The manuscript mostly concentrated on the Evaluation Run and its statistical properties by comparing the ocean model output with the available observations (SST, SLP, T). However, the manuscript missing a detailed analysis of the climate runs which is the main promise of the manuscript. The statistical analysis made for the Evaluation Run (Atmospheric forcings, Sea level variability and circulation patterns, Thermohaline properties, Extreme thermal events) should be repeated for all the climate runs.

As it is known, the open boundary conditions (OBCs) for the ocean models are critical, especially in the Adriatic Sea, the small differences in the salinity and temperature specified at the OBCs significantly affect the dense water formation and physical properties. In the manuscript, how the OBC data were generated to force the ROMS model is not clear, need to be specified the methods and justified that could be used safely in a climate model.

Review of the paper

AdriE: a high-resolution ocean model ensemble for the Adriatic Sea under severe climate change conditions

Authors: Davide Bonaldo, Sandro Carniel, Renato R. Colucci, Cléa Denamiel, Petra Pranic, Fabio Raicich, Antonio Ricchi, Lorenzo Sangelantoni, Ivica Vilibic, Maria Letizia Vitelletti

The manuscript deals with the present and the end-of-century, kilometre-scale ensemble modelling approach for the description of ocean processes in the Adriatic Sea using and ensemble of climate runs in a severe RCP8.5 scenario forced by the SMHI-RCA4 Regional Climate Model driven by CMIP5 General Climate Models as well as evaluation runs for the 1987-2010 period.

The text is well written and results are presented clearly.

The authors show that the main behaviour of the model used is ‘’satisfactory’’. However scenario simulations show results that necessitate a deeper investigation of the role of the model set up and of the forcing. The choice of the lateral boundary conditions is also of a crucial importance.

A deeper investigation of the relatively high theta and S changes in the deep Adriatic is recommended. This is also the case of the weak change of MHWs shown.

I recommend major revisions.

Specific comments:

-15. ‘’ with particularly encouraging results’’

Could authors explain to what extent results are encouraging?

-45. ‘’ (?Denamiel et al., 2021a) ‘’

Please correct if needed.

-50. ‘’ … ranging from the very evolution of the global…’’

Could authors verify this sentence?

-100.   ‘’Potential temperature (θ), salinity (S), momentum…”

Could authors describe how momentum is used in the boundary condition set up?

-105. ‘’…were modulated accordingly with the anomalies computed from Med-CORDEXderived CMCC-CMprofiles (Scoccimarro et al., 2011) in the norhtheasternmost grid cell of the Ionian Sea.’’

Could authors better describe the approach followed?

-185. ‘’Furthermore, although being the only available option the evaluation of SMHI-RCA4, ERA-INTERIM known to be far from the “perfect boundary conditions” hypothesis, particularly in terms of rainfall-related quantities (Bao and Zhang, 2013).’’

This sentence is rather unclear, could authors rephrase it. 

-210.  Whereas the comparisons shown in Fig.2, Fig.3 and 4 show that the model behaviour is rather satisfactory as stated by the authors: ‘’thus performing significantly better than most of the RCMs available for this geographical area’’, it would be interesting to illustrate this by one or two concrete examples.

-230. “This northbound improvement of the model skills suggests that internal dynamics partially compensate for the missing variability component in the boundary conditions.’’ ; “…the fairly good performance on the Northern Adriatic coast permits a more straightforward use in this region, also in terms of boundary conditions for local applications”.

-What is the sampling interval used in Fig.5?

-Are tidal oscillations included in Fig .5?

-The tidal amplitude is known high in the northern Adriatic; could authors discuss the impact of the tidal amplitude of the model performance shown in Fig.5.

-Again authors should further discuss the lateral boundary conditions.

-270. ‘’… , suggesting that SST does not show any macroscopic sign of a spurious drift related to the model implementation “.

 Could authors rephrase this sentence?

-280. ‘’ In turn, while intermediate to high S values are mostly well reproduced, low to mid salinity tends to be overestimated, particularly in the Kvarner Bay and in the Dalmatian Islands.’’

Please better discuss and explain the overestimation of the lowest S values.

-285. ‘’ This suggests that the climate ensemble, whose implementation began before the release of the latest version of MFS (Escudier et al., 2020), should not be considered prone to major elements of obsolescence associated with the use of a previous dataset (Simoncelli et al., 2019)”.

Could authors explain how this can be deduced from Fig.10.

-305. ‘’ In the deep Adriatic, an apparent tendency to underestimate average values of θ and S throughout the year is actually the result of some shortcomings in the description of thermohaline properties in the upper layers.’’

The sentence needs further explanation of the mentioned shortcomings.

-370. Please correct : ‘’hle200 m’’

-375. ‘’Below the upper layer, θ increase varies from + 2.8°C for h=200 m to +1.3°C for h≥800 m, S increase varies from +0.21 to +0.17, and σθ varies between-0.44 and-0.15 kg m−3.’’

Authors should discuss the relatively high values of salinity and theta changes in the deep Adriatic (shown in Fig.15c) and present comparison with results from previous work. Also, why two among the vertical profiles of the theta change are truncated at depths less than ~630 m,  (Fig. 15c left).

-385. ‘’Under this approach, modelled differences between SCE and CTR conditions (expressed as monthly mean cumulative intensity of the events) appear generally minor and in any case only occasionally statistically significant.’’

Here also, authors should mention results from previous work, if available.