Please see the attached pdf document for comments.

General

This paper investigates atmospheric stability conditions over the Arctic Ocean using sounding data obtained during the year-long MOSAiC drift of the research vessel Polarstern. Based on a classification of stability they elaborate thermodynamic and kinematic drivers of the ABL structure.

The paper is very well organized, in most parts well written and presents interesting findings, which are new and will stimulate probably further research. Nevertheless, I suggest some major revisions before its publication.

Major Revisions

1) My most important point concerns the definition of stability. The paper lives from the definition of classes, whose motivation is described in detail in a paper that is not yet finally accepted for publication. For this reason, I find it necessary to obtain more details here. E.g. the question arises if the boundaries of the classes have physical reasons. The definition of weakly stable, strongly stable and so on has long tradition (e.g. Mahrt, 1998), so it needs motivation when these terms are used in a new sense and with new thresholds.

2) In this connection it is important that a physical classification of the surface layer is usually done in terms of z/L where z is height and L is the Obukhov length. L does not seem to be available here (or is not yet available?). The authors investigate just the thermal stability, while dynamical stability is more important for modeling since flux parameterizations depend on it. This investigation could be done in terms of the dependence of the (bulk) Richardson number, which would include the effect of wind speed. Is there a reason why this was avoided? I am not against the consideration of thermal stability but it might be worth to consider the Richardson number dependent stability in addition. At least, a discussion would help in Section 3.3 when the wind speed dependence is studied.

3) I think the description of the Table 2 needs improvement. I have a list of questions here:

Line 190: Why is the near-surface stability representative of the stability within the entire ABL? Decoupling might occur etc….

In line 185 it is said that the lowest value is between 15 and 35 m, but in line 190 the near-surface stability is based on the 42.5 m level. Both together is puzzling.

First row of Table2 : Does this refer to ABL heights smaller than 50 m ?

I understood that the term ‘aloft’ refers to the layer above the ABL but below 1km. But is the given stability then an average over the whole layer? I am asking because often there is a capping inversion with strong stability but in upper layers stability is weaker.

Why is unstable stratification not a part of the classification? SHEBA data showed such cases.

Minor Revisions

1) During MOSAiC, also turbulence was measured at a mast taller than 2 m, are the data not yet available? These data would help for the stability classification, but also to correct the biases of the soundings near the surface that are discussed in section 2.2 (?)

2) The quality of some figures (4, 5, 7) is limited. Please improve the resolution and/or increase their size.

Line 160) Please write clearly that measurements below 35 m are not used (as in the Summary section).

Line 164: There are different definitions of Ri_b in the literature. So, it should be defined here.

Line 225: replace ‘trends’ by ‘tends’

Line 293: ‘wind speed correlated with stability’. Any other finding would be very strange (see definition of Ri_b).

Lines 308-310: Isn’t that very similar in spring? Furthermore, my impression is that the wind speed distribution shows only little dependence on the seasons, and perhaps this whole paragraph can be shortened.

Lines 359-362: It is stressed that for VSM and NN another factor than radiation is the driving factor. So, do you conclude that for all other regimes radiation is the most important factor and those discussed in the previous sections are not that important? This needs clarification.

Line 385: replace ‘is a more’ coupled by is ‘more coupled’

Figure 7 and its description: what is mixing ratio at ABL height? There are often large differences between mixing ratio below and above the ABL top. So, is the value at ABL height still in the ABL (below the inversion)? Does a small variation in height change the conclusion?

Line 524: Better ‘wide range’ for clarity.

Line 538: replace ‘than the’ by ‘than in the’

Line 559: Cloud top cooling enhances turbulence first of all in the center of the cloud and not only below cloud base.

Lines 555-564: There is much complexity concerning the cloud-generated mixing, which is only partly described here. Especially in case of multiple cloud layers, decoupling of the lowest ABL may occur. But decoupling of the surface layer can exist also in case of one StCu layer. This might have been captured by the stability classes accounting for the different stability in the surface layer and aloft, so this could be stressed here. Overall, I think, however, that a detailed description of the impact of clouds on stability needs further research and the information given in this section should be seen as one step towards more understanding. E.g., the given stability classification does not distinguish cloud free air and cloudy conditions. Note that, for details, it is the equivalent potential temperature rather than the virtual potential temperature, which would have to be considered for this goal in addition, which is, however, probably beyond the scope of this paper. Some sentences like this could be included.

Lines 561, 566: ‘enhanced radiation’. In the summary it needs clarification, is it longwave, shortwave, net radiation? What is a net radiation regime, what is a low, high radiation environment?

References: there are several papers cited, which are submitted or in prep. I do not know if this is against ACP rules. The paper Chechin et al. (2022) was now published in ACP (2023).

Reference

Mahrt, L. (1998). Nocturnal boundary-layer regimes. Boundary-layer Meteorology,88, 255-278