In the manuscript the authors describe the construction of a crustal model of the Dinarides starting from the combination of different datasets and analyze the main properties of the model. The construction of the model and the dataset used are accurately described giving the right relevance to the uncertainties of the inputs and of the resulting model. Finaly the authors show the improvement in travle time computation due to the application of their crustal model. I think the manuscript requires only minor corrections and clarifications specified in the attached file.

The manuscript “Referent seismic crustal model of the Dinarides” presents a compilation of geophysical and structural models in the extended area of the Dinarides with the aim of creating a 3D geophysical model of the crust, which is the first step in defining the complex structure of the crust in this area. The background, study area and model construction are described thoroughly. The authors analyse the main crustal interfaces, P- and S-wave velocities, and density of the derived model with emphasis on the uncertainties. At the end, the authors show the relevance of the new model by presenting improvements in travel-time calculations compared to the 1D model. I think that the authors have described their model realistically and in detail, while pointing out certain shortcomings of this model themselves.

I consider the manuscript suitable for publication with some minor corrections and clarifications.

You will find attached some comments, questions and suggestions.

General comments

With this paper the authors present a 3D model of the larger Dinarides region which provides insights into the variation of P-wave velocity (plus S-wave velocity and density) in the sedimentary cover, the crust and uppermost mantle. Knowing about the vertical and lateral configuration of these mechanical properties is essential for a reliable localization of seismic events. For the construction of the 3D model, the authors have integrated various datasets and applied sophisticated interpolation methods to arrive with a gapless representation of parameter configurations (including uncertainty estimates). To validate their resulting 3D model, the authors finally compare measured travel times to those that would be predicted by the 3D model. Due to an improvement in the misfits between measured and modelled travel times as compared to the misfits associated with a standard 1D velocity model, the authors conclude that the 3D model is reliable and useful for future earthquake relocations.

This summary of the main messages of the paper shows that it indeed addresses relevant scientific questions within the scope of SE. The novelty of the paper emerges mainly from the presented maps and profiles which could not have been generated based on any of the input datasets alone, highlighting the importance of the comprehensive data integration approach. The conclusions drawn are substantial and clearly articulated in the corresponding chapter (which is one of the better chapters in terms of writing style and language as compared to other chapters in the manuscript). The title is well chosen and reflects the work done (though I would call it a “reference” seismic crustal model). Certain weaknesses in the descriptions of the methodology, results and interpretation are listed below in the “Specific comments” section. The Discussion chapter misses a critical discussion of the data integration processes (see my comments on this chapter below). A major weakness of the paper regards its presentation style using a structure and line of arguments that have to be revised to avoid misunderstanding and more concisely articulate the authors’ main messages (see suggestions for a new structure of the paper below). Concerning the language, a professional proofreading is recommended. Given the large number of flaws, the “Technical corrections” section below does not specify all of them individually (e.g., where the article “the” is missing and where it should be removed).

Specific comments

Abstract

The abstract does not ideally present the objectives, methodology, main results and implications of the study performed. This could be improved a lot to attract more readers.

Introduction

This chapter misses the clear statements of (i) which problem is at hand, (ii) why the authors are motivated to solve it (purpose of the 3D model!), (iii) which general strategy they have chosen to overcome the problem. Since the envisaged purpose of the model remains unclear in the current version of this chapter, it is impossible to assess and judge the quality of the modelling approach chosen.

Tectonic and geological setting

This chapter should be shortened, i.e. reduced to information that is relevant for the understanding of the modelling approach and later discussion of results. More detailed information about lithological characteristics of the model units (Neogene, Carbonate Complex, crystalline crust) would be required to assess the reliability of the rock physical properties of the final model.

Data / Model construction

The chapters “Data” and “Model construction” are not well structured (e.g., “Data” already includes information on the model construction). Furthermore, the original chapters simply fail in putting someone in the position to correctly reproduce the 3D model when using the same data - which is a primary requirement for a model to be reliable and acceptable. Therefore, for the sake of more clarity and less repetition in the text, I suggest to merge the two chapters into one called “Modelling approach” which would have the following structure:  First, describe and justify the intended differentiation of the model into four discrete model units (Neogene, Carbonate Complex, crystalline crust and mantle). Then, introduce the technicalities of the model building process (e.g., the interpolation method). Finally, present a sub-chapter for each model unit to refer to (i) the used input data sets, (ii) their processing (including equations, uncertainty assessment method) and (iii) all decisions taken by the authors including explanations (e.g., how to compensate for observational gaps; why and how to recalculate sediment thicknesses, etc.). For the later model evaluation process (Discussion), each model unit should already here come with a map showing the locations / distributions of all data that have been integrated into the modelling. Furthermore, this chapter should include all methodological aspects of the model validation process through travel time predictions (in a final sub-chapter).

In general, the authors should avoid to phrase interpretations in this results chapter.

In the submitted version of the manuscript, it remains unclear whether uncertainty has been assessed (calculated using equations – which?) or merely taken over from the original data sources.

The main characteristics of the model

It is not clear how useful it is to present maps showing S-wave velocity and density. In my opinion, they could be removed since they are the result of simple (empirical) P-wave velocity conversions. The resulting maps do not contribute anything to the main conclusion of this study.

Discussion

As an important point of the critical discussion of model limitations, the authors should comment on the fact that they have jointly interpolated P-wave velocities from active seismic profiles and those derived from gravity constrained density values. Is this reliable? There seems to be a systematic jump in velocity when crossing the boundary between the different domains. Does this mean that such abrupt changes in the interpolated maps do not reflect real differences in rock physical properties but inconsistencies due to different methods?

Technical corrections

General: check and synchronize the order of figures in the paper and in the text

Line 33: refer already here to the map of Figure 1 to introduce the location of the study area

Figure 1: add the exact position of all active seismic profiles as this is the most important type of data for the study

Please consider also the detailed comments and suggestions in the uploaded PDF.