The paper is well written, clear in its objectives and effective in its presentation of results. As acknowledged by the authors themselves, the sample used has some significant limitations, but these have been sufficiently discussed and adressed. The main value of the paper is that it provides important data on a relatively little-studied area such as the Mediterranean coasts of France and sheds light on the limitations of the warning channels that can be used for overcoming the so-called last mile issues. The results of the MCA could perhaps have been more clearly explained, but the reviewer acknowledges that this is a subjective assessment and that the paper can therefore be published without further changes.

The manuscript surveys the tsunami risk and tsunami alert perception of the population living or working in five areas on the French coast. The surveyed area is indeed defined within the perimeter of the tsunami evacuation areas. Four of the locations studied are in the south of France and one location is in Corsica. The five areas face the Mediterranean Sea on different sides. 

The issues addressed in the manuscript are essential to develop virtuous interventions and tsunami risk mitigation policies aimed at the local population and tourists as well as to implement more effective hazard and uncertainty communication tools. The study takes an international perspective and increases the data availability on tsunami risk perception and tsunami alert perception in the NEAMTWS area. 

The addressed issues are also of multilevel interest: local, national and European. 

However, the manuscript has many shortcomings that affect various aspects. 

A) conceptual gaps, 

B) methodological shortcomings, 

C) shortcomings in data analysis and output presentation 

D) inaccuracies and errors of form.

Specifically, some references cited in the text are not included in the bibliography. Further bibliography on risk perception and tsunami risk perception is recommended, paying particular attention to the complexity of the constructs and how they are addressed. 

L35 - there are references to specific data without cited bibliography

L64 and others - there are references to the manuscript by Carles et al., 2023 which is published in French. It is recommended to include other references rather than referring to a single native-language text, especially when dealing with important concepts or validation of flood zones with respective parameters. 

L68 - There is intense debate concerning the definition of "natural risks". It is suggested to inform about it. However, it is preferable to use 'natural hazards'. 

L71-94-101 etc. there are important generalised shortcomings on risk perception concept. This also emerges as a contradiction in terms in e.g. L94 and 95. The paper deals specifically with tsunami risk and alert perception, therefore, risk perception must be well structured. 

The authors also survey the perception (a sociologically complex and irreducible concept; Slovic, 1980) by considering only stimuli perceived through the senses. This is a theoretical/conceptual error. Therefore, the theoretical focus is shifted from perception to sensation, a psychological concept, see e.g. Graham and Ratoosh, 1962. 

Section 3.2.2; 4.2.3 and Figure 2. 

Major methodological and data analysis discrepancies emerge. Unfortunately, the appendix is not available to provide a direct comparison with the questionnaire administered to the respondents. However, the questions described in section 3.2.2 do not match the data presented in the graphs and commented in section 4.2.3. Some of the questions have been reformulated and do not accurately reflect the original text. Confusion between questions Q7 and Q8 where part of the text has been removed, losing its meaning.

Furthermore, the authors extracted categories from open questions (e.g. Q8 and Q10) without explaining the methodological procedure. The categories are not mutually exclusive and the data presented in this way are not statistically robust. To carry out such an operation, it is essential to describe the procedure, the instrument used and to present the categories. 

Since the authors have very interesting data and in-depth interviews, I suggest that the authors review the entire manuscript because, as mentioned above, it is of relevant importance and submit it once the shortcomings have been fixed.