The paper is well-organized, with distinct sections explaining cloud computing concepts, service models, deployment models, and their relevance to Operational Ocean Forecasting. It provides a comprehensive overview of cloud computing, including its essential characteristics and how they apply to scientific and operational oceanography. The inclusion of real-world examples, such as NOAA's Open Data Dissemination Program and the Copernicus Service, adds value by grounding theoretical discussions in practical applications. Additionally, it mentions important technologies for cloud-native development, such as Docker, Kubernetes, and HPC-focused container platforms which are crucial in scientific computing environments.

Nevertheless, there are some areas of improvement. The paper focuses heavily on the benefits of cloud computing, but as I will discuss in my comments, it would be valuable to address some downsides or challenges. For instance, data security, cost management, and performance issues in high-performance computing (HPC) settings should be mentioned, as these are relevant concerns for any organization or scientific body adopting cloud solutions. There are some lack of technical precisions that I also comment below. 

The paper could benefit from more discussion on emerging trends like AI/ML integration with cloud computing in operational oceanography, which is becoming increasingly important for predictive models and real-time analytics. In this sense, there are some initiatives happening using EGI such as the iMagine project. 

You could also point out the collaborative aspect of cloud computing. Many cloud-based projects in oceanography, like Copernicus and NOAA, focus on data sharing and collaborative research via Virtual Research Environments. Cloud platforms enable large-scale collaborative environments where multiple stakeholders can work with shared datasets and tools, thus improving international collaboration and research outcomes.

Another topic I am missing in this paper is data interoperability and FAIR principles. A key challenge in oceanography is ensuring that large, distributed datasets follow FAIR principles (Findable, Accessible, Interoperable, Reusable). Cloud computing can enhance data interoperability through standardization of formats, APIs, and access protocols, ensuring that datasets can be easily shared, accessed, and reused by researchers globally.

The tables 1, 2, 3 and 4 provide valuable definitions, models and patterns. The main text could do a better job of integrating and leveraging these definitions more effectively to support the overall analysis. The main text mentions these characteristics but doesn’t always connect them directly to specific use cases in ocean forecasting or other scientific applications. For example, when discussing data storage and management, it could explicitly reference the measured service characteristic to highlight how cloud providers charge based on resource usage. Similarly, the broad network access characteristic could be tied to how cloud services enable global access to oceanographic data from remote locations.

Including an overall landscape of cloud technologies in oceanography would add significant value to the paper. It would provide readers with a broader view of how various cloud technologies are applied across the oceanographic field, helping them understand the diversity of tools, platforms, and strategies currently in use. In fact, a landscape analysis would give readers a complete picture of the range of cloud-based tools, platforms, and applications being used for different tasks in oceanography. This could include everything from data collection and storage to forecasting models, visualization, and collaborative platforms. By presenting a landscape, you can highlight emerging trends (as I suggested earlier) in the field, such as the growing use of AI and machine learning, edge computing, and serverless technologies. This would position the paper as forward-thinking and relevant to ongoing technological advances. You could include a dedicated section (e.g. “Cloud Technology Landscape in Oceanography”), summarizing the technologies used at each stage of oceanographic data collection, analysis, and dissemination. Visual aids such as diagrams or tables could map out which cloud platforms are used for different tasks, helping readers see how different cloud technologies fit into the broader landscape of oceanography.

Attached, I provide some comments related to specific parts of the text.

The document is very well written, with very clear English that it is easy to follow and understand.

The manuscript presents a basic state of the art of the cloud concept and cloud providers angled to modelers to inform them with the advantages of running their models in a cloud environment.

The manuscript contains enough relevant references.

My only concern is the manuscript is not particularly innovative or exciting; presents cloud concepts that have been used for almost two decades. I recognize that the target audience may not be familiar with those concepts, and there is nothing wrong with presenting them again if they lead somewhere. And that is the biggest problem with the manuscript; while I see where the authors are leading the reader, there is are no enough strong arguments in section 2 to properly inform or convince a modeler that the cloud approach is the right one.

I am not against the publication of this work, but I would like to ask the authors to revise their work;

The introduction should introduce the problem they are trying to solve, not just the concept of the cloud. The content of the current introduction is relevant, but reading it I don't understand the problem the paper tries to address.

Section 1 is fine, informative. The authors talk about Linux containers in page 6, but I believe they are trying to describe container technology, which is not just used for Linux, they can be used for any OS. Linux containers is an umbrella term used for container technologies under Linux. Please clarify this potentially swapping Linux containers for just containers. In the same page the authors claim that Docker has not made strides into the HPC world due to technical limitations. There is no text or reference to substantiate such claim.

Section 2. This should be the main part of the manuscript where the authors should work more, better articulating how modelers can leverage cloud technologies. The two examples provided are relevant but reading them they are just informative, there is no clear narrative giving the reader a cohesive view. This could be address with some extra text after the three examples.

There is no closing section; what have we learnt?, what are the future cloud technological developments and trends the reader should keep an eye on?.

I am more than happy to provide the authors with further comments if they have any questions.