Comment on Mid-Holocene reinforcement of North Atlantic atmospheric circulation variability from a western Baltic lake sediment record by Markus Czymzik et al.

This is a very interesting study which attempts to provide insights into the North Atlantic Oscillation (NAO) variability throughout most of the Holocene. The detailed sedimentology from lake Kalksjon, west-central Sweden, is excellent and provides some justification for its use as a qualitative reconstruction of the NAO.  Overall, I found the manuscript well written and will be of high interest for the paleoclimate community studying the climate during the Holocene for this region. It is therefore suitable for Climate of the Past. However, to assess this archive as a reconstructed “qualitative” record of the NAO, I believe there needs to be further examinations.   

Major comments:

Different NAO reconstructions show periods of coherency and often no coherency at all. This is partly because the NAO itself exhibits non-stationary behavior, so the use of one single location may not capture the whole variability. One aspect that may have been overlooked in this paper is the other existing NAO proxies. Have you explored other records that may be sensitive to large-scale and long-term NAO fluctuations? That said, from 1800BP to the present, the NAO from Olsen et al. (2012) and the TOC record don’t seem to match quite well, but perhaps if you plot different reconstructed NAO, a better co-variability (correlation) may be seen. I am thinking of speleothems in Europe and North Africa as well as the new one from Becker et al. (2020). See also Wassenburg et al. 2016 Nat Geosci, Baker et al. 2015, etc. 

Figure 12 shows the relationship between paleo NAO (Olsen et al. 2012) and the TOC content in the studied lake. When sampling both records to the lowest resolution of the corresponding record; do you find any significant correlations? Same comment for the Siberian High.

Also, the lack of coherence between your record and other NAO proxies could be explained by other mode of variability that may have been more persistent in the past. The Scandinavian Blocking, for example, accounts for ~27% of the winter North Atlantic variability. A persistent Scandinavian Blocking in winter would translate to cooler conditions in the region, thereby presumably increased ice cover time (in turn less productivity). Any thoughts on this?  I suggest to add more discussion around line 345.

Some other minor comments:

Figure 3: Given you are dealing with decadal to centennial scale variability, I think an average (monthly?) of temperature and precipitation would improve visualization.

Figure 4 shows the composite with the XRF cluster stratigraphy. It should be located after figure 7 (PCA of the elements).

Figure 4: Add average temporal resolution for each cluster, i.e., mm/year

Figure 10: the sharp decline in TOC falls within 4.2k BP. Do you consider that your proxy responded to the 4.2k event?

Figure 12: we don’t see much the Trouet et al. NAO, the color is too pale. Also, why not the selection of the Ortega et al. reconstructed NAO?

Figure 12: the lines of the SH index are outside the x-axis.

Figure 13: the spectral peaks don’t seem to match quite well, perhaps a cross-spectral analysis would give something better. I would suggest moving this to the supplement.

Line 89: could you provide more information as to how the grain-size was extracted?

Section 3.3: provide information on how many thin-sections were produced.

Section 3.5: Why no pollen analysis on other SDU?

Section 4.3: Add a Table showing the matrix correlation between µ-XRF data

Line 104: 3s is deem low. Is this a typo?

Line 105: why only these elements? For example, why is Ca omitted?

Line 106: How did you build the µ-XRF composite? Elements often decrease (increase) their values at both edge of the sediment sections. Did you remove those data?

Line 239: There is no Ca profile showing these peaks. Maybe add into supplement.

Line 364-365: Please rephrase

Comment on Mid-Holocene reinforcement of North Atlantic atmospheric circulation variability from a western Baltic lake sediment record by Markus Czymzik et al

I agree with reviewer 1 that this is a very interesting and well-written paper that provides new insight into NOA variability using a Swedish lake record. I am, however, not an expert in NAO and NAO variability so I have focused my review on other aspects of the paper.

Comments:

Line 18    Maybe a minor comment, but I think that LKJ would be a better acronym than KKJ

Line 19    I prefer CE (Common Era) rather than AD

Line 42    Here it might be useful if you define what you mean with the western Baltic region (or the western Baltic Sea region?). There are at least ten different definition of the “Baltic Region” in Wikipedia

Line 55    I checked the Swedish Land Survey online maps and long. 13°03’E is more correct.

Line 59    west

Line 63    Figure 3 shows daily temperatures and precipitations, not mean monthly temperatures, see also comment by Rev. 1

Line 105  I agree with Rev. 1, why was not e.g. Zr and Ca analysed

Line 228  I would use “concentrations of the artificial radionuclides” rather than “contents”

Line 238  See comment by Rev. 1. Was Ca measured after all?

Line 266  I was slightly confused by the discussion about the isolation of the lake basin from Ancient Lake Vänern. I have not read the Risberg et al paper in any detail, but the uplift history of the area is complicated with evidence of irregular postglacial isostatic uplift. Hence, the comment that the isolation of L. Skjutsbolstjärnet located at a similar location (and what is meant by that?) and height supports your isolation point must be clarified (or deleted?). The isolation age of L. Skjutsbolstjärnet is c. 9600 uncalibrated C14 years which calibrates to c. 10,900 cal a BP

Line 269 cal is missing before “a BP” (Stanton et al. 2010)

Line 313  typo? Should be east of L. Kälksjön

Line 318  My alternative interpretation of the TOC contents would be “a sharp increase until ca 4500 cal a BP with a drop at ca. 5300 cal a BP”, i.e. the increase continues another 1000 years, followed by more stable values until recent times. I would also add δ¹³C to Figure 10.

Line 331  Does this refer to winter or annual temperatures, precipitation and windiness?

Line 341  I am not certain that the reference to Almquist-Jacobson is 100% correct here. Almquist-Jacobson’s sites are situated in an area that may have been settled by humans later than the river and lake valleys in Värmland. There are archaeological findings north and north-west of Torsby dating back to the early Neolithic (3800-3300 BC)

Line 364-374      Here you should also discuss proxy records from Sweden showing climate shifts at this time, e.g. the peat record from Store Mosse (Kylander et al., 2013; QSR) and lake level records from L. Bysjön (Digerfeldt, 1988; Boreas). These could also be shown to Fig. 12.

Line 546  Change Väners to Vänern

Table 1    If possible, give weights and type of material (e.g. terrestrial or lacustrine plant remains?)

Figure 2   Give the source of the historic document. Museum, library?

Figure 4   Sediment deposition units. Give the age range for each SDU, if possible, not only the onset