This manuscript takes advantage of a long-term hydrographic data set from a series of stations encircling Iceland to assess major characteristics of upper-ocean variability and to evaluate long-term trends in the region.  The study shows that salinity governs stratification at stations northwest of  Iceland, while temperature governs stratification to the south.  To the north of Iceland, alternating impacts from the North Icelandic Irminger Current and the East Icelandic Current lead to a mixed response.

The data set used for the study offers a rich supply of information, and the authors have chosen an interesting question to pursue.  Parts of the analysis would benefit from more detail.  I feel that the manuscript will likely be suitable for publication after careful revision.

1. The overall analysis of the manuscript addresses several distinct issues that are not always tightly linked together. One focal point is temperature vs salinity controls on mixed-layer depth or stratification, including consideration of the seasonal cycle.  A second thread considers long-term trends in mixed-layer temperature and salinity along with the quantities with which they correlate.  A third aspect assesses the multi-year linear trends in mixed-layer temperature in summer and winter.  Analyses explore the mixed-layer evolution on the seasonal scale using a one-dimensional mixed-layer model and look at historic temperature vs salinity domination on a regional and seasonal scale.  These are interesting and related analyses, but they are not fully linked together to provide clear and targeted interpretation of the results.  For example, the long-term temperature trends in Figure 6 are interesting but not well connected with the rest of the analysis.  In a rewriting, the manuscript should be more tightly focused to identify clear and linked results, well grounded in robust statistics. 
2. One gap in the manuscript is a lack of statistical detail. This gap is particularly noticeable in Figure 5, in which the authors show time series of mixed-layer temperature and salinity anomalies, mixed-layer depth, and the North Atlantic Oscillation.  The authors discuss correlations between these records but do not report correlation coefficients or statistical significance.  To show that the patterns that the authors observe in their plots are robust, they should report quantifiable statistical metrics.
3. Similarly in Figure 6, the authors fit linear trends to plotted mixed-layer temperature records, but they do not specify the slope of the linear trends. In addition, the caption to Figure 6 does not indicate what the p values represent.   These pieces of information should be added. 
4. Given the discussion in Figure 5 and given the character of the records in Figure 5, I was surprised by the decision to fit trends in Figure 6. The discussion in Figure 5 emphasizes the specific relations between plotted quantities rather than long-term trends, so I was expecting Figure 6 to report correlations.  It would be interesting to see the correlations between NAO, mixed-layer depth, and MLT mapped out for the full set of stations.
5. Since the analysis of Figure 6 focuses on trends, and the overall goals of the manuscript are directed toward alpha and beta oceanic regimes, the authors could/should expand the manuscript discussion to indicate how the trends (and regression coefficients, perhaps) inform their understanding of alpha vs beta ocean regions.
6. Line 133. “great score”.  In this usage, “great” sounds like a word that expresses an opinion.  This point needs to be quantified, and more neutral wording should be used to express the skill of the density threshold method.
7. Lines 206-207. “The southern stations ST5 and SB5, have a minimal contribution from salinity, which may be associated with the numerous river discharges and the proximity to the continental shelf.”  This is an interesting point.  Does the fresh water budget support this hypothesis?  It would be useful to quantify the volume of freshwater discharge and its expected impact on salinity.  Precipitation or oceanic circulation would be other factors that could influence salinity.
8. Figure 3. The station labels are much smaller than the other figure labels and are too small to read clearly.  The figure should be redrafted with larger labels.
9. Line 214. “hydrographic onset”.  The meaning of this is unclear.  Does it refer to the top of the hydrographic profile or the seasonal onset of a change in the hydrographic profile?
10. Figure 4. The figure shows open circles for deep mixed layers.  The justification for this is not clear, since deep mixed layers can be as dynamically relevant as shallow mixed layers.  Further explanation is needed.
11. Line 270. Correlations with the NAO should be quantified.  As noted above, the manuscript should report correlation coefficients and evaluate statistical significance.
12. Figure 5. The repetition of panels g, h, and i seems unnecessary.  Could the NAO time series be superimposed on the panels above (along with the addition of concrete correlation statistics)?   
13. Line 286. “aliasing”.  The term “aliasing has a specific meaning in time series analysis, and the usage here seems inconsistent with that usage.  This could be described as “superimposed on”.
14. Lines 295 and following. Choice of one-dimensional model.  The Price-Weller-Pinkel model has been used extensively over the last four decades for upper ocean analyses.  It is not the only possible model, and other recent studies have made use of GOTM or a stripped-down form of KPP.  Thus, it’s important to justify the choice of the PWP model.
15. Lines 295 and following. The focus of the one-dimensional mixed layer analysis on winter only also needs clarification and should be more carefully described to explain that the analysis is really looking a the winter-to-summer transition.  The   PWP model has previously been used over a broad range of latitudes and for all seasons.  Thus, a priori, there's not an obvious reason to exclude summer.
16. Figure 7 calculations. How is the mixed-layer model initialized?  Does it start with stratification typical of February?  It's surprising that the mixed layer in the model appears to deepen at the outset.  I would have expected it to be initialized with a profile that matches the climatological observations.  This should be explained.
17. Figure 7 color scale. Please check colors.  Green/red contrasts can be challenging for readers with limited color vision.
18. Line 389. “not correlated with the NAO”.  The lack of correlation should be quantified in the main body of the text, particularly if it is referenced in the conclusions.  Better language would specify that the correlation with the NAO is not statistically different from zero.
19. Line 390. “at times reaching 2 degrees C”.  This number should be reported as a rate, in units of change in temperature per unit time.  Please specify the time interval over which this estimate is computed. 
20. Minor grammatical points
    • Line 17: “sections” -> “section”
    • Lines 19 and 20: It would be good to make a decision about consistent capitalization of regions (“South” or “south”, etc.)
    • Line 24. “alternate the temperature and salinity contribution to stratification”.  Wording is unclear.  Maybe the authors could write, “while in the North, the North Icelandic Irminger Current and East Icelandic Current alternate seasonally, shifting the region between temperature-dominated and salinity-dominated stratification.”
    • Line 28. Remove comma after “locally”.
    • Line 32. “their link”  Wording is confusing.  The word their implies a plural reference point, but the grammatical structure of the sentence does not clearly identify what this reference should be.  Maybe “This study provides an unprecedented and detailed description of the seasonal to multi-decadal variability of mixed-layer depth and stratification around Iceland, showing links between this regional variability and the changing North Atlantic….”
    • Lines 42-46.  Capitalization and punctuation are inconsistent for numbered points.  All three items could be capitalized as separate sentences, or all three could be started with lower case letters, with semi-colons to separate the items.  But mixed punctuation and capitalization is confusing.
    • Line 45. Remove “with”
    • Line 55. “drives” -> “drive”
    • Line 57. “heat fluxes are the main drivers” or “heat flux is the main driver”?
    • Line 57. “on the center” -> “in the center”
    • Line 58. “Nordic Seas have been previously described as a ‘melting pot’”.  Inconsistent plurals.  The Nordic Seas region is a melting pot?  Or Nordic Seas are melting pots?
    • Line 59. “Nordic Seas are also a large repository”.  Same thing.  “The Nordic Seas region is a large repository”?
    • Line 65. “of the Arctic Amplification” -> “of Arctic Amplification”
    • Line 65. “the decrease” -> “a decrease”
    • Line 74. “forcings” -> “forcing”
    • Line 74. “to control” -> “for controlling”
    • Line 74. Add comma after “mixing”
    • Line 76. “of the strong” -> “of strong”
    • Line 99. “hinders” -> “hinder”
    • Line 101. Add comma after “MLD”
    • Line 114. “IB” -> “IH”
    • Lines 128-137. It’s standard practice to subscript theta in sigma_theta.
    • Line 137. The line following equation (10) continues the sentence containing equation 1 and should not be capitalized or start a new paragraph.  Equation 1 should be punctuated with a  comma rather than a period
    • Line 137. “decomposed on” -> “decomposed to show”
    • Line 138. “contribution of the salinity” -> “contribution of salinity”
    • Line 146. “Where” is a continuation of the sentence containing equations (3) and (4).  No capitalization and no indenting.
    • Line 148. Add comma after “salinity” since this is a compound sentence.
    • Line 151. “have” -> “has”. (The sentence structure implies that only one component needs to have an impact, so the verb should assume a singular subject.)
    • Line 164. No indent.  Please check all equations for this issue.
    • Line 187. “strike out” has a couple of distinct usages, but this reads as if it is using the baseball metaphor, which means to fail completely. Maybe use “are strikingly saltier”.
    • Line 205. Missing words.  Maybe "despite the fact that stratification ...."
    • Line 251. “correlates” -> “correlate”
    • Line 258. “in the order” -> “on the order” OR “are the same order of magnitude as”
    • Line 261. “neither …. nor” is not used correctly here.  Change to “do not seem correlated with the MLT/MLS or with the ….:
    • Line 264. “it is”.  The text is not clear about the meaning of “it”.  Clarify whether “it” is station LB8 or the winter MLD.
    • Line 266. Start a new sentence:  “front, fresher” -> “front. Fresher”.  Add a verb:  “MLs associated” -> “MLs are associated”
    • Line 270. “of NAO” -> “of the NAO”
    • Line 271. “MLS” -> “MLS,”
    • Lines 325-326. “distribution (Fig. 8) … southern (northern) … alpha- (beta-) …”  Avoid using opposites in parentheses since opposites are also used for clarifications (e.g. Fig. 8 is likely not a match to “northern”).  In general, this opposite-in-parentheses structure is difficult for readers to parse.  If the point is worth making, then it can be spread into two sentences.
    • Line 325 and discussion of Fig. 8. The method underlying the results in Fig. 8 is shown in the Introduction. Here the text could reference Equation (5) to point readers to the relevant aspect of the computational approach.
    • Line 338. “200” -> “2000”?
    • Line 364. “northwestward, hence”. This is a comma splice.  Start a new sentence instead.
    • Line 366. Don’t put opposites in parentheses.  Write a clear two-part sentence instead.
    • Line 370. “variability, they”.  Comma splice.  Start a new sentence with “they”
    • Line 374. Add a comma after “ML’s”

Please see the attached document CommentsToAuthors.pdf