Manuscript: Rock Glacier Inventories (RoGI) in 12 areas worldwide using a multi-operator consensus-based procedure.

Line Rouyet et al. submitted to ESSD

I enjoyed reading this paper, and to my understanding this approach of mapping rock glaciers based on a given set of assumptions and definitions seem to be well founded and carried out in a thought through manner. I find the considerations accounted for in the manuscript to be meaningful and interesting, and the paper seems like a natural next step following the work carried out by the RGIK group. There are a few things I think could improve the manuscript a bit, and this would mainly be for readability and overview.

Section 2.1: you describe a bit how the exercise was performed, but could you perhaps elaborate a bit here? For instance, were mapping teams assembled randomly, or did they comprise individuals with local expertise? Table 1 suggests that the principal investigator of each team possesses some or significant area-specific knowledge, but what about other team members?  From my own background I know that it can be quite difficult to 1. change my opinion about familiar terrain, and 2. communicate my interpretation of landforms to local experts in areas that I lack familiarity with. I assume that you found ways to handle this, but it would be nice to know.

In table 1 you include the numbers of RGUs within each area, this is really a result. Instead, consider including information on the materials available in each study area, such as orthophotos (with resolution), DEMs (with resolution), and InSAR quality and availability. This extra information might fit better in chapter 3 after you go through the content of the input data.

Figure 3: The text and symbols are quite small, making it difficult to read.

Figure 4: The legend contains an entry for "not a rock glacier," which is not previously mentioned in the manuscript. It is acknowledged in Section 5.1 as potentially misinterpreted landforms, serving as an educational element. However, it may be more appropriate to exclude this annotation from the figures. Instead, incorporating it into attribute tables or supplementary materials like you have done could facilitate learning, so it might be useful to keep.

Chapter 4: Consider restructuring this chapter to begin with the description of study areas (Section 4.1 onward) before presenting Figures 5-8. This change could enhance the comprehension and interpretation of the figures.

I appreciate how challenges and uncertainties across regions and operators are summarized, reflecting the inherent difficulties in interpreting geomorphology in certain areas. I also like that you address some issues connected to how we traditionally have interpreted intact landforms as active, while in the kinematical and more recent definitions these are considered transitional or even relict. Maybe such observations could challenge the value we add to the current movement rates in high-arctic areas.

In general, I especially enjoyed reading the quality assessments of the different products mapped, i.e. the results of this exercise. I think you go through the different points carefully and thoroughly, and I gained some new insights while reading.

Map material:

To me this looks good, and with some help from the descriptions in the appendices it was quite easy to navigate in the mapped material.

From the text, I cannot read whether you did some “user sensitivity” tests or what to call it. It appears there are notable differences between the extended and restricted rock glacier outlines between some regions. Were systematic, regional differences in extended and restricted RG areas assessed? You mention this a bit in sec. 5.2.1. and 5.2.3., and maybe this issue is mainly addressed in the attribute tables as low outline reliability. However, when I had a look at one of the rock glacier outlines (RGU707506N277873E) in Finnmark where there is a rather large difference between the restricted and the extended outline of the front, the extended front position is marked with 2 (high reliability) while the restricted front position is marked with 1 (medium reliability), while both have 0 (low reliability) connected to their upslope margins. The uncertainties of the upslope margins are well accounted for in the text, but from the mapped material to me it looks like it is the front positions that are uncertain in this specific case. I had a look at the other rock glaciers in the same area, and the ones I had a look at seem to be classified in the same way. (I only looked into a few in the vicinity of RGU707506N277873E.)

Additionally, I am a bit confused by the discrepancies between rock glacier outlines and MA polygons. Could you clarify why these sometimes overlap and other times do not? While only "certain" rock glaciers are outlined, many polygons with MA values are neither marked as "uncertain" nor as "not a rock glacier" in the mapped material. Conversely, there are instances where RG outlines exist without corresponding MA polygons (e.g. Disko, Greenland).