Revising chronological uncertainties in marine archives using global anthropogenic signals: <em>a case study</em>

Irvalı, Nil; Ninnemann, Ulysses S.; Olsen, Are; Rose, Neil L.; Thornalley, David J. R.; Mjell, Tor L.; Counillon, François

doi:https://doi.org/10.5194/egusphere-2023-2845

Preprints

https://doi.org/10.5194/egusphere-2023-2845

Preprints

15 Dec 2023

| 15 Dec 2023

Revising chronological uncertainties in marine archives using global anthropogenic signals: a case study

Nil Irvalı, Ulysses S. Ninnemann, Are Olsen, Neil L. Rose, David J. R. Thornalley, Tor L. Mjell, and François Counillon

Abstract. Marine sediments are excellent archives for reconstructing past changes in climate and ocean circulation. Overlapping with instrumental records they hold the potential to elucidate natural variability and contextualize current changes. Yet, dating uncertainties of traditional approaches (e.g., up to ± 30–50 years, for the last two centuries) pose major challenges for integrating the shorter instrumental records with these extended marine archives. Hence, robust sediment chronologies are crucial and most existing age model constraints do not provide sufficient age control, particularly for the 20^th century, which is the most critical period for comparing proxy records to historical changes. Here we propose a novel chronostratigraphic approach that uses anthropogenic signals such as the oceanic ¹³C Suess effect and spheroidal carbonaceous fly ash particles to reduce age model uncertainties in high-resolution marine archives. As a test, we apply this new approach to a marine sediment core located at the Gardar Drift, in the subpolar North Atlantic, and revise the previously published age model for this site. We further provide refined estimate of regional reservoir corrections and uncertainties for Gardar Drift.

Received: 29 Nov 2023 – Discussion started: 15 Dec 2023

Download & links

Preprint (PDF, 1950 KB)

Supplement (291 KB)

Download & links

Nil Irvalı, Ulysses S. Ninnemann, Are Olsen, Neil L. Rose, David J. R. Thornalley, Tor L. Mjell, and François Counillon

Status: final response (author comments only)

RC1:
'Comment on egusphere-2023-2845', James David Scourse, 19 Jan 2024
This paper reports a new approach to the establishment of reliable age models for very recent marine sediment cores (last 200 years). This is often a difficult and sometimes intractable problem for very recent cores and core-tops, because of the problems inherent in offsets between, for instance, radiocarbon, which has very poor precision for the last few hundred years, and age models based on Pb and Cs isotopes. This problem becomes significant because this is exactly the time period for which precise and accurate age models are required to calibrate proxy with instrumental data. These problems are rehearsed and explained well in this generally very well written contribution. In practice, most age models for recent marine sediments bring any data into play that might help construct and refine an age model and these data typically augment ¹⁴C and Pb/Cs (notably tephra). This article – which is a case study - focuses on two additional approaches, using the oceanic ¹³C Suess effect and spheroidal carbonaceous fly ash particles (SCP). The section on the Suess effect outlines an approach that will be of interest and use to many in the community, and the approach is novel; the section on SCP is included here as an accessory technique and is of less novelty. I have major concerns over the approach used to determine the age of the core-top.

Specific comments

I wonder whether the title should highlight/reflect the Suess effect approach? I think this is the most significant part of the paper and reference to this in the title would help flag this significance.

Lines 47-48: the ¹⁴C bomb-spike is introduced here as a confounding factor that increases uncertainties but it can provide a useful additional basis for assessing age if sufficient serial samples are available to define the spike.

Line 193: This sharp decline is only present in the final, single, 0.5 cm sample so the sampling resolution here could be problematic. Having higher resolution to define this decline more clearly would make the argument stronger. This issue is exacerbated considering the certain impact of bioturbation in the sediments, and the likely lateral variability in signals generated by bioturbation. Although in lines 321-322 the authors state that there were no visible signs of bioturbation, they acknowledge that bioturbation is common (actually ubiquitous) and that this will likely influence age distributions in the top 10 cm of the core. If there was no bioturbation the core would be laminated. I’m therefore more concerned with this somewhat over-interpreted approach to estimating core-top age than with the age-depth modelling. The authors should either consider strengthening this argument or deleting this section of the MS.

Line 423: The term “ultra-high-resolution” should be reserved for archives that have annual to subannuual resolution.
Citation: https://doi.org/10.5194/egusphere-2023-2845-RC1
- AC1: 'Reply on RC1', Nil Irvali, 11 Apr 2024
  
  We thank Prof. James Scourse for his positive feedback and helpful suggestions. Please find our detailed responses in the attached pdf file.
  
  Citation: https://doi.org/10.5194/egusphere-2023-2845-AC1
RC2:
'Comment on egusphere-2023-2845', Anonymous Referee #2, 26 Feb 2024
Irvali and co-authors present a very interesting study with new ideas for establishing improved chronologies of modern sediments. Such efforts are much needed in paleoceanography because better chronologies for the upper, most recent sediments will allow the comparison of proxy data with instrumental time series and can therefore contribute to improving proxy calibrations and reconstructions.
The new data presented in this study include measurements of stable isotopes on foraminifera to detect the anthropogenic Suess effect and the occurrence of SCPs as additional age controls for the last few centuries.
Although the research idea and the data produced are of high quality, I have some major concerns about the chosen methods, the reporting of the data and structuring of the manuscript.
In the introduction, the authors list various available techniques for dating recent marine sediments, and state that all of them have their own limitations and uncertainties. With so much data available for the studied sediment core, the ideal approach would thus be to combine all information into an integrated optimized age model for the last few centuries. However, the authors choose to create a core chronology based only on stable carbon isotopes measured on one species of foraminifera and their correlation to a model (based on the Suess effect), complemented by radiocarbon dating. The radiocarbon dating uses a reservoir age which is obtained from the abovementioned Suess correlation, so this is not independent either. The available Pb-210 and Cs-137 information is discarded and only included in the discussion, and the new SCP data is not actually used to build the age model but only to confirm the findings based on the isotopes.
On lines 108-109, the authors argue that the Cs-137 concentrations below 4 cm core depth were too low to detect and therefore these data were not used. One could argue that this is a result in itself; the anthropogenic Cs-137 isotope only occurs above 4 cm. I would consider this valuable information for the age model.
The same is true for the radiocarbon age of the top of the core (Table 1, date KIA34242). The authors correctly state that this age indicates the presence of bomb carbon, and therefore younger than ~1957. This is another constraint that can be used to include in the age model. I don’t understand why it is “not included in the age model”. It is not possible to use it as a normal radiocarbon date, but an age constraint in the form of a maximum age should be possible to incorporate in the core chronology.
The interpretation of the Suess effect in the foraminiferal isotope measurements is steered towards the final conclusions of the authors and does discuss other options. Only bulloides, and the combined stack are compared to the model output (Suppl. Table 1), but not the other individual species? Why not?
The curve fits are compared to various depth intervals from 12-0 to 5-0 cm (Lines 203-204), but why stop there? What about the curve fit coefficients at 4-0 cm, or even up to 1-0 cm?? The best fit is found for 7.5- 0 cm with r=0.73. Is this significant compared to the other intervals? For 0-5 cm, the r value is 0.69. Is that difference significant? And what would it be for 0-4 cm etc?
After this simple statement: “7.5 cm must be 1800 AD”, this is taken as fact, without any further discussion or even consideration of uncertainty. In SI Fig. 3 or in Fig. 4, one could imagine a very good fit between the red and blue curves if the red was more compressed and shifted to the right.
The core top age is then determined as 1977, based on the findings above and a very simplified comparison between the upper parts of the isotope model data, and the bulloides measurements (Figure 3). This figure, instead, to me highlights the main difference between the data. One change is extremely abrupt, showing the major changes all in the top samples, while the other is gradual, with a decrease over 200 years. This difference is not sufficiently discussed. Also, this figure illustrates how the biggest change in bulloides is all in the top 1 cm of the core, and the moving average curve is here left out.
If the core top would actually be 1977, where is the rest of the sediment? Multi-cores are known to preserve the sediment-water interface, so what happened to the last 34 years?
From here onwards, this new age model is simply taken as the truth and no uncertainty is reported whatsoever. The term “known-age” is described between quotation marks, but that is not the same as discussing or reporting uncertainties. At some point (Line 345) a ±3-year uncertainty is reported, but it is not clear where this value comes from?
The authors claim to be able to deduce better estimates of the local ΔR values with lower uncertainties. They briefly acknowledge possible bioturbation, but don’t discuss that the new ΔR values are based on the assumption that 0 cm is 1977 and 7.5 cm is 1800. The uncertainties derived from this should be included in the new reservoir age estimates.
Figure 7 shows a combination of some of the data, compared to the old Pb-210 age model. The very short discussion that follows just repeats how the new age model was made but fails to explain why the Pb-210 doesn’t match. What could be reason for the mismatch? And would it not be possible to find a solution that satisfies all data? Perhaps the raw Pb-210 data could be reevaluated, and not just the resulting age-depth model.
The manuscript often lacks a clear distinction between introduction, methods, results, interpretation or discussion. An example is paragraph 3.5 on SCP analyses. It combines most of the above in a single page.
In summary, the presented research idea is very promising, but in the current state, the methods and discussion do not make a convincing case that the new chronology is more reliable than the old one.
Specific comments:
Lines 44 – 86. Several studies on recent marine sediments have used the increase of mercury concentrations as an anthropogenic marker for the last century. This could be added to your list. Example of a study from north of Iceland: https://doi.org/10.1371/journal.pone.0239373

Lines 116-117: bulloides was picked from the 250-300 µm size fraction, while inflata was picked from the 250-350 µm size fraction. Is this a typo, or did you include a 300 µm AND a 350 µm sieve?

Line 155: “We set the starting point in time to 1800,…”. How does the record look before that? Does it still look similar to the measured isotopes?

Lines 191-193 does not give an objective description of the isotope results. Natural variability is described as “over the 10-44 cm core interval” which already includes the interpretation that thereafter, the changes are of anthropogenic origin. Why not explore the option that “natural variability goes to 4 cm? Or 1 cm depth? Why stop at 10?

Lines 210-211: “suggesting 7.5 cm must be 1800 AD”. This is too strong a statement. What about the uncertainty of this method and any critical discussion?

Lines 230-231: “gives us a rough estimate of which curve is most similar to our target curve (i.e., d13CSE_0-200), and overall agrees with our initial finding”. Again, this is simply not critical enough. Would it also have agreed if you had other findings? Probably.

Line 287: “the ΔR in the region is highly variable”. What is meant here are the ΔR values based on the online database of calib.org, which are just a few observations. It is not the same as the actual ΔR values, so this should be clarified.

Line 323: Bioturbation is not limited to the top 10 cm of the core. Every depth level was once the top of the sediment, so bioturbation affects the entire core. The mixing or resulting smoothing of data is then over a ~10-cm window.
Citation: https://doi.org/10.5194/egusphere-2023-2845-RC2
- AC2: 'Reply on RC2', Nil Irvali, 11 Apr 2024
  
  We thank the reviewer for their thoughtful feedback and suggestions. Please find our detailed responses in the attached pdf.
  
  Citation: https://doi.org/10.5194/egusphere-2023-2845-AC2
RC3:
'Comment on egusphere-2023-2845', Anonymous Referee #3, 14 Mar 2024
Geochronology Manuscript egusphere-2023-2845 "Revising chronological uncertainties in marine archives using global anthropogenic signals: a case study" by N. Irvali et al.
This paper proposes a new chronostratigraphic approach that uses the oceanic ¹³C Suess effect and spheroidal carbonaceous particles (SCP) to improve the age models of marine sediment archives that cover the last few hundred to thousand years and could then serve to extend instrumental records back in time.
Extending instrumental records back in time is an important goal since it is the only way to improve our understanding of decadal to multidecadal climate variability and how this variability is currently affected by climate change. Marine sediments are one key climate archive in this respect but uncertainties associated with traditional dating techniques of marine sediment cores (e.g. ¹⁰Pb, ¹⁴C) are too large (up to ±30-50 y) to actually build continuous times series from sediment and instrumental records.
The paper is clearly written and its topic deserves publication. However, a number of points should be improved, as detailed below, before it can be accepted for publication.
Main comments
My first major comment is that there is a relatively large uncertainty with respect to the date of the starting point of the decline in atmospheric d¹³C and with respect to the new core top age. But these uncertainties are both critical in the definition of the final revised age model. It is thus necessary to explore the impact of the uncertainty of these boundaries age on the final age model. This could be done by Monte Carlo or any other technique.

Regarding the starting point of the decline in atmospheric d13C set at 1800, an error bar of ± 40 y seems reasonable given that the industrial revolution is dated between 1760 and 1840, depending on the authors. Note that this uncertainty does actually lead to an uncertainty in the definition of the core top age, which in turn is key in the definition of the new age model using the Bacon age-depth modeling software.
In addition, l. 408, the authors discuss the possible impact of the changes in the subpolar gyre circulation over the 20^th century and productivity decline in this region on the G. bulloides d¹³C. They write “we suspect the uncertainty based on natural climate variability to be minor in our core top age estimate”. In my opinion, in a scientific article, it is required to go one step further than "suspecting". One way to go would be to carry out a sensitivity study to various school cases.
My second major comment concerns the estimate of the “known ages” for depths 2.5 cm, 4 cm and 5.5 cm by reading the corresponding ages from the d¹³C_{SE_50}curve (Fig. 4): doing so the authors assume that the sedimentation rate is constant between 0 and 7.5 cm, which leads to a constant sedimentation rate over the top 7.5 cm of the core by construction (as can be seen on Fig. 5b). This is a strong assumption which affects the final age model. It should be clearly described and its validity should be discussed.

A third important comment concerns the SCP profile : its resolution is too low to allow a verification of the age model obtained using the oceanic ¹³C Suess effect. The published SCP profiles of Rose (2015) seem to indicate that the SCP concentration peaks between 1970 and 1990 in regions adjacent to the core site, which does not match the SCP profile in the studied core plotted vs the revised age model in Fig. 6. This should be discussed. Also, an additional figure in the supplementary material showing the studied core SCP profile superimposed on published SCP profiles of Rose (2015) would be useful.

8 is practically not discussed. What does the revised age model suggest in terms of the relative phasing between the Iceland-Scotland overflow vigor and the Atlantic Multidecadal Variability?

Concerning the assessment of the ∆R to be applied to the radiocarbon dates: why do not the authors extract it from the GLODAP data set? This should provide a nice alternative to the CALIB marine reservoir database. It would be interesting to compare the ∆R currently computed by the authors with the ∆R based on the GLODAP data set. This would provide another estimate of the uncertainty associated with ∆R.

The conditions of validity of the assumption of Transient Steady State should be discussed. For instance, this assumption does not hold in case of changes in ocean circulation.

Line 349-350: the average sedimentation rate over 0-44 cm is not really meaningful since it is larger below 30 cm than over the 7.5-30 cm depth interval, and larger below 7.5 cm than above 7.5 cm (Fig. 5).

More minor comments:
88: “that uses” should be replaced by “that use”

To ease the reading, l. 181 should read “In the North Atlantic, inflata calcifies between 200 and 400 m south of 57°N, and between 100 and 200 m north of 57°N”.

185-189: an additional figure of the d¹³C stack together with the average ¹³C Suess effect change over 0-200 m would be useful in the supplementary material.

195: “even” seems unnecessary.

200: replace “found” by “computed” or “determined”

210: suppress “We show that,”

281: replace “would be” by “is”

301: replace “5 cm” by “5.5 cm”

4 and 5 must contain a typo because they don’t yield an uncertainty of ± 38 y for the weighted mean of ∆R, contrarily to what is indicated in Table 2.

378: replace “confirm” by “confirms”

7: the legend is too small

397-402: this is a repetition of what is written earlier in the article. Repetitions should be avoided.

412: replace “demonstrate” by “illustrate”

439: check the syntax.
Citation: https://doi.org/10.5194/egusphere-2023-2845-RC3
- AC3: 'Reply on RC3', Nil Irvali, 11 Apr 2024
  
  We thank the reviewer for their thoughtful feedback and suggestions. Please find our detailed responses in the attached pdf.
  
  Citation: https://doi.org/10.5194/egusphere-2023-2845-AC3

Nil Irvalı, Ulysses S. Ninnemann, Are Olsen, Neil L. Rose, David J. R. Thornalley, Tor L. Mjell, and François Counillon

Supplement

https://doi.org/10.5194/egusphere-2023-2845-supplement

Nil Irvalı, Ulysses S. Ninnemann, Are Olsen, Neil L. Rose, David J. R. Thornalley, Tor L. Mjell, and François Counillon

Viewed

Total article views: 350 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
247	75	28	350	28	15	14

HTML: 247
PDF: 75
XML: 28
Total: 350
Supplement: 28
BibTeX: 15
EndNote: 14

Views and downloads (calculated since 15 Dec 2023)

Month	HTML	PDF	XML	Total
Dec 2023	72	21	9	102
Jan 2024	53	17	4	74
Feb 2024	35	14	5	54
Mar 2024	34	11	1	46
Apr 2024	53	12	9	74

Cumulative views and downloads (calculated since 15 Dec 2023)

Month	HTML	PDF	XML	Total
Dec 2023	72	21	9	102
Jan 2024	53	17	4	74
Feb 2024	35	14	5	54
Mar 2024	34	11	1	46
Apr 2024	53	12	9	74

Viewed (geographical distribution)

Total article views: 354 (including HTML, PDF, and XML) Thereof 354 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 26 Apr 2024

Short summary

Marine sediments are excellent archives for reconstructing past changes in climate and ocean circulation. Yet, dating uncertainties, particularly during the 20^th century, pose major challenges. Here we propose a novel chronostratigraphic approach that uses anthropogenic signals, such as the oceanic ¹³C Suess effect and spheroidal carbonaceous fly ash particles, to reduce age model uncertainties in high-resolution marine archives over the 20^th century.


Total:	0
HTML:	0
PDF:	0
XML:	0