Overall, I appreciate this content of this paper / study and the work that the authors did in it and in writing it up. I think this is a valuable resource for the software and information science community, and hope that it will be circulated more widely than in just in the Earth science community.

In particular, Section 3.1 is quite useful.

General comments:

The authors could mention CIG (https://community.geodynamics.org) somewhere as complementary to the work studied in this paper.

Adding blank lines to separate new paragraphs would be helpful. This is done in some parts of the paper, but not others. In particular, it would be helpful in the references section.

I find some of the terminology here a little confusing. When I heat models today, I think of machine learning or AI. The models here are more modeling/simulation programs or functions. Note that the FAIR4RS principles are about Research Software. There is also a group working on FAIR4ML, where ML is machine learning models. If the authors want to keep using "models", it should clearly be defined at the start.

Similarly, when looking at Figure 2, it's unclear to me if "publish data model and records" is discussing the software or the data that it produces. And in the F row, what metadata is being discussed? Metadata about data or metadata about the software? This is made more clear in the paper text, but the figure/caption could also be clarified.

Specific comments (with line numbers):

46 - perhaps mention https://www.researchsoft.org/tf-actionable-fair4rs/

75 - Please add the names of the two model catalogs to the caption

332 - I strongly disagree with idea here that using a CC0 dedication/license is appropriate. While Creative Commons says this can be done (https://wiki.creativecommons.org/wiki/CC0_FAQ#May_I_apply_CC0_to_computer_software.3F_If_so.2C_is_there_a_recommended_implementation.3F), it mentions that OSI does not approve this, and given that many projects consider the use of an OSI-approved license the definition of open source, software that has a CC0 dedication may not be considered open source.  Also, even CC says that CC0 is a dedication, not a license.  See https://opensource.org/blog/public-domain-is-not-open-source for OSI's view.

455 - it might be worth mentioning LLMs here, as this is the technology that is being most tested for this purpose.

551 - JOSS could be cited - One recent paper is https://doi.org/10.31274/jlsc.18285 (Note that I am an author of this.)

577 - It would be useful to mention SciCodes.org here, and for the authors to participate in it if they don't already, or at least to make sure that their lessons get back to that community.

Summary and feedback:

The manuscript demonstrates how the FAIR4RS principles can be applied to numerical models, offering a clear and practical framework for other modeling initiatives to follow. ​The paper compares the Community Surface Dynamics Modeling System (CSDMS) Model Repository and the U.S. Geological Survey (USGS) Model Catalog and provides valuable real-world case studies that highlight both successes and challenges in implementing FAIR principles. ​ This comparison allows readers to understand the practical applications and benefits of the FAIR principles in different organizational contexts. ​Additionally, the paper emphasizes the importance of community engagement and user feedback in developing solutions that are tailored to the needs of the scientific modeling community.

Comments:

• From the perspective of a model user, easy installation and clear instructions for the setup and execution are critical for effective reuse. Even when models are openly available, complex installation procedures or incomplete documentation can make them difficult to adopt in practice. Users benefit greatly from straightforward installation processes, well-documented dependencies, and clear guides for running and adapting the model. Including example workflows or usage demonstration can further help new users understand how to use the model. Emphasizing how these aspects are done by the two catalogues will strength the discussion of the model adoption and reuse (Reusability)
• In section 2.4, the paper could include more examples of how models have been reused across projects or disciplines to illustrate the practical benefits of FAIR principles. The example could include: workflow, required adaptations, and the time saved compared with a “pre‑FAIR” approach.
• The paper has omitted in the “Interoperability & reusability” discussion the portable images using containerization technologies like Docker and Apptainer. These tools are essential for enhancing the FAIR aspects and enabling easier access and functionality for model users. These technologies facilitate the creation and sharing of reproducible research environments. For instance, Docker allows user to package applications with their dependencies into portable containers and makes them accessible across various systems. Similarly, Apptainer emphasizes security and use in high-performance computing context.

Recommendation:

The manuscript is well organized and easy to read, and offers a valuable, community‑oriented account of how FAIR‑4RS can be used for Earth‑surface models. I recommend minor revision based on my previous points.