In my opinion, the major revision I'm requesting will improve the study by providing a clear physical meaning of the MHW variability in each individual period (historical and RCP 8.5).

Please see the pdf included in the zip file

Review of Far-Future Climate Projection of the Adriatic Marine Heatwaves: a kilometre-scale experiment under extreme warming” by Cléa Denamiel.

This work presents an analysis of marine heatwave characteristics in the Adriatic Sea under historical conditions (1987–2017) and a far-future extreme warming scenario for 2070–2100 based on RCP 8.5. 

I have two main comments about this paper. First, is that the author should better present, from the beginning, the methodology used (the Pseudo-Global Warming approach) together with the limitations of this approach. The description in the Methods section is rather succinct, and the presentation of what the limitations of this approach could be comes rather late in my opinion (it is mentioned in the introduction but the implications are not clear). My second comment is that there should be also a description of the differences between the historical and RCP8.5 runs, in terms of extrema, variability, intensity and trends. Without this we cannot put in context the differences observed between both runs presented in the paper. 

Here are my comments in order of appearance:

Figure 1. I'd limit the bathymetry colorbar to the depths found in the Adriatic. As it is now, we can clearly see Thyrrhenian bathymetry which is irrelevant, and the Adriatic is rather homogeneous

Line 125-129. I find it confusing what it is mentioned about the shifting of the RCP8.5 results: is this only for visualisation purposes? Why is this done? 

Line 147: "All the intensities": which intensities

Line 165: "deepest part": I'd suggest to add some bathymetry contours to Fig 2

Line 168: central eastern Adriatic: maybe refer to the pre-defined regions?

Figure 3. Add (left) and (right) to the description of panel (e) in the caption. The colors used in panels (a) and (b) are rather random. I would suggest to use shades of the same colours for each decade (e.g. shades of blue for the first decade, shades of orange for second decade and shades of purple for the last one. I would avoid red and green on the same figure). What are the dotted lines in panel (e)? (one is straight and the other seems to follow the histograms). The colours of the histograms represent the same thing as the numbers on the x-axis? and what about the height of the histograms? Because I do not understand why the yellow histograms for the monthly panel are higher than the RCP8.5 histograms but are still yellow.

Line 188. Here you mention for the first time that there are no MHWs for years 9-11 in both runs. This comes as a surprising fact, moreover since this happens in both runs, which makes it look suspicious that the PGW approach is not able to separate from the variability from the base/historical run. This is discussed in the Discussion section, but I think a brief comment (indicating this is fully described in the Discussion section) should be added here.

Line 200. Is the word skewed the correct one here? For me skewed indicates there is a bias/error in the data, and not just that it show a trend towards a specific value (which is what is mentioned here). 

Line 222. Are MHWs less frequent in RCP8.5 because they are longer? (i.e. there is less time with no MHW so less possibilities for a new MHW to develop)

Line 223-225. I do not understand what the author refers to in this sentence. If it is what I think I understand, doesn't a correlation of 0.9 show again that the PGW is not effectively separating events from the historical run from the RCP8.5?

Figures 5 and 6. Again, there is a dotted line which is not described in the caption, and no units in the y-axis of panels a.b.c (left). What is bottom temperature and OHC *intensity* ? 

Line 270. values of 0.25degC etc "above the climatology" I guess? 

Figure 7. There is no description in the text of how the "percentage of events primarily driven by air-sea fluxes" is determined.

Line 326. I would not call these other studies "independent observations" since they might use the same datasets. Just "other studies" would be more adequate.

Line 374 and following: how are the percentage of changes in the discharges decided? Why is this not discussed in the Methods section?

Section 4.2.2 Here the author mentions the Easter Mediterranean Transient as a possible cause of the absence of MHW in years 9-11, but no proof or demonstration is given, so this is just a suggestion provided by the author that needs to be verified. Line 427 says no previous work has linked EMT to MHW suppression, but this work doesn't provide the link either. And again, the fact that the MHW absence is also noticed in the RCP8.5 run should be better explained as it reflects in my opinion that the PGW fails to detach from the base variability. There is some mention in line 447 but this should be made clearer before.