This paper provides a comprehensive overview of the publication process of the EGU journals, with particular focus on the value and practice of open access and open review. The paper is a valuable documentation of the evolution of open access journals and engagement of authors, reviewer, editors and the general scientific community. I want to commend what the leadership teams of the EGU journals have done and congratulate them for the accomplishment in a relatively "short" period of ~25 years.

I only have a few minor comments listed below.

1. Figure 10 on page 34, the processing time between the acceptance date to publication date appears to be very long. I wonder why this is so.

2. It will be valuable to the readers if the authors can provide their perspectives on how the publication process and quality of the journals can be further improved. While the final section contains some discussions, they are mostly about the technical aspects. Readers will interested to hear about existing challenges and broader perspectives.

3. Line 217: "the" should be "they".

4. Line 265: "coudl" should be "could".

5. The sentence from 618 to 620 needs to be improved for better clarity.

6. Line 911, "Initially" is used in the preceding sentence. It can be removed in the second sentence. 

7. Line 1022: "references", do you mean "referees"?

8. Line 1139: "If data are not publicly accessible at the time of publication, the data statement must specify when and where they will become available and how readers can access them until then." Is there a follow-up to make sure that the authors indeed follow what they say?

General comments

This paper provides a detailed overview of the approach to open access publishing, preprinting and open peer review developed by the journal Atmospheric Chemistry and Physics, as well as other journals of the European Geosciences Union (EGU), in close collaboration with the publisher Copernicus.

I very much enjoyed reading the paper. It offers in-depth insights into the historical development of the open science practices of the EGU journals, the motivations for developing these practices, and the development of similar practices by other journals and publishers. The paper also provides detailed statistics demonstrating the level of adoption of the open science practices of the EGU journals.

Additionally, the paper shows that EGU has been an absolute frontrunner in this space. Nowadays there are many more journals and platforms adopting open science practices similar to those used by the EGU journals, and some of these journals and platforms are regularly praised for their contribution to innovation in scientific publishing. However, it needs to be acknowledged that many of these innovations were already in use by the EGU journals two decades ago.

The authors also make a strong case for open peer review: “Such disclosure of reviewer comments upon publication of final papers should be regarded a minimum standard for OA publishing, in order to counteract low scientific standards of (semi-)predatory and fraudulent journals that are solely motivated by the publishers’ financial interests. The mere post-publication of reviewer reports, however, inherently leads to a kind of bias and loss of information because only the reports of finally accepted papers are shown, whereas the reviews for rejected manuscripts are lost.” (l. 259-263). I fully agree with this argument, and I fully support the authors’ recommendation to “introduce appropriate elements of transparency and open peer review in all peer-reviewed OA publications” (l.1213).

Specific comments

l. 246, “concerns about the rigor of peer review by the Web of Science”: Web of Science evaluates journals to decide whether they can be selected for indexing in its databases. I think it is confusing to refer to this process as ‘peer review’. The process is quite different from what is usually understood as peer review.

l. 272, “referee reports can be entered to platforms like ORCID or Publons”: Is this correct? It seems that Publons doesn’t exist anymore. It has partly been integrated into the Web of Science platform.

l. 347, “whereas the former was fully sponsored by research funders, institutions and societies”: I don’t think this is correct. F1000Research applies an APC model.

l. 362-364, “For example, the French-led Peer Community in (PCI, 2016) facilitates open peer review of preprints deposited on the Episciences platform (CCSD, 2017) or in the French ’Hyper Article en Ligne’ repository (HAL, 2001).”: My understanding is that PCI accepts preprints from a large number of preprint servers, not only the ones mentioned in this sentence.

l. 366-367, “However, the small number of papers in the Peer Community Journals (∼100 per year, across a wide range of scientific disciplines) exemplifies its limited success and popularity.”: I think this sentence is inaccurate. Only a subset of the articles reviewed and recommended by PCI are published in the Peer Community Journal. What matters is not the number of articles published in the Peer Community Journal but the total number of articles reviewed and recommended by PCI.

l. 903-905, “The APCs in Table 3 are comparable to fees applied on publishing platforms, such as those managed by F1000 (F1000 APCs; Open Research Europe (2020), Section 2.3.2). However, on these platforms, no type-setting and copy-editing is applied unlike for final articles in EGU journals”: I am not sure if this is correct. It seems to me that F1000 does perform type-setting.

Technical corrections

l. 1022: I think ‘references’ should be ‘referees’.