Please see comments in the attached document

The paper untitled “Investigating extreme marine summers in the Mediterranean Sea” - by Dimitra Denaxa Gerasimos Korres, Emmanouil Flaounas, and Maria Hatzaki – investigates Marine Heat Waves (MHW) during “extreme marine summers” in the Mediterranean Sea. The authors use a reanalysis to identify MHW characteristics and understand the main drivers explaining regional differences during particularly hot summers.

In general, I found the paper is well-written with a straightforward general structure. I really appreciated efforts of authors to clearly describe their results, even if some parts could be shortened to reduce the long length of the manuscript.

However, I have several concerns from the beginning of the paper regarding methodological choices:

1. I do not understand why the authors decide to only consider 4 events as “extreme marine summers”. Even if Figure 4 gives an idea about the distribution of summer SST, this is only at three locations. The authors need to give a more precise definition of what they call “extreme marine summer” regarding the distribution.
2. My second concern is a lack of justification about applying methodologies for each “grid point”. Results gives a very statistical point of view. Is there any criterion about spatial scales? This is confusing as they give physical interpretation associated with regional patterns. I would at least explain why they are not considering EMS composites.
3. If I understand Figure 3 correctly, it needs more explanation since patterns might results from one “very extreme” event for instance, and might not be representative of the 4 events. This also might be an issue in other figures.
4. It misses a discussion about the interest for the community of this new term “ESM”.

I think these points need to be addressed before introducing this new concept. A major revision is required to clarify these points.

Minor comments:

L128 – “derives from a combination of HadISST2 and OSTIA datasets” – ERA5 SST is based on a reanalysis which includes model and observations datasets. I would reformulate.

L133 – “In addition to ERA5 (reference SST dataset), we use the CMEMS L4 satellite SST product (EU Copernicus Marine Service Product, 2022c) for the period 1982-2019, at 0.05°x0.05° grid spacing, to cross check the quality of the reference dataset against a high-resolution observational SST dataset.” What did you compare? What did you conclude?

L143 – “We then define EMSs, separately at each grid point, as the four summers with the highest average JAS SST, i.e., exceeding the 95th percentile of the 71 available summer periods from 1950 to 2020. “ As the authors want to introduce a new concept, it needs more explanation and justification.

L239 – “we provide examples of different patterns for three grid points of the domain ». Why these points? Are they representative a regional patterns?

Figure 4 – This figure is interesting, but this arises questions. I would imagine with such “Extreme Marine Summer” concept, that the spatial scales would be close to those of the basin. Maybe, the authors should show several plot with seasonal SST anomaly during two or three typical ESM.

L627 – “These values indicate that oceanic processes are primarily responsible for the observed EMS SSTs in these areas “ – You often refer to oceanic processes, but can you give references that support this result.

Dear editor,

I apologise for taking so much time to review this paper.

The writers have investigated the concept of Marine Extreme Summers (EMS) in the Mediterranean Sea, their spatial distribution and potential drivers. In principle, the EMS idea sounds interesting, Despite the fact that the international literature on marine heatwaves (MHWs,a widely accepted framework for the identification of extreme temperatures in the ocean) has shown that a shift on the mean sea surface temperatures of the ocean is responsible for the development of extreme temperature in the oceans, the writers have tried to deal with the contribution and role of the SST variance in the development of extreme ocean temperatures in the Mediterranean. They have done a considerable amount of work in trying to prove the drivers of multiple EMS in the basin, concluding on the significant contributions from increased air-sea heat fluxes, reduced turbulent heat fluxes (mainly Latent) and winds, along with shallower than normal Mixed layer depth. Although this is a well known result for individual extreme ocean temperature events of the past in the Mediterranean Sea , it has not been shown before for multiple extreme events.

However, due to the length of the paper and the absence of coherence in its structure, writing and presentation of results, i could not a) get a clear message or outcome from it, b) understand exactly how the EMS are defined, or what is their usefulness/added value compared to the MHW definition. As i was approaching the description of the drivers behind EMS, i was striving to understand the methodologies used, i felt very often different methodologies were used to prove the same thing (e.g Sections, 3.3.1,3.3.2,3.3.3) without getting any new knowledge. In general, there was no structure in the paper which was discouraging, difficult and time-consuming to try to understand what the writers were doing most of the time, despite my expertise on marine extremes. I also find a lot of results are often repeated, some methodologies were only used in the beginning (e.g. the de-trending from the summer period) and then they were never mentioned for the reminder of the paper.

I find the paper badly written, and it should be re-written with the help from a native speaker. Since right now i could not get one clear message, rather many different results that very often were only inferred but not well-described or proven.For this reason I also provide a line-by-line review hoping that it will help the writers improve their manuscript. 

At the moment I advise against the publication of this paper, unless major revisions are performed in which case the concept of EMS should be better described and once there is a clear message, the idea should be highlighted. 

Regards