Review comments to the paper ”DANRA: The Kilometer-Scale Danish Regional Atmospheric Reanalysis”, manuscript ID essd-2025-610.

General comments:

The paper describes the DANRA high resolution regional reanalysis and compares the output from this to the global ERA5 reanalysis. The latter is considered to be the current state of the art reanalysis so any added value is a measure of good quality. It is very nice to read a well written paper that discuss and describe a well made and good quality reanalysis since there is an added value in the DANRA data set compared to ERA5. Even though the manuscript is well written and mostly clear, there are a few questions that arise and also some editorial issues. I would recommend the paper to be published after these questions and comments have been answered and updated in the manuscript.

1. First a somewhat provocative question: What is the purpose of creating this reanalysis? And who are the intended users? The authors claim that it is an important data set that is asked for by both public and private bodies and the high resolution is important for local applications. It is also touched upon at the very end of the paper but not very much. It would be beneficial for the paper and for understanding the need for this reanalysis if a few real use cases were presented.

2. Another major thing is the observation usage. Since observations and the use of observations, is a very crucial part of a reanalysis, as the authors also point out, it would be nice with a bit more information on what observations are used and how. The background error statistics for example, how is that derived, for what period and is it the same during the entire reanalysis? A figure of how the number of observations changes during the period or how the coverage changes would also be nice to see.

3. Several acronyms are not defined in the paper (HARMONIE-AROME, ACCORD...). Please look through and define the acronyms the first time they appear.

4. In the manuscript, only three parameters are verified, temperature at 2 meters, wind at 10 meters and precipitation. This is of course interesting for the extreme weather events presented but from a model evaluation perspective it would be interesting to also see verification of other model parameters. I am confident that the authors have looked at this so what is the motivation for not including more verification in the paper?

5. The authors mainly refer to the higher resolution being the reason for the improvements compared to ERA5. Are the improvements only attributed to the higher resolution or are there other factors contributing to this, like additional observations, better parametrisations or similar?

Specific comments:

Lines 42-43: It says that HARMONIE-AROME Cycle 40h1 is the operational model in Denmark and Norway. That might have been true at the time but that information is old by now. I think that this should be stated by writing e.g. “the operational model at the time of the project start” or something similar. Besides HARMONIE-AROME Cycle 40h1 was used by several more countries than Denmark and Norway.

Line 45: It says “...improved modelling for the ‘cold’ climate condition…” but it is never mentioned what those improvements contain.

Line 195: “It would be faster if index files were pre-computed and GRIB files were uncompressed”. How much faster and what would this mean for storage space?

Lines 199-200: “users should not have to fetch and read too many time steps” and “too many grid points”. How many are too many? Please be more specific or reformulate.

Figure 7: DANRA always shows lower wind speed than ERA5. In the end of the period, when the bias is positive, this is good but during the first half of the period it is not good. Any explanation to why?

In addition, the transition from negative to positive bias occurs when the number of verification observations is constant. How do you explain this transition? Are there any other observations introduced in the data assimilation during this time?

Line 256: There is a ? after North Jutland.

Line 272: “...resulting in a markedly higher wind speed than CERRA.” It should most likely be ERA5 instead of CERRA.

Line 283: Another ?. Missing reference?

Line 284: I don’t understand what the authors are trying to say with this sentence? Please be a bit more specific.

Line 320 and Figure 9: The small dots marking the synop stations are really hard do see, especially in areas with high temperature (dark colours)

Line 321: “these variability” should probably be “this variability”.

Line 346: Summary and outlook. I can mostly see a summary and not really any outlook. Am I missing something or could this be extended e.g with some use cases as mentioned in the general comments or more specific ideas and plans for what this data set could be used for?

Review for the manuscript on “DANRA: The Kilometre-Scale Danish Regional Atmospheric

Reanalysis”

This paper presents a kilometre-scale  reanalysis system developed exclusively for Danish weather- and climate-related studies and tools developments. 

General comments:

The DANRA reanalysis modelling system and observations are mainly based on well-established modelling and observation infrastructures, such as Arctic, European and global Copernicus reanalysis systems. These systems are well documented. This paper describes specific modelling and observation processing that enhance the system over the Danish area of interest.  The Authors demonstrate the added value of the system through verification procedures and case studies by comparing DANRA-produced datasets with the above-mentioned reanalysis datasets. In the moment where data-driven model development is in focus, I found it important that DANRA analyses are made available for this purpose.

Depending on the development and studies' needs, more information about the upper-air datasets would be valuable and would make this paper more comprehensive. Although I agree that the added values from HARMONIE-AROME-based model systems are mostly in better description of the near-surface parameters and fields. 

 Minor comments/suggestions:

line 4:  ...climate reanalysis model or system that…

line 17: … weather and climate evolution or change

line 20: ...public sectors.

Line 34: I would define the model discretisation better by separating the horizontal and vertical discretisation. 

Line 75: describe what parameters are analysed/changed instead of “near-surface and soil properties”

line 92: I’d use another word than “extensive array” to make it more explicit

line 104: used in the ERA5 reanalysis and archived in MARS

line 116: what does it mean “CERRA (2018-2021)” I don’t understand this

line 128: in observation reporting via

line 130: I’d add a connecting word, like Meaning, the final … because the first sentence is hanging somehow…

line 133 : 1994, through …, were …    use commas

line 137: this looks to me a continuation of the previous sentence, so shouldn’t be in separate paragraph. 

Line 161: “exhibits significant drops in magnitude and at hourly intervals” I’d make it clear that the drops are observed every 3 hours and the datasets are hourly. 

Line 256: resolve the question mark. And instead of “wind speeds” “wind intensities”

line 269:  (Fig. 8c)

line 283: resolve the question mark

line 287: please find better wording than “most crucial elements in NWP” 

line 294: “had travelled” sounds a bit strange; I’d use other words. 

Line 300: “the time point when”, sounds strange to me

line 358: post-processing with shorter line

line 375: delete (Danish Reanalysis) since what DANRA is well explained above.