Spiky fluctuations and scaling in high-resolution EPICA ice core dust fluxes

Lovejoy, Shaun; Lambert, Fabrice

doi:https://doi.org/10.5194/cp-15-1999-2019

Articles | Volume 15, issue 6

https://doi.org/10.5194/cp-15-1999-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/cp-15-1999-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 15, issue 6

Research article

|

17 Dec 2019

Research article |

| 17 Dec 2019

Spiky fluctuations and scaling in high-resolution EPICA ice core dust fluxes

Shaun Lovejoy and Fabrice Lambert

Download

Final revised paper (published on 17 Dec 2019)
Preprint (discussion started on 03 Jan 2019)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Review', Michel Crucifix, 20 Mar 2019
- AC1: 'Response to Reviewer 1', Fabrice Lambert, 19 Jun 2019
RC2: 'general comments', Anonymous Referee #2, 20 May 2019
- AC2: 'Response to Reviewer 2', Fabrice Lambert, 19 Jun 2019
RC3: 'Review for „Spiky Fluctuations and Scaling in High-Resolution EPICA Ice Core Dust Fluxes” by Lovejoy and Lambert', Anonymous Referee #3, 22 May 2019
- AC3: 'Response to Reviewer 3', Fabrice Lambert, 19 Jun 2019

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

ED: Reconsider after major revisions (27 Jun 2019) by Carlo Barbante

AR by Fabrice Lambert on behalf of the Authors (30 Jun 2019) Author's response Manuscript

ED: Referee Nomination & Report Request started (17 Jul 2019) by Carlo Barbante

RR by Anonymous Referee #3 (06 Aug 2019)

Suggestions for revision or reasons for rejection

Review report for the revised version of “Spiky Fluctuations and Scaling in High-Resolution EPICA Ice Core Dust Fluxes” by Lovejoy and Lambert

General remarks and major comments:

Overall, the Authors have greatly improved the presented manuscript in the second version. It is now much clearer, and the arguments of the Authors are more easily followed and especially the description of the results and their discussion has won a lot. However, before the publication of the paper there are a few more issues that should be addressed. Especially the second one is still major in my view.

The Authors always refer to “dust flux measurements” multiple times throughout the manuscript. Even though this might be a small detail, I think it is misleading as the dust flux is not directly measurable but a derived quantity that incorporates both measurements of dust concentrations as well as (model) inferred accumulation rates. Even worse, the dust flux data used here is the result of a combination of a range of different dust and dust proxy measurements. I suggest the authors reformulate for example to dust flux data, record or reconstruction.

Thank you for adding the description and discussion of the data set that was employed in section 2.4 of the paper. One major thing, however, is still missing: The dust flux reconstruction uses 55 cm resolution accumulation rate reconstructions that are based on the empirical conversion of deuterium isotopic ratios of the water to accumulation rates. For analysis of variability beyond the resolution of 55 cm this poses a serious issue: Any variability beyond the resolution of the accumulation rate reconstruction is purely a result of the variability of the dust concentration in the ice. In turn this means that if the variability of the dust flux is interpreted one interprets different records at different time scales, i.e. fluxes at scales >55 cm and concentrations at scales <55 cm. Especially due to the fact that the time scales represented by these depth resolutions changes with depth in the core the effect on the fluctuation analysis is not entirely obvious and likely not straight forward.
The authors should thoroughly test the effect of this on their results and clearly discuss the effect in the paper.

The discussion of the uncertainties related to the analysis that was added to the conclusions is better placed in the discussion, in my opinion. Furthermore, I feel that the “we live in a single realization” argument is a little bit of an easy answer to the concerns raised in the first round of reviews. I will leave it up to the editor, whether a more thorough uncertainty analysis using i.e. simulations needs to be added or if the presented argument is sufficient.

Minor comments:

P2L25ff: In the added explanation of the macroweather-climate transition scale, the authors elude twice to “new long frequency processes” and “new, slow internal sources of variability” without going into much more detail in the remainder of the paper. I suggest to either at least clearly hypothesize and discuss the nature of these “new” processes and sources of variability or remove the “new” from each of the sentences.

P14L26ff: Where do you show the change in correlation, is the reference missing or is this planned in a future publication?

P16L31ff: The whole paragraph about DO-events is repeated from above, I assume this is an editing error.

Hide

RR by Michel Crucifix (03 Sep 2019)

Suggestions for revision or reasons for rejection

The authors have addressed all comments point-by-point. Results are significant and the study should be published. I am still left with a couple of comments, all are mostly editorial comments, or at least, can be adressed with adequate edition.

p.2: "indicating that the new long frequency processes become dominant". Not clear what is the "new" long-frequency process (in what sense is it 'new' and does it 'become' dominant". Perhaps be more explicit about what understood as of anthropogenic origin?

p.4 ("respect the scale symmetry") and p.11 ("Scaling is a statistical symmetry"). Turbulence theory (as, among others, given in the Lovevoy / Schertzer book, ch. 2) tell us how symmetry causes scaling, but the article here misses a proper definition of what is a "scale symmetry", or in what sense "scaling is a statistical symmetry". Isnt'it more accurate to state that "(statistical) scaling is a consequence of (physical) symmetry"?

p.8 : "that may not globally be the case" : be more accurate or explicit : "although this may not be the case everywhere".

p. 11, l. 3: "in dimensional or nondimensional time that statistical" -> "in dimensional or nondimensional time, statistical"

p. 10, l. 12: remove second occurrence of "clearly"

p. 16, l. 32: "thanks to the scaling, very few processes are Gaussian". Deterministic theory is very at ease with explaining non-Gaussian fluctuations without resorting to scaling arguments. The Lorenz 63 system maps Gaussian-distributed initial conditions onto a non-Gaussian distribution of final states.

p. 20, ll. 17- 24 : This author response to my earlier comment left me a bit unsatisfied. It is commonplace in statistics to provide estimators and estimator variances for stochastic processes, a problem which can generally be addressed with
either a frequentist or a Bayesian framework. Perhaps the most straightforward and classical example is the estimation of the mean of a random process, but other trivial and less trivial examples include the problem of estimating the coefficients of an auto-regressive processes, spectrum estimation, parameter estimation of non-linear stochastic dynamical systems.... So if I understand correctly the author's problem is the lack of a reasonable 'generic stochastic process' model, which they could derive a distribution from. This seems rather different than the objection they raise, which they claim is related to the problem of having a single draw from a stochastic process.

Hide

ED: Reconsider after major revisions (13 Sep 2019) by Carlo Barbante

AR by Anna Wenzel on behalf of the Authors (25 Sep 2019) Author's response

ED: Publish as is (24 Oct 2019) by Carlo Barbante

Post-review adjustments

AA: Author's adjustment | EA: Editor approval

AA by Fabrice Lambert on behalf of the Authors (26 Nov 2019) Author's adjustment Manuscript

EA: Adjustments approved (04 Dec 2019) by Carlo Barbante

Short summary

We analyze the statistical properties of the eight past glacial–interglacial cycles as well as subsections of a generic glacial cycle using the high-resolution dust flux dataset from the Antarctic EPICA Dome C ice core. We show that the high southern latitude climate during glacial maxima, interglacial, and glacial inception is generally more stable but more drought-prone than during mid-glacial conditions.