the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 Jan 2022
14C plateau tuning – A misleading approach or trendsetting tool for marine paleoclimate studies?
Abstract. On the basis of minor time scale adjustments including the synchronization on IntCal20 the Suigetsu-based atmospheric 14C plateau structures are shown to be authentic. Their global significance is demonstrated by the coherence with the tree ring record 10 to 15 cal. ka and by coherent features in the 14C record of Hulu Cave back to 35 cal. ka. The suite of atmospheric structures can be recognized in high-resolution ocean sediment records independent of various processes leading to partial distortion of a sediment record. This provides a unique tool for global stratigraphic correlation and paleoceanographic studies as shown by supplementary figures and tables from 19 cores obtained from key locations in the world ocean.
- Preprint
(816 KB) - Metadata XML
-
Supplement
(3879 KB) - BibTeX
- EndNote
Status: closed
-
CC1: 'Comment on cp-2021-173', Edouard Bard, 20 Mar 2022
14C plateau tuning – A misleading approach for marine paleoclimate studies
Edouard Bard1 and Timothy J. Heaton2
1 CEREGE, Aix-Marseille University, CNRS, IRD, INRAE, Collège de France,
Technopôle de l’Arbois, Aix-en-Provence, France
2 School of Mathematics and Statistics, University of Sheffield, Sheffield S3 7RH, UKSee the attached PDF
-
CC2: 'Reply on CC1', Pieter M. Grootes, 02 Apr 2022
Discussion contribution Climate of the Past.
'Comment on cp-2021-173', by Edouard Bard and Timothy J. Heaton.
14C plateau tuning – A misleading approach for marine paleoclimate studies.
First reply and invitation for further clarification by Pieter M. Grootes
We (Sarnthein and Grootes) welcome the discussion contribution of Bard and Heaton (B&H) to ‘14C Plateau Tuning – A misleading approach or a trendsetting tool for marine paleoclimate studies?’. This continues the evaluation, started with Sarnthein et al 2020, of Plateau Tuning (PT) as a paleoceanographic research tool to evaluate small 14C signals in a patchy, noisy record where traditional methods fail.
We submitted to Climate of the Past (CP) to provide a summary of our PT results for 19 cores and to invite community discussion of the new PT technique and its, for some, sometimes controversial results. After considerable discussion, the paper was published late 2020. B&H submitted comments with serious objections in late January 2020 - yet listed as received by CP on 23 December 2020 - after Sarnthein et al 2020 had been published (Bard and Heaton, CP 2021, p.1701-1725). B&H raised 17 objections spread over two chapters. As pointed out in our 2021 CP comments, it seems ‘Its aim is to demonstrate that Plateau Tuning (PT) is fraught with problems and should not be used’. The present B&H comment appears to follow this line.
B&H start by referring to objections in their long 2021 CP-paper and the detailed discussion of those objections in our comments and their answers. We answered in 2021 in detail B&H objections related to the difficulty of identifying plateaus and to the lack of statistical robustness (in 2.3, 3.2, 3.4, 3.7, and 3.8). B&H selectively grouped comments and ignored several, both in their rebuttal and in the final paper.
Our present paper is the promised update of Sarnthein et al., 2020 that brings together all our data sets using the new Suigetsu time scale of Bronk Ramsey et al 2020, showing the information that may be obtained from ‘difficult’ sediment records by using Plateau tuning. To answer the question in its title we need to consider, in addition to the present B&H comments, some earlier discussion points that B&H failed to address in CP 2021.For a meaningful discussion I invite B&H’s further response to the points listed below.
The difference between PT and IntCal20:
PT aims to extract new information regarding variations in internal ocean dynamics and ocean-atmosphere exchange over last glacial-deglacial-Holocene times from ocean sites that lack clear chronostratigraphic markers for detailed age control, e.g., as listed in ‘Outlook’ of B&H 2021. PT hunts for small 14C signals in a patchy, noisy record that will generally fail statistical tests for robustness. This is even true for the fine structure of the Suigetsu atmospheric 14C record.
IntCal20, by contrast, is a statistically robust tool to translate 14C ages into calendar ages. To achieve robustness, decadal to centennial information has been smoothed, which makes it less suitable for PT, as illustrated by Fig. 2 of the present B&H comment, comparing the Bayesian splines of Suigetsu and IntCal20.PT robustness.
PT robustness
As pointed out earlier: Absence of statistical proof is not proof of absence. The lack of statistical robustness of a single PT 14C record, emphasized in the statistical perspective of B&H 2021, is compensated in PT by much work, that documents consistency of the derived 14C plateau sequence with local sedimentology, stratigraphy, and the multi-parameter sediment record, and with similar plateau sequences developed elsewhere. The probability that a specific 14C fluctuation is caused by noise decreases with each new record in which this fluctuation is found, thus building robustness. The set of 19 records provides a base to check the global consistency.
Incorrect PT assumptions B&H 2021 used in formulating their objections.
Bard is an experienced paleoceanographer and his list of potential problems of sediment records is realistic. Yet, contrary to his writing, these problems have been recognized by Sarnthein and coworkers in developing PT. In our comments 2021 and in the manuscript presently submitted, we pointed out that B&H’s objections were based on their misunderstanding of PT and that several of their supposed ‘assumptions made in PT’ were incorrect.
Physical impossibility of PT plateau schedule
B&H discussed (in 2.3) in great detail the timing of 14C fluctuations in the PT-tuned Suigetsu record. They objected that PT tuning resulted in a highly irregular behaviour of the 14C clock and, when moved from 14C age to 14C concentration Δ14C, in physically unrealistic jumps in atmospheric 14C concentration. B&H repeat this objection in the present comment.
They ignore our comments to 2.3 that the jumps were the result of our use of constant-age plateaus that was dictated by the lack of data sufficient to define a plateau slope. An irregular 14C clock, moreover, is the logical consequence of fluctuating atmospheric 14C concentrations.
Bayesian spline and age plateaus.
B&H made the Bayesian spline plot of Suigetsu Δ14C values to demonstrate the physically unreasonable consequences of PT. Despite our comments, B&H so far do not acknowledge the surprising agreement between their Bayesian spline and the PT record of atmospheric plateaus and jumps, especially after the recent time scale update. Figure 2 of their present comment - unfortunately including a typo in our Table 1; thanks to B&H for finding it – again demonstrates this. Fig. 2 also shows the loss of fine structure in IntCal20.
-
CC2: 'Reply on CC1', Pieter M. Grootes, 02 Apr 2022
-
RC1: 'Comment on cp-2021-173', Anonymous Referee #1, 29 Mar 2022
Sarnthein and Grootes: submitted to Climate of the Past
The format and discussion of this document better resembles a comment than a scientific manuscript, making this very difficult to understand / review. I can only guess that the authors intended for this text to be submitted as a comment instead of a stand-alone manuscript. It is therefore simplest to reject this submission and allow the authors to correct their mistake or ask them to submit a manuscript that at least partially follows the standard formatting of a scientific publication.
A brief listing of the problems with this document
The title of this document suggests the reader will be provided with an expert review of the Plateau tuning technique for identifying the calendar age. However, the Introduction does little to introduce the reader to the relevant argument, instead directing the reader to first fully understand the discussions of several earlier publications. This is a missed opportunity. If the rebuttal to this comment is that the review was already published, then I see no reason why the text provided here should be a stand-alone manuscript. Submit it as a comment to the already peer-reviewed manuscripts.No Results section or Discussion is provided, with the text immediately moving to arguments that I (a putative expert) have difficulty following.
Every figure has a different X-axis range, with some have the calendar age move in different directions. These figures are not professional.
Citation: https://doi.org/10.5194/cp-2021-173-RC1 -
AC1: 'Comment on cp-2021-173', Michael Sarnthein, 08 Apr 2022
Discussion contribution to Climate of the Past.
'Comment on cp-2021-173', by Edouard Bard and Timothy J. Heaton:
"14C plateau tuning – A misleading approach for marine paleoclimate studies."
Reply by Michael Sarnthein
Introduction
Though in disagreement with most arguments and objections recently raised by Bard and Heaton (B&H), we (Grootes and I) see a great chance in taking their concerns seriously, since similar reservations may also come from other colleagues in the science community. The comments below try to counter the issues of B&H with patience one-by-one.
p.1, §1 and §2. B&H refer to pitfalls of plateau tuning (PT) they had already discussed in their CP paper of B&H 2021, though not addressed in the manuscript presently under discussion for CP.
Contrariwise, the CP paper of B&H (2021) has strongly suffered from ignoring most of our detailed discussion remarks we had returned to their 17 objectives 2021 (comments of Sarnthein & Grootes, 2021). In the present comment, B&H may hopefully be able to better recognize our answers to their objections. However, it appears inappropriate now to reiterate our critical remarks on B&H statements of Feb. 2021 but focus on two key objectives, (1) the existence of atm 14C plateaus and (2) their proper reflectance in (19) ocean sediment cores.
p.1, §2. B&H refer to Skinner & Bard (2021) seeing unexplained discrepancies between local MRA estimates based on PT and that of other authors: As long as Skinner & Bard (2021) don't present any objective evidence to decide which actually was the true and which the wrong MRA estimate for a site or for closely neighbored sites under discussion, I prefer to stick to the well-defined estimates deduced by PT.
p.1, §3 and §4. B&H conclude: "Sarnthein & Grootes only consider one narrow aspect of PT – whether long atmospheric 14C-age plateaus exist and can be reliably identified. We are unfortunately unconvinced , , On the basis of the Suigetsu 14C record the authenticity of atmospheric 14C age plateau structures is now reliably confirmed by means of 3, completely independent techniques of signal evaluation (details listed below), therefore plateaus >300 yr long are indeed regarded as robust (shorter ~250 yr is too short to be reasonably defined).
B&H demur other key PT requirements: Whether these atmospheric 14C-age plateaus transfer to the marine environment and, even if they do, whether they can be reliably identified in extremely sparse 14C samples from sediment cores , , ,The transfer of atmospheric plateau signals to marine surface waters and (planktic) sediment records is broadly discussed in our synthesis paper 2020, in Grootes & Sarnthein (Feb. 2021), and now, in a graphic display of 19 sediment records, where a sampling resolution better than 50-150 yr is found adequate to identify the suite of 14C plateaus >250-300 yr long each.
B&H claim that PT was applied to sediment cores for which no independent timescale is available , ,
In contrast to the claim of B&H all results of PT were strictly reconciled (and discussed) with independent 'conventional' stratigraphic tie points of SST and planktic d18O records, and with further age markers such as ash layers for each single core record examined.
What Sarnthein and Grootes claim in their new work
p.1, §5. "We remain unclear how a reader can be confident that the hypothetical 14C-age plateaus have been identified objectively"
Based on both the 14C record of Suigetsu and that of tree rings (10 - ~15 cal ka) the suite of atmospheric plateau structures has been clearly reproduced by means of 3 independent evaluation techniques, hence is regarded robust. Either based on visual inspection or on the 1st derivative of the slope of 14C yr vs. cal yr the plateau boundaries we had inferred merge within <100 yr (Sarnthein et al., 2015) and differ from those based on the (smoothing) Bayesian spline by no more than 50 - 250 yr (Fig. 1), a difference tolerable. Different from the other two techniques, however, the Bayesian spline did not eliminate in advance up to 10 % obvious 14C age outliers that mark most 14C plateaus, hence may diffuse some plateau boundaries.
p.2, §2. This new paper therefore provides limited new insight over their review Sarnthein et al. published 2020.
Indeed, our recent, IntCal20-based revision of the age tie points and MRA for 19 marine data sets did not aim for any new results but just for a more accurate age control, crucial for any proper correlation to other marine and/or non-marine paleoclimate records.
Lack of independence between Lake Suigetsu and Hulu Cave Chronologies
p.2, §3and top of §4. , , , , the updated Suigetsu calendar age chronology , , , , , is not independent of the Hulu Cave 14C record.
Various aspects of the Suigetsu calendar age chronology and their link to that of Hulu Cave and various caveats were broadly discussed in our synthesis 2020, thus won't need to be repeated in this comment.
p.2, §4 – §6. (an admonition concerning our Fig. 4, not "Fig .3") , , , , one must be extremely careful not to overstate one’s confidence that similarities in the two records after tuning provide robust and repeated evidence for the simultaneous presence of 14C-age plateaus
Indeed, our Fig. 4 won't serve for overstating the confidence in the similarity of 14C plateau structures found in Suigetsu and Hulu records, but just highlights the lack of discrepancies. As broadly discussed in 2020, we are fully aware, and more so than other authors, that both the timing of Hulu and the correlated Suigetsu records don't present a purely atmospheric signal but hide away in part major, in part minor potential shifts in the Dead Carbon Fraction (DCF) of the Hulu stalagmite, short-term shifts yet unknown.
Variation in Atmospheric 14C levels
p.2, §7. and p.3, §1-2 and §3. B&H 2022 address a crucial issue in questioning the Suigetsu-based evidence for a long "staircase" of 14C-age plateaus 15 – 30 cal ka per analogy to that found in the tree ring record 0 – ~15 cal. ka, i.e., the problem of reliably separating short-term wiggles (e.g., 100 yr and less) from actual plateau-style trend inversions in D14C extending over about 300 – 1200 yr.
- B&H may have overlooked in our approach that medium-term atmospheric 14C variations, in particular 14C jumps are clearly identified and reproduced with just a minor degree of subjectivity (chosen width of Kernel window) by means of the 1st derivative of 14C yr vs cal yr we employ.
- This approach that helps us "to see the wood for the trees", a sincere problem to some colleagues, a problem also reflected in B&H Fig. 1. PT does not intend to compete with IntCal in creating a novel age calibration program but just aims for a centennial-scale raster of atmospheric 14C structures suitable for global age correlation.
- Interesting to note that the length of 14C plateaus and the extent of 14C jumps have changed significantly over different regimes of glacial-to-deglacial climate: Plateaus up to 1000 yr long covered ~80% of the peak glacial and early deglacial period 30–14 cal. ka (as outlined by B&H 2021). From ~14–8.6 cal ka, the plateau length dropped to several hundred years (our Fig. 3 and Sarnthein & Werner, 2018). Later on, Holocene plateaus rarely exceeded >250 yr, except for a single case near 2.5 cal ka ("Hallstatt" plateau).
- The potential origin of the variable length of 14C plateau structures was broadly discussed by Sarnthein et al. (2020) in the context of changes in oceanography and the global carbon cycle, that were more prominent prior to 14 cal. ka than during the Holocene.
Updating of hypothesized 14C-age plateaus
p.4, §4. B&H found parts of Fig. 1 incompatible with data listed in Table 1.
Indeed, the base level of plateau 7 in Fig. 1 had unfortunately suffered from a plotting error and had to be lifted from 480 to 494‰ D14C. Thanks to the advice of B&H the compatibility of plateau 7 has now gained a lot. Conversely, Table 1 rightly puts the end of plateau 10a to 25.96 cal. ka as plotted in Fig. 1, consistent with the record of the Bayesian spline, different from an age of 25.65 cal. ka suggested by B&H.
p.4, §5 and §6 and p.5, Fig. 2b. B&H claim that the suite of 14C-age plateaus was physically unreasonable, especially, the suite of massive and instantaneous D14C jumps suggested by means of visual inspection.
Conversely, the overwhelming consistency of trends inferred for the Suigetsu atmospheric 14C record on the basis of three independent evaluation techniques (our Fig. 1) provides strong support in favor of the authenticity of the "rung ladder" of 14C plateau and jump structures over peak glacial to deglacial times.
This authenticity of 14C plateau structures is not impaired by a comparison with the IntCal20 record (B&H, Fig. 2b), where the structures are largely lost for the following reasons:
The IntCal20 record aims for a maximum in robustness of 14C-age calibration being based on averaging, hence smoothing a broad variety of the different terrestrial and marine 14C records employed, many of them presenting an only indirect record of atmospheric D14C where unknown changes in MRA and DCF are ignored. Inversely, the purely atmospheric Suigetsu record is less well resolved, only indirectly dated, and less well age calibrated than IntCal20, but capable to display the primary suite of structures preserved in the raw 14C record, hence less suited for any fine-scale age calibration at estimates <50–100 yr.
Transferal of atmospheric 14C variations to the ocean
p.5, §3. B&H dispute the reproducibility of plateau identification in marine sediments. s
We are pleased to give once more some guidance already published (but ignored by B&H).
The objectivity of identification of a plateau suite is regarded robust as it is supported by (1) the use of the 1st derivative in all 14C records to separate small-scale noise (extending over less than ~100 yr) from major long-term shifts (>300 yr) in the 14C-age trend, (2) a strict alignment of the age scale derived by PT with that of independent but paired conventional age tie points of d18O, SST, and other sediment records such as ash layers, (3) a frank discussion of alternative plateau alignments where questionable, (4) a detailed discussion of each PT-based time scale with scales published for neighbor sites in an ocean basin (as being displayed in the primary papers on PT records published).
B&H demur the sparsity of marine 14C records suited for PT because of short-term changes in sedimentation rate and MRA complicating the practice of PT.
This problem is linked to the costly efforts of high-resolution PT for sediment cores with strictly hemipelagic bulk sedimentation rates >10–20 cm/ky as suggested by conventional means of stratigraphy, excluding cores with turbidites. Though surprises in age assignment are found inevitable and only resolved by means of the package of tests listed above.
p.5, last § and p.6, §1. B&H question the ground-truthing of short-term changes of sedimentation rate and hiatuses by independent lines of evidence.
Different from this suspicion hiatuses and other major changes in sedimentation rate were independently confirmed in most PT-dated core sections by means of visual inspection and high-resolution photography of sediment sections and studies of the sediment composition.
Ground-truthing of 14C plateaus is also achieved by (1) plateaus lengths exceeding >10–20 cm, in part >50 cm thick, densely sampled hemipelagic sediment sections, (2) by characteristic internal structures in the suite of atmospheric plateaus 27-10 cal ka clearly reproduced in a sediment section, (3) by a match with conventional stratigraphic tie points such as DO event 1.
p.6, §2. B&H evoke the potential signal alteration by bioturbation and other sediment mixing processes.
A detailed response is given in our recent article under discussion (L.149–160).
p.6, §3-§5. Minor issues in the critique uttered by B&H
p.6, §3. B&H bemoan changes in the plateau assignment of 14C plateaus for four Pacific records, results finally stored under Küssner et al. (2020) in the PANGAEA database.
Such changes between the initial submission of a manuscript and its last state are common and a legitimate result of scientific discussions amongst different coauthors on the best-possible plateau assignment, the final results of which were included into both the PANGAEA database (2020) and the paper of Küssner et al., 2020 (though unfortunately rejected). Likewise, subsequent comments of coauthors cited by B&H present a legitimate private, though scientifically untested opinion of these authors.
p.6, §4. B&H contest the direct match of marine 14C records to those of the atmosphere.
MRA and the quality of 14C signal transfer depend on the time lags between surface ocean (habitat of planktic foraminifers, in particular of near-surface dwellers) and atmosphere, values that depend on regional mixing and lateral advection of surface waters. Only a minor portion of the time lags can be covered by simple box models listed by B&H. In case the time lag between 14C signals in local surface waters and global atmosphere may be as large as 20 years and more, the objection to PT was hardly relevant, since the error range of PT-based age tie points is far larger (up to 50-100 yr). Grootes & Sarnthein displayed a more extensive reasoning in their contribution of the discussion of the paper of B&H (2021).
p.6, §5. B&H postulate that the benthic 14C record of South China Sea site GIK17940 should actually resemble the 14C plateau structures, since the water column has been largely mixed down to 2 km depth.
In part this postulate may indeed be realistic (Fig. S6), far more than B&H may have recognized when inspecting Fig. S6. A perfect plateau structure, however, fails because of the insufficient sampling resolution of benthic 14C ages and an ocean mixing intensity that still was somewhat erratic over the long time range studied both in terms of intensity and potential changes of ocean water masses admixed.
Summary
By means of their critical remarks B&H provided a great and highly appreciated opportunity to explain in more detail a number of ongoing questions on the reasoning of PT, in our previous explanations perhaps not sufficiently spelled out and/or specified. One may now expect that a number of readers will try to reproduce more easily the solid base of PT and will employ it as a trendsetting tool for their marine paleoclimate studies.
References
Bard, E. and Heaton, T.J.: On the tuning of plateaus in atmospheric and oceanic 14C records to derive calendar chronologies of deep-sea cores and records of 14C marine reservoir age changes. Climate of the Past, 17, 1701–1725. https://doi.org/10.5194/cp-17-1701-2021
Grootes, P.M. and Sarnthein, M.: Community comment on “On the tuning of plateaus in atmospheric and oceanic C records to derive calendar chronologies of deep-sea cores and records of C marine reservoir age changes” by Edouard Bard and Timothy J. Heaton, Clim. Past Discuss., https://doi.org/10.5194/cp-2020-164-CC2, 2021.
Küssner, K., Sarnthein, M., Michel, E., Mollenhauer, G., Siani,s G., and Tiedemann, R.: Glacial-to-deglacial reservoir ages of surface waters in the southern South Pacific. PANGAEA, https://doi.org/10.1594/PANGAEA.922671, 2020.
Küssner, K., Sarnthein, M., Michel, E., Mollenhauer, G., Siani G., and Tiedemann, R.: Glacial-to-deglacial reservoir ages of surface waters in the southern South Pacific. Paleoceanography and Paleoclimate, 47 ms-pp. (2020, rejected)
Sarnthein, M. and Grootes, P. M.: Community comment on “On the tuning of plateaus in atmospheric and oceanic C records to derive calendar chronologies of deep-sea cores and records of C marine reservoir age changes” by Edouard Bard and Timothy J. Heaton, Clim. Past Discuss., https://doi.org/10.5194/cp-2020-164-CC1, 2021.
Sarnthein, M., Küssner, K., Grootes, P.M., Ausin, B., Eglinton, T., Muglia, J., Muscheler, R., Schlolaut, G. Plateaus and jumps in the atmospheric radiocarbon record – Potential origin and value as global age markers for glacial-to-deglacial paleoceanography, a synthesis. Climate of the Past, 16, 2547–2571, doi: 10.5194/cp-16-2547-2020, 2020.
Sarnthein, M. and Werner, K.: Early Holocene planktic foraminifers record species specific 14C reservoir ages in Arctic Gateway, Marine Micropaleontology, 135, 45-55.
DOI:10.1016/j.marmicro.2017.07.002, 2018.
Skinner, L.C. & Bard, E.: Radiocarbon as a dating tool in paleoceanography, Review in Geophysics, 60, e2020RG000720, doi.org/10.1029/2020RG000720, 2022.
Citation: https://doi.org/10.5194/cp-2021-173-AC1 -
RC2: 'Review Comment on cp-2021-173', Anonymous Referee #2, 21 Apr 2022
In this manuscript, the authors seek to establish the existence and timing of ‘plateaus’ in the atmospheric radiocarbon record, and to demonstrate that these are also present in marine records from around the world. On this basis, the authors seek to argue that radiocarbon plateaus identified in marine records can be stratigraphically aligned to correlative plateaus identified in the atmospheric record, allowing calendar ages to be transferred to the marine records (and therefore allowing for ‘marine reservoir age’ offsets to be determined). This method of chronostratigraphic alignment has been termed ‘plateau tuning’ (PT).
This manuscript is quite unusual, as it does not appear to advance any new observations/data, arguments, models or insights. Some adjustments are made (again) to the proposed timing of plateaus identified in the atmospheric radiocarbon record, but this does not really make any difference to what has been proposed by the authors in several papers since 2007. A recent ‘review’ of the ‘PT method’ and its results was published by the authors just last year in this same journal. Primarily, it seems, the manuscript seeks to publish a rebuttal of a prior piece of work produced by Bard & Heaton (B&H) that was also reviewed and published in Climate of the Past last year. The latter was also accompanied by several pages of commentary by Sarnthein and Grootes, which was in turn responded to by Bard and Heaton over the course of the discussion phase of the manuscript.
Unfortunately, I find it impossible to recommend that this manuscript be accepted for publication. There are three main reasons for this: 1) it does not appear to present an original piece of research, and insofar as it presents adjustments, they are not important enough for publication on their own merit; 2) its arguments against B&H are not coherent (regardless of whether or to B&H are correct); 3) the vast bulk of figures and tables referred to in the manuscript are included in a ‘supplement’ that has not actually been produced/included. On the latter point, the promise of a compilation of all the available PT data in useful tables would have been at least one welcome contribution: but it turns out that the intention of the authors was to include ~20 disparate data tables that are already available on PANGEA and that are not at all useful in reproducing the PT data that have been published to date by the authors (it took me days to do this, and the results are not the same as what the authors have published in many cases, which is both worrying and annoying). The fact that the PT data (and associated MRA etc.) that have been produced by the authors over several years, and presented in a series of ‘global synopsis’ papers, cannot be easily reproduced by others using the multitude of available data tables, is particularly worrisome. The same can be said for the fact that only one (?) PT study exists that does not include the authors of this study (the champions of the PT approach). Incidentally, this might already answer the question of whether or not it is a ‘trend setting’ tool.
With regard to the second point raised above, the authors state that they reject the arguments of B&H based on the basis of how plateaus are identified (i.e. as ‘sequences’, like a sort of Morse code), and on the basis that B&H use a 1998 box model to support their arguments. Regardless of the validity of B&H’s remarks, I don’t see how either of these points represent a coherent basis on which to reject a criticism of the PT method, where that criticism is founded in large part on the proposed difficulty of objectively identifying plateaus (let alone sequences of plateaus) in a noisy marine radiocarbon record whose offset from atmospheric radiocarbon varies over time, as well as the proposal that sedimentary processes (such as simple - and highly likely - sedimentation rate changes during periods such as Heinrich Stadial 1, or the Younger Dryas) can also produce ‘plateaus’ in the 14Cage-depth domain, without these being causally linked to atmospheric radiocarbon variability. In addition, the claim that a 1998 box-model is somehow incorrect because of its vintage seems to miss the point: the key purpose of deploying such a model is surely to illustrate in a very simple way how the phasing and amplitude-attenuation of an input signal will be altered (filtered) in the ocean, depending on the timescale on which the signal can be communicated to the ocean, and the frequency/duration of the signal variability. You can do this with a very complex biogeochemical coupled ocean-atmosphere numerical model if you like, but if it did not show a simple phase-attenuation relationship like the box model, it would mean that the complex model had a problem! In fact, by playing around with numerical model outputs it can be shown that they do show the same principles as a 2 box-model, and that should not be surprising, as it is an expression of a simple and fundamental physical principle: parts of the ocean that have small MRA offsets (such as the tropical ocean, MRA ~400 14Cyears) can respond quickly and can pick up shorter fluctuations from the atmosphere, whereas parts of the ocean that have large MRA offsets will take longer to pick up the atmospheric signal (since a larger MRA means that the isotopic exchange timescale for that water is longer) and will pick up a smoothed and lagged response. The limits of applicability of the PT method could readily be analysed and qualified in such a theoretical context, but the authors don’t do this unfortunately.
Ultimately, the manuscript sets out to answer the question posed in the title: “is the ‘plateau tuning’ (PT) approach a misleading approach or a trend-setting tool”? I would note that, at worst, PT could be both misleading and trend-setting, and my major concern is that the authors clearly wish for it to be the latter, but have not really (either in the present manuscript, or over the course of several publications that appear to present the same datasets repeatedly) demonstrated that the PT approach is indeed viable, either in theory or in practice. As suggested above, this is not to say that some sort of defence cannot be made, in theory at least. But the authors (still) have not managed to do this. My own view is that the chronostratigraphic principles that the authors wish to apply are not completely crazy: yes, the atmospheric radiocarbon record has ‘wiggles’ and these would be transferred to other reservoirs that exchange CO2 with the atmosphere rapidly enough to pick them up. However, the conditions under which these wiggles can be recorded in other reservoirs, such as the ocean, and the biases (in amplitude and phasing especially) that will inevitably and predictably arise (even prior to the complications of sedimentation changes, bioturbation, sampling/analytical noise etc.) need to be accepted and addressed by the authors at some point if this debate is to move in a useful direction. I can think of a variety of ways to test the PT method in theory (using models), and in practice using data, and I wonder why the authors have never done something similar. If a scientific study that achieved such goals was produced, it would be a welcome and useful addition to the literature (as B&H has proven to be, insofar as it stimulates critical thinking). Such a study would best come from the authors of the present study, who appear to be the main (if not the only?) champions of the PT method; however, this is not what the current manuscript provides.
Citation: https://doi.org/10.5194/cp-2021-173-RC2 -
AC2: 'Comment on cp-2021-173_Response to Review#1', Michael Sarnthein, 26 Apr 2022
The comment was uploaded in the form of a supplement: https://cp.copernicus.org/preprints/cp-2021-173/cp-2021-173-AC2-supplement.pdf
-
AC3: 'Comment on cp-2021-173 _ Response to Review #2', Michael Sarnthein, 26 Apr 2022
The comment was uploaded in the form of a supplement: https://cp.copernicus.org/preprints/cp-2021-173/cp-2021-173-AC3-supplement.pdf
Status: closed
-
CC1: 'Comment on cp-2021-173', Edouard Bard, 20 Mar 2022
14C plateau tuning – A misleading approach for marine paleoclimate studies
Edouard Bard1 and Timothy J. Heaton2
1 CEREGE, Aix-Marseille University, CNRS, IRD, INRAE, Collège de France,
Technopôle de l’Arbois, Aix-en-Provence, France
2 School of Mathematics and Statistics, University of Sheffield, Sheffield S3 7RH, UKSee the attached PDF
-
CC2: 'Reply on CC1', Pieter M. Grootes, 02 Apr 2022
Discussion contribution Climate of the Past.
'Comment on cp-2021-173', by Edouard Bard and Timothy J. Heaton.
14C plateau tuning – A misleading approach for marine paleoclimate studies.
First reply and invitation for further clarification by Pieter M. Grootes
We (Sarnthein and Grootes) welcome the discussion contribution of Bard and Heaton (B&H) to ‘14C Plateau Tuning – A misleading approach or a trendsetting tool for marine paleoclimate studies?’. This continues the evaluation, started with Sarnthein et al 2020, of Plateau Tuning (PT) as a paleoceanographic research tool to evaluate small 14C signals in a patchy, noisy record where traditional methods fail.
We submitted to Climate of the Past (CP) to provide a summary of our PT results for 19 cores and to invite community discussion of the new PT technique and its, for some, sometimes controversial results. After considerable discussion, the paper was published late 2020. B&H submitted comments with serious objections in late January 2020 - yet listed as received by CP on 23 December 2020 - after Sarnthein et al 2020 had been published (Bard and Heaton, CP 2021, p.1701-1725). B&H raised 17 objections spread over two chapters. As pointed out in our 2021 CP comments, it seems ‘Its aim is to demonstrate that Plateau Tuning (PT) is fraught with problems and should not be used’. The present B&H comment appears to follow this line.
B&H start by referring to objections in their long 2021 CP-paper and the detailed discussion of those objections in our comments and their answers. We answered in 2021 in detail B&H objections related to the difficulty of identifying plateaus and to the lack of statistical robustness (in 2.3, 3.2, 3.4, 3.7, and 3.8). B&H selectively grouped comments and ignored several, both in their rebuttal and in the final paper.
Our present paper is the promised update of Sarnthein et al., 2020 that brings together all our data sets using the new Suigetsu time scale of Bronk Ramsey et al 2020, showing the information that may be obtained from ‘difficult’ sediment records by using Plateau tuning. To answer the question in its title we need to consider, in addition to the present B&H comments, some earlier discussion points that B&H failed to address in CP 2021.For a meaningful discussion I invite B&H’s further response to the points listed below.
The difference between PT and IntCal20:
PT aims to extract new information regarding variations in internal ocean dynamics and ocean-atmosphere exchange over last glacial-deglacial-Holocene times from ocean sites that lack clear chronostratigraphic markers for detailed age control, e.g., as listed in ‘Outlook’ of B&H 2021. PT hunts for small 14C signals in a patchy, noisy record that will generally fail statistical tests for robustness. This is even true for the fine structure of the Suigetsu atmospheric 14C record.
IntCal20, by contrast, is a statistically robust tool to translate 14C ages into calendar ages. To achieve robustness, decadal to centennial information has been smoothed, which makes it less suitable for PT, as illustrated by Fig. 2 of the present B&H comment, comparing the Bayesian splines of Suigetsu and IntCal20.PT robustness.
PT robustness
As pointed out earlier: Absence of statistical proof is not proof of absence. The lack of statistical robustness of a single PT 14C record, emphasized in the statistical perspective of B&H 2021, is compensated in PT by much work, that documents consistency of the derived 14C plateau sequence with local sedimentology, stratigraphy, and the multi-parameter sediment record, and with similar plateau sequences developed elsewhere. The probability that a specific 14C fluctuation is caused by noise decreases with each new record in which this fluctuation is found, thus building robustness. The set of 19 records provides a base to check the global consistency.
Incorrect PT assumptions B&H 2021 used in formulating their objections.
Bard is an experienced paleoceanographer and his list of potential problems of sediment records is realistic. Yet, contrary to his writing, these problems have been recognized by Sarnthein and coworkers in developing PT. In our comments 2021 and in the manuscript presently submitted, we pointed out that B&H’s objections were based on their misunderstanding of PT and that several of their supposed ‘assumptions made in PT’ were incorrect.
Physical impossibility of PT plateau schedule
B&H discussed (in 2.3) in great detail the timing of 14C fluctuations in the PT-tuned Suigetsu record. They objected that PT tuning resulted in a highly irregular behaviour of the 14C clock and, when moved from 14C age to 14C concentration Δ14C, in physically unrealistic jumps in atmospheric 14C concentration. B&H repeat this objection in the present comment.
They ignore our comments to 2.3 that the jumps were the result of our use of constant-age plateaus that was dictated by the lack of data sufficient to define a plateau slope. An irregular 14C clock, moreover, is the logical consequence of fluctuating atmospheric 14C concentrations.
Bayesian spline and age plateaus.
B&H made the Bayesian spline plot of Suigetsu Δ14C values to demonstrate the physically unreasonable consequences of PT. Despite our comments, B&H so far do not acknowledge the surprising agreement between their Bayesian spline and the PT record of atmospheric plateaus and jumps, especially after the recent time scale update. Figure 2 of their present comment - unfortunately including a typo in our Table 1; thanks to B&H for finding it – again demonstrates this. Fig. 2 also shows the loss of fine structure in IntCal20.
-
CC2: 'Reply on CC1', Pieter M. Grootes, 02 Apr 2022
-
RC1: 'Comment on cp-2021-173', Anonymous Referee #1, 29 Mar 2022
Sarnthein and Grootes: submitted to Climate of the Past
The format and discussion of this document better resembles a comment than a scientific manuscript, making this very difficult to understand / review. I can only guess that the authors intended for this text to be submitted as a comment instead of a stand-alone manuscript. It is therefore simplest to reject this submission and allow the authors to correct their mistake or ask them to submit a manuscript that at least partially follows the standard formatting of a scientific publication.
A brief listing of the problems with this document
The title of this document suggests the reader will be provided with an expert review of the Plateau tuning technique for identifying the calendar age. However, the Introduction does little to introduce the reader to the relevant argument, instead directing the reader to first fully understand the discussions of several earlier publications. This is a missed opportunity. If the rebuttal to this comment is that the review was already published, then I see no reason why the text provided here should be a stand-alone manuscript. Submit it as a comment to the already peer-reviewed manuscripts.No Results section or Discussion is provided, with the text immediately moving to arguments that I (a putative expert) have difficulty following.
Every figure has a different X-axis range, with some have the calendar age move in different directions. These figures are not professional.
Citation: https://doi.org/10.5194/cp-2021-173-RC1 -
AC1: 'Comment on cp-2021-173', Michael Sarnthein, 08 Apr 2022
Discussion contribution to Climate of the Past.
'Comment on cp-2021-173', by Edouard Bard and Timothy J. Heaton:
"14C plateau tuning – A misleading approach for marine paleoclimate studies."
Reply by Michael Sarnthein
Introduction
Though in disagreement with most arguments and objections recently raised by Bard and Heaton (B&H), we (Grootes and I) see a great chance in taking their concerns seriously, since similar reservations may also come from other colleagues in the science community. The comments below try to counter the issues of B&H with patience one-by-one.
p.1, §1 and §2. B&H refer to pitfalls of plateau tuning (PT) they had already discussed in their CP paper of B&H 2021, though not addressed in the manuscript presently under discussion for CP.
Contrariwise, the CP paper of B&H (2021) has strongly suffered from ignoring most of our detailed discussion remarks we had returned to their 17 objectives 2021 (comments of Sarnthein & Grootes, 2021). In the present comment, B&H may hopefully be able to better recognize our answers to their objections. However, it appears inappropriate now to reiterate our critical remarks on B&H statements of Feb. 2021 but focus on two key objectives, (1) the existence of atm 14C plateaus and (2) their proper reflectance in (19) ocean sediment cores.
p.1, §2. B&H refer to Skinner & Bard (2021) seeing unexplained discrepancies between local MRA estimates based on PT and that of other authors: As long as Skinner & Bard (2021) don't present any objective evidence to decide which actually was the true and which the wrong MRA estimate for a site or for closely neighbored sites under discussion, I prefer to stick to the well-defined estimates deduced by PT.
p.1, §3 and §4. B&H conclude: "Sarnthein & Grootes only consider one narrow aspect of PT – whether long atmospheric 14C-age plateaus exist and can be reliably identified. We are unfortunately unconvinced , , On the basis of the Suigetsu 14C record the authenticity of atmospheric 14C age plateau structures is now reliably confirmed by means of 3, completely independent techniques of signal evaluation (details listed below), therefore plateaus >300 yr long are indeed regarded as robust (shorter ~250 yr is too short to be reasonably defined).
B&H demur other key PT requirements: Whether these atmospheric 14C-age plateaus transfer to the marine environment and, even if they do, whether they can be reliably identified in extremely sparse 14C samples from sediment cores , , ,The transfer of atmospheric plateau signals to marine surface waters and (planktic) sediment records is broadly discussed in our synthesis paper 2020, in Grootes & Sarnthein (Feb. 2021), and now, in a graphic display of 19 sediment records, where a sampling resolution better than 50-150 yr is found adequate to identify the suite of 14C plateaus >250-300 yr long each.
B&H claim that PT was applied to sediment cores for which no independent timescale is available , ,
In contrast to the claim of B&H all results of PT were strictly reconciled (and discussed) with independent 'conventional' stratigraphic tie points of SST and planktic d18O records, and with further age markers such as ash layers for each single core record examined.
What Sarnthein and Grootes claim in their new work
p.1, §5. "We remain unclear how a reader can be confident that the hypothetical 14C-age plateaus have been identified objectively"
Based on both the 14C record of Suigetsu and that of tree rings (10 - ~15 cal ka) the suite of atmospheric plateau structures has been clearly reproduced by means of 3 independent evaluation techniques, hence is regarded robust. Either based on visual inspection or on the 1st derivative of the slope of 14C yr vs. cal yr the plateau boundaries we had inferred merge within <100 yr (Sarnthein et al., 2015) and differ from those based on the (smoothing) Bayesian spline by no more than 50 - 250 yr (Fig. 1), a difference tolerable. Different from the other two techniques, however, the Bayesian spline did not eliminate in advance up to 10 % obvious 14C age outliers that mark most 14C plateaus, hence may diffuse some plateau boundaries.
p.2, §2. This new paper therefore provides limited new insight over their review Sarnthein et al. published 2020.
Indeed, our recent, IntCal20-based revision of the age tie points and MRA for 19 marine data sets did not aim for any new results but just for a more accurate age control, crucial for any proper correlation to other marine and/or non-marine paleoclimate records.
Lack of independence between Lake Suigetsu and Hulu Cave Chronologies
p.2, §3and top of §4. , , , , the updated Suigetsu calendar age chronology , , , , , is not independent of the Hulu Cave 14C record.
Various aspects of the Suigetsu calendar age chronology and their link to that of Hulu Cave and various caveats were broadly discussed in our synthesis 2020, thus won't need to be repeated in this comment.
p.2, §4 – §6. (an admonition concerning our Fig. 4, not "Fig .3") , , , , one must be extremely careful not to overstate one’s confidence that similarities in the two records after tuning provide robust and repeated evidence for the simultaneous presence of 14C-age plateaus
Indeed, our Fig. 4 won't serve for overstating the confidence in the similarity of 14C plateau structures found in Suigetsu and Hulu records, but just highlights the lack of discrepancies. As broadly discussed in 2020, we are fully aware, and more so than other authors, that both the timing of Hulu and the correlated Suigetsu records don't present a purely atmospheric signal but hide away in part major, in part minor potential shifts in the Dead Carbon Fraction (DCF) of the Hulu stalagmite, short-term shifts yet unknown.
Variation in Atmospheric 14C levels
p.2, §7. and p.3, §1-2 and §3. B&H 2022 address a crucial issue in questioning the Suigetsu-based evidence for a long "staircase" of 14C-age plateaus 15 – 30 cal ka per analogy to that found in the tree ring record 0 – ~15 cal. ka, i.e., the problem of reliably separating short-term wiggles (e.g., 100 yr and less) from actual plateau-style trend inversions in D14C extending over about 300 – 1200 yr.
- B&H may have overlooked in our approach that medium-term atmospheric 14C variations, in particular 14C jumps are clearly identified and reproduced with just a minor degree of subjectivity (chosen width of Kernel window) by means of the 1st derivative of 14C yr vs cal yr we employ.
- This approach that helps us "to see the wood for the trees", a sincere problem to some colleagues, a problem also reflected in B&H Fig. 1. PT does not intend to compete with IntCal in creating a novel age calibration program but just aims for a centennial-scale raster of atmospheric 14C structures suitable for global age correlation.
- Interesting to note that the length of 14C plateaus and the extent of 14C jumps have changed significantly over different regimes of glacial-to-deglacial climate: Plateaus up to 1000 yr long covered ~80% of the peak glacial and early deglacial period 30–14 cal. ka (as outlined by B&H 2021). From ~14–8.6 cal ka, the plateau length dropped to several hundred years (our Fig. 3 and Sarnthein & Werner, 2018). Later on, Holocene plateaus rarely exceeded >250 yr, except for a single case near 2.5 cal ka ("Hallstatt" plateau).
- The potential origin of the variable length of 14C plateau structures was broadly discussed by Sarnthein et al. (2020) in the context of changes in oceanography and the global carbon cycle, that were more prominent prior to 14 cal. ka than during the Holocene.
Updating of hypothesized 14C-age plateaus
p.4, §4. B&H found parts of Fig. 1 incompatible with data listed in Table 1.
Indeed, the base level of plateau 7 in Fig. 1 had unfortunately suffered from a plotting error and had to be lifted from 480 to 494‰ D14C. Thanks to the advice of B&H the compatibility of plateau 7 has now gained a lot. Conversely, Table 1 rightly puts the end of plateau 10a to 25.96 cal. ka as plotted in Fig. 1, consistent with the record of the Bayesian spline, different from an age of 25.65 cal. ka suggested by B&H.
p.4, §5 and §6 and p.5, Fig. 2b. B&H claim that the suite of 14C-age plateaus was physically unreasonable, especially, the suite of massive and instantaneous D14C jumps suggested by means of visual inspection.
Conversely, the overwhelming consistency of trends inferred for the Suigetsu atmospheric 14C record on the basis of three independent evaluation techniques (our Fig. 1) provides strong support in favor of the authenticity of the "rung ladder" of 14C plateau and jump structures over peak glacial to deglacial times.
This authenticity of 14C plateau structures is not impaired by a comparison with the IntCal20 record (B&H, Fig. 2b), where the structures are largely lost for the following reasons:
The IntCal20 record aims for a maximum in robustness of 14C-age calibration being based on averaging, hence smoothing a broad variety of the different terrestrial and marine 14C records employed, many of them presenting an only indirect record of atmospheric D14C where unknown changes in MRA and DCF are ignored. Inversely, the purely atmospheric Suigetsu record is less well resolved, only indirectly dated, and less well age calibrated than IntCal20, but capable to display the primary suite of structures preserved in the raw 14C record, hence less suited for any fine-scale age calibration at estimates <50–100 yr.
Transferal of atmospheric 14C variations to the ocean
p.5, §3. B&H dispute the reproducibility of plateau identification in marine sediments. s
We are pleased to give once more some guidance already published (but ignored by B&H).
The objectivity of identification of a plateau suite is regarded robust as it is supported by (1) the use of the 1st derivative in all 14C records to separate small-scale noise (extending over less than ~100 yr) from major long-term shifts (>300 yr) in the 14C-age trend, (2) a strict alignment of the age scale derived by PT with that of independent but paired conventional age tie points of d18O, SST, and other sediment records such as ash layers, (3) a frank discussion of alternative plateau alignments where questionable, (4) a detailed discussion of each PT-based time scale with scales published for neighbor sites in an ocean basin (as being displayed in the primary papers on PT records published).
B&H demur the sparsity of marine 14C records suited for PT because of short-term changes in sedimentation rate and MRA complicating the practice of PT.
This problem is linked to the costly efforts of high-resolution PT for sediment cores with strictly hemipelagic bulk sedimentation rates >10–20 cm/ky as suggested by conventional means of stratigraphy, excluding cores with turbidites. Though surprises in age assignment are found inevitable and only resolved by means of the package of tests listed above.
p.5, last § and p.6, §1. B&H question the ground-truthing of short-term changes of sedimentation rate and hiatuses by independent lines of evidence.
Different from this suspicion hiatuses and other major changes in sedimentation rate were independently confirmed in most PT-dated core sections by means of visual inspection and high-resolution photography of sediment sections and studies of the sediment composition.
Ground-truthing of 14C plateaus is also achieved by (1) plateaus lengths exceeding >10–20 cm, in part >50 cm thick, densely sampled hemipelagic sediment sections, (2) by characteristic internal structures in the suite of atmospheric plateaus 27-10 cal ka clearly reproduced in a sediment section, (3) by a match with conventional stratigraphic tie points such as DO event 1.
p.6, §2. B&H evoke the potential signal alteration by bioturbation and other sediment mixing processes.
A detailed response is given in our recent article under discussion (L.149–160).
p.6, §3-§5. Minor issues in the critique uttered by B&H
p.6, §3. B&H bemoan changes in the plateau assignment of 14C plateaus for four Pacific records, results finally stored under Küssner et al. (2020) in the PANGAEA database.
Such changes between the initial submission of a manuscript and its last state are common and a legitimate result of scientific discussions amongst different coauthors on the best-possible plateau assignment, the final results of which were included into both the PANGAEA database (2020) and the paper of Küssner et al., 2020 (though unfortunately rejected). Likewise, subsequent comments of coauthors cited by B&H present a legitimate private, though scientifically untested opinion of these authors.
p.6, §4. B&H contest the direct match of marine 14C records to those of the atmosphere.
MRA and the quality of 14C signal transfer depend on the time lags between surface ocean (habitat of planktic foraminifers, in particular of near-surface dwellers) and atmosphere, values that depend on regional mixing and lateral advection of surface waters. Only a minor portion of the time lags can be covered by simple box models listed by B&H. In case the time lag between 14C signals in local surface waters and global atmosphere may be as large as 20 years and more, the objection to PT was hardly relevant, since the error range of PT-based age tie points is far larger (up to 50-100 yr). Grootes & Sarnthein displayed a more extensive reasoning in their contribution of the discussion of the paper of B&H (2021).
p.6, §5. B&H postulate that the benthic 14C record of South China Sea site GIK17940 should actually resemble the 14C plateau structures, since the water column has been largely mixed down to 2 km depth.
In part this postulate may indeed be realistic (Fig. S6), far more than B&H may have recognized when inspecting Fig. S6. A perfect plateau structure, however, fails because of the insufficient sampling resolution of benthic 14C ages and an ocean mixing intensity that still was somewhat erratic over the long time range studied both in terms of intensity and potential changes of ocean water masses admixed.
Summary
By means of their critical remarks B&H provided a great and highly appreciated opportunity to explain in more detail a number of ongoing questions on the reasoning of PT, in our previous explanations perhaps not sufficiently spelled out and/or specified. One may now expect that a number of readers will try to reproduce more easily the solid base of PT and will employ it as a trendsetting tool for their marine paleoclimate studies.
References
Bard, E. and Heaton, T.J.: On the tuning of plateaus in atmospheric and oceanic 14C records to derive calendar chronologies of deep-sea cores and records of 14C marine reservoir age changes. Climate of the Past, 17, 1701–1725. https://doi.org/10.5194/cp-17-1701-2021
Grootes, P.M. and Sarnthein, M.: Community comment on “On the tuning of plateaus in atmospheric and oceanic C records to derive calendar chronologies of deep-sea cores and records of C marine reservoir age changes” by Edouard Bard and Timothy J. Heaton, Clim. Past Discuss., https://doi.org/10.5194/cp-2020-164-CC2, 2021.
Küssner, K., Sarnthein, M., Michel, E., Mollenhauer, G., Siani,s G., and Tiedemann, R.: Glacial-to-deglacial reservoir ages of surface waters in the southern South Pacific. PANGAEA, https://doi.org/10.1594/PANGAEA.922671, 2020.
Küssner, K., Sarnthein, M., Michel, E., Mollenhauer, G., Siani G., and Tiedemann, R.: Glacial-to-deglacial reservoir ages of surface waters in the southern South Pacific. Paleoceanography and Paleoclimate, 47 ms-pp. (2020, rejected)
Sarnthein, M. and Grootes, P. M.: Community comment on “On the tuning of plateaus in atmospheric and oceanic C records to derive calendar chronologies of deep-sea cores and records of C marine reservoir age changes” by Edouard Bard and Timothy J. Heaton, Clim. Past Discuss., https://doi.org/10.5194/cp-2020-164-CC1, 2021.
Sarnthein, M., Küssner, K., Grootes, P.M., Ausin, B., Eglinton, T., Muglia, J., Muscheler, R., Schlolaut, G. Plateaus and jumps in the atmospheric radiocarbon record – Potential origin and value as global age markers for glacial-to-deglacial paleoceanography, a synthesis. Climate of the Past, 16, 2547–2571, doi: 10.5194/cp-16-2547-2020, 2020.
Sarnthein, M. and Werner, K.: Early Holocene planktic foraminifers record species specific 14C reservoir ages in Arctic Gateway, Marine Micropaleontology, 135, 45-55.
DOI:10.1016/j.marmicro.2017.07.002, 2018.
Skinner, L.C. & Bard, E.: Radiocarbon as a dating tool in paleoceanography, Review in Geophysics, 60, e2020RG000720, doi.org/10.1029/2020RG000720, 2022.
Citation: https://doi.org/10.5194/cp-2021-173-AC1 -
RC2: 'Review Comment on cp-2021-173', Anonymous Referee #2, 21 Apr 2022
In this manuscript, the authors seek to establish the existence and timing of ‘plateaus’ in the atmospheric radiocarbon record, and to demonstrate that these are also present in marine records from around the world. On this basis, the authors seek to argue that radiocarbon plateaus identified in marine records can be stratigraphically aligned to correlative plateaus identified in the atmospheric record, allowing calendar ages to be transferred to the marine records (and therefore allowing for ‘marine reservoir age’ offsets to be determined). This method of chronostratigraphic alignment has been termed ‘plateau tuning’ (PT).
This manuscript is quite unusual, as it does not appear to advance any new observations/data, arguments, models or insights. Some adjustments are made (again) to the proposed timing of plateaus identified in the atmospheric radiocarbon record, but this does not really make any difference to what has been proposed by the authors in several papers since 2007. A recent ‘review’ of the ‘PT method’ and its results was published by the authors just last year in this same journal. Primarily, it seems, the manuscript seeks to publish a rebuttal of a prior piece of work produced by Bard & Heaton (B&H) that was also reviewed and published in Climate of the Past last year. The latter was also accompanied by several pages of commentary by Sarnthein and Grootes, which was in turn responded to by Bard and Heaton over the course of the discussion phase of the manuscript.
Unfortunately, I find it impossible to recommend that this manuscript be accepted for publication. There are three main reasons for this: 1) it does not appear to present an original piece of research, and insofar as it presents adjustments, they are not important enough for publication on their own merit; 2) its arguments against B&H are not coherent (regardless of whether or to B&H are correct); 3) the vast bulk of figures and tables referred to in the manuscript are included in a ‘supplement’ that has not actually been produced/included. On the latter point, the promise of a compilation of all the available PT data in useful tables would have been at least one welcome contribution: but it turns out that the intention of the authors was to include ~20 disparate data tables that are already available on PANGEA and that are not at all useful in reproducing the PT data that have been published to date by the authors (it took me days to do this, and the results are not the same as what the authors have published in many cases, which is both worrying and annoying). The fact that the PT data (and associated MRA etc.) that have been produced by the authors over several years, and presented in a series of ‘global synopsis’ papers, cannot be easily reproduced by others using the multitude of available data tables, is particularly worrisome. The same can be said for the fact that only one (?) PT study exists that does not include the authors of this study (the champions of the PT approach). Incidentally, this might already answer the question of whether or not it is a ‘trend setting’ tool.
With regard to the second point raised above, the authors state that they reject the arguments of B&H based on the basis of how plateaus are identified (i.e. as ‘sequences’, like a sort of Morse code), and on the basis that B&H use a 1998 box model to support their arguments. Regardless of the validity of B&H’s remarks, I don’t see how either of these points represent a coherent basis on which to reject a criticism of the PT method, where that criticism is founded in large part on the proposed difficulty of objectively identifying plateaus (let alone sequences of plateaus) in a noisy marine radiocarbon record whose offset from atmospheric radiocarbon varies over time, as well as the proposal that sedimentary processes (such as simple - and highly likely - sedimentation rate changes during periods such as Heinrich Stadial 1, or the Younger Dryas) can also produce ‘plateaus’ in the 14Cage-depth domain, without these being causally linked to atmospheric radiocarbon variability. In addition, the claim that a 1998 box-model is somehow incorrect because of its vintage seems to miss the point: the key purpose of deploying such a model is surely to illustrate in a very simple way how the phasing and amplitude-attenuation of an input signal will be altered (filtered) in the ocean, depending on the timescale on which the signal can be communicated to the ocean, and the frequency/duration of the signal variability. You can do this with a very complex biogeochemical coupled ocean-atmosphere numerical model if you like, but if it did not show a simple phase-attenuation relationship like the box model, it would mean that the complex model had a problem! In fact, by playing around with numerical model outputs it can be shown that they do show the same principles as a 2 box-model, and that should not be surprising, as it is an expression of a simple and fundamental physical principle: parts of the ocean that have small MRA offsets (such as the tropical ocean, MRA ~400 14Cyears) can respond quickly and can pick up shorter fluctuations from the atmosphere, whereas parts of the ocean that have large MRA offsets will take longer to pick up the atmospheric signal (since a larger MRA means that the isotopic exchange timescale for that water is longer) and will pick up a smoothed and lagged response. The limits of applicability of the PT method could readily be analysed and qualified in such a theoretical context, but the authors don’t do this unfortunately.
Ultimately, the manuscript sets out to answer the question posed in the title: “is the ‘plateau tuning’ (PT) approach a misleading approach or a trend-setting tool”? I would note that, at worst, PT could be both misleading and trend-setting, and my major concern is that the authors clearly wish for it to be the latter, but have not really (either in the present manuscript, or over the course of several publications that appear to present the same datasets repeatedly) demonstrated that the PT approach is indeed viable, either in theory or in practice. As suggested above, this is not to say that some sort of defence cannot be made, in theory at least. But the authors (still) have not managed to do this. My own view is that the chronostratigraphic principles that the authors wish to apply are not completely crazy: yes, the atmospheric radiocarbon record has ‘wiggles’ and these would be transferred to other reservoirs that exchange CO2 with the atmosphere rapidly enough to pick them up. However, the conditions under which these wiggles can be recorded in other reservoirs, such as the ocean, and the biases (in amplitude and phasing especially) that will inevitably and predictably arise (even prior to the complications of sedimentation changes, bioturbation, sampling/analytical noise etc.) need to be accepted and addressed by the authors at some point if this debate is to move in a useful direction. I can think of a variety of ways to test the PT method in theory (using models), and in practice using data, and I wonder why the authors have never done something similar. If a scientific study that achieved such goals was produced, it would be a welcome and useful addition to the literature (as B&H has proven to be, insofar as it stimulates critical thinking). Such a study would best come from the authors of the present study, who appear to be the main (if not the only?) champions of the PT method; however, this is not what the current manuscript provides.
Citation: https://doi.org/10.5194/cp-2021-173-RC2 -
AC2: 'Comment on cp-2021-173_Response to Review#1', Michael Sarnthein, 26 Apr 2022
The comment was uploaded in the form of a supplement: https://cp.copernicus.org/preprints/cp-2021-173/cp-2021-173-AC2-supplement.pdf
-
AC3: 'Comment on cp-2021-173 _ Response to Review #2', Michael Sarnthein, 26 Apr 2022
The comment was uploaded in the form of a supplement: https://cp.copernicus.org/preprints/cp-2021-173/cp-2021-173-AC3-supplement.pdf
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 889 | 252 | 50 | 1,191 | 93 | 35 | 41 |
- HTML: 889
- PDF: 252
- XML: 50
- Total: 1,191
- Supplement: 93
- BibTeX: 35
- EndNote: 41
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1