1200 years of warm-season temperature variability in central Scandinavia inferred from tree-ring density

Zhang, Peng; Linderholm, Hans W.; Gunnarson, Björn E.; Björklund, Jesper; Chen, Deliang

doi:https://doi.org/10.5194/cp-12-1297-2016

Articles | Volume 12, issue 6

https://doi.org/10.5194/cp-12-1297-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/cp-12-1297-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 12, issue 6

Research article

|

03 Jun 2016

Research article |

| 03 Jun 2016

1200 years of warm-season temperature variability in central Scandinavia inferred from tree-ring density

Peng Zhang, Hans W. Linderholm, Björn E. Gunnarson, Jesper Björklund, and Deliang Chen

Download

Final revised paper (published on 03 Jun 2016)
Preprint (discussion started on 19 Feb 2015)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

SC C35: 'Comparison with previous results', Vladimir Matskovsky, 20 Feb 2015
- AC C53: 'Reply to Vladimir Matskovsky', Peng Zhang, 04 Mar 2015
RC C146: 'Encouragement for improved research', Anonymous Referee #1, 25 Mar 2015
- AC C781: 'Response to referee1', Peng Zhang, 24 Jun 2015
RC C305: 'Review', Anonymous Referee #2, 05 May 2015
- AC C787: 'Response to referee2', Peng Zhang, 24 Jun 2015

Peer-review completion

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

AR by Peng Zhang on behalf of the Authors (24 Jun 2015) Author's response Manuscript

ED: Referee Nomination & Report Request started (25 Jun 2015) by Eduardo Zorita

RR by Anonymous Referee #3 (28 Jul 2015)

Suggestions for revision or reasons for rejection

The article presents a temperature reconstruction for central Scandinavia based on maximum-latewood-density that updates and extends the existing proxy archive. The new data, generated by the authors, are supposed to replace an older dataset with a weaker environmental signal and a partly unknown provenance. Although this is generally a very valuable contribution to the proxy network in this area, I would be very cautious with citing or using the new reconstruction in its current state, since I identified some shortcomings regarding the soundness of the applied methods and the reproducibility of the study.

Major comments

- In the previous round of reviews the referees already noted the partly sparse methodological remarks. The technical part was improved, but there are still considerable gaps in the description of the quantitative analyses. The reconstruction technique, for example, is not described at all, neither in the method section nor elsewhere, although this is an essential information to replicate the study. There are also more details required regarding the detrending (see below) and the error estimates. The latter is only very briefly addressed in the figure captions.
- I wonder why the authors used negative exponential functions to approximate the raw MXD data. Usually a so called Hugershoff function fits much better for MXD data. This might not affect the overall trends of the RSFi-chronology but low replicated periods with a low mean tree age (like the early MCA) might be altered.
- Although I do see the benefit and the need of adjusting the mean of different tree-ring sites, there are two major points of criticism w.r.t. this method: (1) The information given is again not sufficient to replicate the study! Even the paper cited (Zhang et al. 2015) is not elucidating since it only uses a subset of the proxy data. For example, it remains unclear how sites that do not cover the (whole) 1300-1550 period were treated. It would also be desirable to find the actual values that were added/subtracted for each site somewhere in the manuscript/supplement. In consideration of other potential biases (see (2)) this would yield an estimate for the size of the effect. (2) Additionally, it seems very important to report the mean age of the samples in the window of overlap. Basically the fact that dendrochronologists detrend their samples w.r.t. age stems from the assumption that the mean value differs significantly with age. Hence, a homogeneous age distribution is an ultimate prerequisite for applying such a “mean-adjustment”. This needs to be stated clearly and proven(!) in some way. Otherwise, the detected long-term cooling in the adjusted chronology could well be perceived as a remnant of age-trends. In order to validate the applied methodology I strongly recommend displaying age-aligned Regional Curves for samples in the common period (1300-1550) before and after mean-adjustment in a supplementary figure.
- Grammar and language still need improvement, especially in the revised sections.
- The revised manuscript puts much weight on the comparison with other reconstructions (instead of refining the methods section). For a fair judgement, however, some additional issues should be considered: (1) Error-estimates should be displayed, at least for the new reconstruction, to identify really significant differences. (2) Temperature-anomalies should be compared to temperature-anomalies. There’s no obvious need to use z-scores. (3) For the large-scale comparison, the NH reconstruction by Christiansen and Ljungqvist (2012) is not the ideal choice. In the IPCC ensemble of reconstructions this is the one that clearly sticks out due to its high low frequency variability. And as already stated by the authors, the seasonal window differs.
- It’s a pity that most samples terminate in 2006/7. Only 5 samples extend to 2011. Generally the interest in the most recent years of climate reconstructions is fairly high. Especially for tree-ring studies, which can be affected by the so called ‘divergence problem’. AN update based on old samples will be outdated soon.
- I didn’t find a statement about data availability in the end of the main text. All data used in this study should be made available.

Minor comments

- Lines 89 and 104: Replace ‘glacifluvial’ with ‘glaciofluvial’.
- Line 58-67: Is the MCA really the key target of this story? There’s eventually not much light brought into this period.
- Line 124: Remove dot in ‘(Table. 1)’.
- Line 126-128: Sentence unclear. It implies that the authors know what the temperature difference between MCA and LIA was, which I doubt. Rewrite.
- Line 143 and 152: ‘subtracted from or divided by’ - this is not termed correctly. Should be a-b and a/b, not a-b and b/a. Reword.
- Line 166: Add information about the exact settings for the detrending (like: negative exponential functions used for the individual fitting, ratios or residuals between fitted function and raw data,…).
- Line 173: In the figures, EPS-values don’t seem to have a 25-year period of overlap.
- Line 237: Replace ‘multidecadal’ with ‘multicentennial’.
- Line 238: Be more consistent with using either the term multi- or two-curve RCS.
- Line 269: ‘20th century’?
- Line 321-322: In which frequency domain?
- Figure 1: It’s hard to locate the map section within Scandinavia. References (borders, towns,…) missing.
- Figure 4: The sampling site should be marked.
- Figure 6: Correct spelling in the legend.

Hide

RR by Anonymous Referee #2 (31 Jul 2015)

ED: Reconsider after major revisions (01 Aug 2015) by Eduardo Zorita

Dear Dr. Zhang,

Thank you for submitting the revised version of the manuscript to Climate of the Past Discussion.
This version has now been evaluated by two reviewers. Unfortunately, reviewer #1 declined to review the revised version and a third reviewer had to been identified. This third reviewer had access to the previous reports and also evaluated whether this revised version addresses the comments and concerns of the previous two reviewers.

As you can see in the attached reports, the opinion of the reviewers on the revised version is still not positive. Both raised major concerns that need to be addressed.

Reviewer #2, who also reviewed the original submission, acknowledges that while the revised version represents some improvement with respect to the initial submission, the revisions are not deep enough. In particular, this reviewer feels that the comparison with other reconstructions does not really place the your new temperature reconstructions into the right perspective. In this regard, reviewer #3 also expresses concerns that the reconstruction of the Northern Hemisphere temperature by Ljungqvist and Christiansen is not the most relevant target for this comparison, as this reconstruction is sort of an outlier within the reconstructions presented in the last IPCC report.

Reviewer #3, in addition, complains that the methodological description of your work is insufficient to evaluate the quality of the reconstruction. Important methodological aspects are still missing or are unclear - the report provides quite specific comments on this regard.

Overall, I think that the opinion of both reviewers essentially overlaps, in that a new round of revision and review is needed I would encourage you to closely pay attention to the recommendation of the reviewers, as this is already a version that has been revised once.

Both reviewers also complain that this version contains numerous spelling and grammatical errors. Although the final version of the manuscript will undergo copy-editing by the technical editors, I would also strongly suggest to put great care in the formatting of the manuscript, as the evaluation of a carelessly submitted version hinders the work of the reviewers and may nudge them towards a general negative stance on the manuscript. If you have the chance, please seek the help of a native English speaker.

Please, also submit a point-by-point rebuttal to the points raised by the reviewers, in particular of those by reviewer #3.

Despite these critical evaluations, both reviewers agree that this new data set is an important contribution to the field of dendroclimatology, and I am looking forward to recommending the publication of this study. However, the reviewer's concerns should be first fully addressed.

Sincerely,

Eduardo Zorita

Hide

AR by Peng Zhang on behalf of the Authors (02 Feb 2016) Author's response Manuscript

ED: Referee Nomination & Report Request started (04 Feb 2016) by Eduardo Zorita

RR by Anonymous Referee #3 (18 Mar 2016)

Suggestions for revision or reasons for rejection

This manuscript is now considerably improved regarding the transparency of methods and general readability. All my comments have been addressed, most of them in a satisfactory way. However, in my opinion the study is still not ready for publication for the following reasons:
1) I need to recommend another thorough language check. There are less but yet numerous spelling mistakes in the manuscript.
2) The application of the “mean-adjustment” among sub-sites is still not fully convincing to me. The new supplementary figures that address the initial criticism do not disperse my doubts. In fact the figures confirm that the dataset lacks two prerequisites: Mean age in the reference period differs considerably among sub-sites and most sub-sites do not cover what is called the “common period”. As indicated in my original remarks both conditions can induce an offset in the mean MXD-values, which is spuriously suppressed by adjusting the mean. Possibly, these effects are small, but this would have to be proven first! There are many tree-ring datasets out there which are a combination of nearby sub-sites and some of them could potentially be treated with a “mean-adjustment”. But before advocating this new method, these issues should be sorted out.
3) The discussion was largely revised. It is now more focused on the new findings of this story. Unfortunately, the authors are often too vague and thus some of their conclusions appear somewhat bold. One such example is that two reconstructions are referred to as “rather similar” (rather subjective term!) in line 248, although exchanging the records would certainly alter the results and maybe their interpretation (Figure 6 indicates a different timing for some of the cold and warm extremes identified in panel d if using the RCS- instead of the RSFi-reconstruction). The discussion could also gain conciseness if the authors would follow more closely their line of argument, for example regarding their broad discussion about potential causes for the MCA (lines 254-271) which seems pretty misplaced.

In summary, I still believe this study could be a useful contribution to paleoclimate and tree-ring research, although I do not recommend publication of the manuscript in its current state.

Hide

RR by Mary Gagen (03 Apr 2016)

ED: Publish subject to minor revisions (review by Editor) (05 Apr 2016) by Eduardo Zorita

AR by Peng Zhang on behalf of the Authors (30 Apr 2016) Author's response Manuscript

ED: Publish as is (15 May 2016) by Eduardo Zorita

AR by Peng Zhang on behalf of the Authors (18 May 2016) Manuscript

Short summary

We present C-Scan, a new Scots pine tree-ring density based reconstruction of warm-season (April-September) temperatures for central Scandinavia back to 850 CE, extending the previous reconstruction by 250 years. Our reconstruction indicates that the warm-season warmth during a relatively-warm period of last millennium is not so pronounced in central Scandinavia, which adds further detail to our knowledge about the spatial pattern of surface air temperature on the regional scale.

1200 years of warm-season temperature variability in central Scandinavia inferred from tree-ring density

Download

Interactive discussion

Peer-review completion

1200 years of warm-season temperature variability in central Scandinavia inferred from tree-ring density