar while instead inside the trench it would have settled twice a thick layer.
Line 504 As you say there are strong similarity between the cores, with the suggested correction you could get a better agreement in dating. One should not expect a perfect agreement due to the approximation assumed, but you can get a match better than 14 years.
In line 506 you quote the differences in sedimentation: you could calculate the magnitude of this effect comparing the inventory in core ANS-52-17 times the width of the mouth of the trench with the total inventory calculated inside the trench, so you will estimate how effective is the ice scour as a “sediment trap”.
In supplementary materials: if you write r(i) in g cm-2 yr-1 and s(i) in cm yr-1 and r(i) in g cm-3 you do not need any conversion factor (and in fact you do not use any in your tables). You have correctly used a unit conversion to calculate r(i) from A(i) and c(i) but it was not explicitly declared.
I am sorry to propose this heavy change (the general scheme of the discussion will remain the same, but a lot of date should be slightly corrected) but I made my suggestion because I really appreciated the aim of this study and I think it is worthy of a supplementary effort.
I found this problem quite interesting, and I hope to have been helpful.
Author Response
Dear Reviewer, thank you for reading our new version of the paper carefully and helping to improve it, this was very appreciated. Below you can find a point-by-point responses in red to your comments and in the manuscript, you can find all the corrections and integrations.
Point 1: I have appreciated the very accurate answers of the authors to many of my questions.
I think however that there are still a couple of issues to be discussed. The basis of the CF model is that, if one assumes a constant 210Pbex flux towards the sediment, the total inventory (Bq m-2) in a fixed area remains constant. Hence, it is possible to estimate the age of one “cut” dividing the total inventory by the inventory underlying that cut, using the third formula in supplementary materials. This calculus holds, for example, in core ANS-52-17 with no lateral constraint.
But the ANS-52-16 core has some lateral boundaries, so it requires an approach outside the canon. The boundary of the ice scour should be a layer of sediment compressed and mixed by the interaction with the iceberg and, in the assumption that the trench is old enough, with no 210Pbex content. If one assumes that the 210Pbex flux incoming into a trench section is constant in time the inventory in the whole section of trench remains constant. So, if one assumes that specific activities are spatially homogeneous across the trench, to get the correct date of one “cut” one should get the total inventory in the trench under that depth, multiplying the inventory of each slice of the core DAi times the width of the trench at that depth (let’s call it W(i)) and summing on slices to calculate new A’(i) values. To get r(i) you should calculate r(i)= lA’(i)/(c(i)W(i)). s(i) calculation does not change.
Response 1: We made a calculation according to the modified formulas of the CF model and designated it as CFw (Table S3 in the Supplementary Materials; Discussion 5.2). The calendar age obtained from the CFw calculation turned out to be no more than 10 years earlier than the age obtained from the CF model (Figure 5, curve “Pb-210 age, ice scour, CFw/45”). We believe that this age adjustment is not significant for the age estimate of the ice scour, which turned out to be 1810±30.
Point 2: If you compare the SAR between the two cores, you will observe a similar feature: a sharp increase occurred between 2002-2004 in core 16 and between 1988-1994 in core 17. If this reflects a significant environmental change, it should have occurred in the same year in both cores. The new calculation may reduce the time gap between the two cores.
Response 2: We have made an attempt to improve the SAR correlation between cores. According to the new calculation CFw compared to the background surface, the lag of sharp changes in SAR in the ice scour has decreased by no more than 10 years, but still remains and reaches about 10 years (Figure 6, C). Considering that the “ice scour” core is more detailed and has lower uncertainty than the “background” core, for better SAR correlation between the cores, the adjustment of the layers ages is necessary first of all for the background surface.
Point 3: Another critical point is the estimate of the missing inventory: figure 4 shows the differences between a pure exponential decay and the data experimentally observed, so I do not think that the exponential decay is a good estimate. My suggestion is to use a time marker to estimate the missing inventory (par 3.5.2 in SC and RF 2012 reference) and the 1963 peak to check if the estimate is accurate enough. As time marker you can use for example the abrupt change in sedimentation rate or the 1954 signal of 137Cs input.
Response 3: We have estimated the missing inventory in the layers below the 43 cm not penetrated by the corer using a time marker. Information about this has been added to section 5.1.
Point 4: Line 491 as I have tried to explain before, the difference in 1 cm in width in each layer is correctly not significant. But if you use the usual CF formulae, the same mass flux at the surface and around 40 cm depth should have produced the same sar while instead inside the trench it would have settled twice a thick layer.
Response 4: See Response 1.
Point 5: Line 504 As you say there are strong similarity between the cores, with the suggested correction you could get a better agreement in dating. One should not expect a perfect agreement due to the approximation assumed, but you can get a match better than 14 years.
Response 5: See Response 2.
Point 6: In line 506 you quote the differences in sedimentation: you could calculate the magnitude of this effect comparing the inventory in core ANS-52-17 times the width of the mouth of the trench with the total inventory calculated inside the trench, so you will estimate how effective is the ice scour as a “sediment trap”.
Response 6: A quantitative evaluation of the sediment trap effect was carried out in section 5.4. It showed that the studied ice scour accumulated 4.7 times more lead than the background surface.
Point 7: In supplementary materials: if you write r(i) in g cm-2 yr-1 and s(i) in cm yr-1 and r(i) in g cm-3 you do not need any conversion factor (and in fact you do not use any in your tables). You have correctly used a unit conversion to calculate r(i) from A(i) and c(i) but it was not explicitly declared.
I am sorry to propose this heavy change (the general scheme of the discussion will remain the same, but a lot of date should be slightly corrected) but I made my suggestion because I really appreciated the aim of this study and I think it is worthy of a supplementary effort.
I found this problem quite interesting, and I hope to have been helpful.
Response 7: We removed the mention of the conversion factor from Supplementary Materials.
Round 3
Reviewer 1 Report
I appreciated the analyses done on the ice scour core.
I think that the paper can be published in the present form and I thank the authors for their effort
