General comments:

This is a welcome and well-presented addition to the literature on the synoptic patterns associated with clear-air turbulence. The authors revisit this topic, which has been largely neglected for decades, with new and improved ways of identifying and analyzing both the synoptic patterns and CAT. The key results of the classification are two main synoptic modes that account for 70% of all moderate-or-greater CAT in the study, which are useful for both practical forecasting considerations and future research on CAT mechanisms. The primary, but minor, weaknesses are a relative lack of discussion of non-local generation of CAT near (and not-so-near) convection for the second mode (II), and some missing literature references that would provide a broader context for the research and the results.

Specific comments:

Line 39: Did Hislop and Bannon specifically connect the Richardson number with CAT? Rustenbeck, in the April 1963 Monthly Weather Review, lists other researchers but indicates that there was up to that point no consensus on whether or not low Ri correlated with CAT.  https://journals.ametsoc.org/view/journals/mwre/91/4/1520-0493_1963_091_0193_taorcw_2_3_co_2.xml



Lines 80-89: Is the RWB referred to specifically cyclonic RWB, i.e. LC2, or is the RWB of both types? LC1 events (anticyclonic RWB) can concentrate low PV, so are they indicative of WCB events? What is the reasoning or the literature trail recommending identifying WCBs as the defining synoptic feature, over a pure RWB interpretation?

Lines 91-93: Rather than citing later, post-PIREPs-use literature, a better source for the limitations of PIREPs is Schwartz, 1996 Weather and Forecasting, https://journals.ametsoc.org/view/journals/wefo/11/3/1520-0434_1996_011_0372_tquopi_2_0_co_2.xml

Lines 169-170 and 276-279: the definition of "clear" used permits quite moist conditions. While this is consistent with how CAT is defined operationally, it likely retains quite a bit of convectively generated CAT that propagates beyond the cloud layers, both vertically and also horizontally. The research on near(and not-so-near)-cloud turbulence, e.g. Lane et al. 2011 Bulletin of the AMS (https://journals.ametsoc.org/view/journals/bams/93/4/bams-d-11-00062.1.xml) and follow-ons, should be cited and incorporated.

Line 274 or so: In Figure 6b, the notable extent of the deformation pattern to the north for type II cases could imply that these occur on the equatorward flank of the jet stream, as would be expected for low-to-negative PV situations.

Lines 276-279 and 294-295: More generally, the triggering of CAT locally via non-local means, e.g. gravity wave propagation, is not discussed other than in one paragraph early on (lines 70-78). CAT is likely to be, in some and perhaps many instances, the product of dynamical adjustment processes taking place on the synoptic-or-smaller scales distant from the CAT event itself, ultimately ending up as KHI locally. While this study focuses on synoptic patterns, the authors should incorporate more caveats indicating that what happens locally may have non-local triggers. This is definitely true for type I cases, but also possible for type II cases; see Thompson and Schultz 2021 Geophysical Research Letters (10.1029/2021GL092649) for a modeling study of inertial instability in the jet stream and its connections to both gravity wave generation and CAT.

Lines 307-319: With regard to the Ellrod indices, it would be interesting to see how TI3, which includes divergence tendency (Ellrod and Knox 2010 Weather and Forecasting, https://journals.ametsoc.org/view/journals/wefo/25/2/2009waf2222290_1.xml; see Lee et al. 2023 Geophysical Research Letters for a recent application, at https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2022JD037679), would perform. The inclusion of divergency tendency was intended to capture adjustment processes characterized by more divergent flows. Would TI3 capture type II events even better?  This is not required, and could be "future work."

Lines 375-377, Figure 12d: This is arguably an important result, because it clearly portrays the unusually high frequency of moderate-or-greater CAT (or at least turbulence) in negative-PV regions. Furthermore, this result is independent of the classification scheme used, although it does line up well with WCBs. The only similar plot (using absolute vorticity) I recall is from a relatively obscure source, using much less data: Sparks, W. R., S. G. Cornford, and J. K. Gibson, 1977: Bumpiness in clear air and its relation to some synoptic-scale indices. Geophysical Memoirs 121, 53 pp.

Lines 405 and Figure 13: I'm surprised that there aren't more type II events over the north Atlantic, for example in ridges/with WCBs. Is this related to undersampling and/or the threshold for the definition of an "event" (lines 537-546)?  Also, what is the spatial distribution of the unclassified 30%?

Lines 485-489: Here, again, I encourage the authors to be careful to acknowledge that not all CAT is going to have an in situ trigger, but may be related to dynamical adjustment processes that are not local.  In particular, there are likely to be type II events that turn out to be "near"-cloud turbulence events rather than CAT.

Lines 490-495: an update of the classic synoptic patterns for CAT (e.g., Fig. 12.1 in Sharman and Lane's Aviation Turbulence book), based on 21st-century data and tools, would be a very useful result from this line of research.

Line 563: Some near-cloud turbulence could be screened out by requiring the aircraft and reanalysis temperatures much less (not just less) than the cloud-top temperature. Was this attempted?

I enjoyed reading this paper (see attached pdf file for review comments).