Sustaining the virtuous loop

In the previous activity, you explored the components that make up the end-to-end ocean forecasting system. But these components do not work in a straight line — observation in, product out, and then stop. Instead, they form a virtuous loop: the products the system delivers, together with feedback from the people who use them, flow back to improve future forecasts. This article looks at what keeps that loop turning — how ocean data is managed, shared, quality-checked, and continuously refined.

Data management: the backbone of the loop

For the loop to function, the data moving through it must be well managed. A robust data management strategy is crucial for handling ocean predictions and developing advanced operational services (Tanhua et al., 2024). Without it, data would be difficult to locate, combine, or reuse — and the loop would break down before it could deliver value to users.

The FAIR principles

To be useful across the whole loop, ocean data should be produced in agreement with the FAIR principles (Wilkinson et al., 2016):

• Findable — data should be easily discoverable through metadata and searchable resources
• Accessible — data should be readily available to users, allowing for collaboration and reuse
• Interoperable — data should be compatible with other systems and formats, enabling seamless integration across different platforms
• Reusable — data should be freely available for use in a wide range of research and applications, promoting innovation and knowledge sharing

FAIR data is what allows one part of the loop to build on another, and what lets a service developed in one region be adapted and reused somewhere else.

Tools for processing and production

Turning raw model output into usable ocean information requires the right tools — software that helps analyse, visualise, and process oceanographic model data. Widely used examples include:

• OceanSpy — for high-level analysis and visualisation of ocean model datasets
• XGCM — for general circulation model post-processing
• Xarray — the core package for working with labelled, multi-dimensional arrays
• Cartopy and Matplotlib — for visualisation and mapping

These tools make it possible to move efficiently from numerical results to products people can interpret and act on.

Standards, best practices, and dissemination

Producing fit-for-purpose services also depends on the adoption of standards and best practices, so that data and methods are consistent and trustworthy across different centres and countries (Pearlman et al., 2019). Finally, the loop needs reliable protocols for disseminating products — for example through data stores or FTP — so that finished ocean information actually reaches the users who need it.

Assessing reliability: validation and intercomparison

How do we know a forecast can be trusted? This is where the loop feeds back on itself. To assess service reliability, forecasting systems adopt protocols for ocean data validation — comparing predictions against observations (Sotillo et al., 2025) — and for model intercomparison — comparing different systems against one another (Hernandez et al., 2025). The results of this assessment reveal where a system performs well and where it needs improvement, directly shaping the next cycle of forecasts.

Closing the loop: users, capacity, and outreach

The loop is only virtuous if it responds to the people it serves. The Copernicus Marine Service, for instance, evolves its operational services in response to both users’ needs and advances in science — a clear example of user feedback driving system improvement. Sustaining this cycle also depends on dedicated capacity development and outreach, ensuring that more people worldwide can access, understand, and contribute to ocean forecasting (Diarra et al., 2025).

Why it is a loop, not a chain

Each element here — data management, FAIR principles, tools, standards, validation, and user feedback — does more than pass information forward. Together they send information back: validation reveals what to fix, users signal what they need, and capacity development brings new contributors in. This is what turns a one-way value chain into a continuously improving virtuous loop.

Key takeaways

• A strong data management strategy, built on the FAIR principles (Findable, Accessible, Interoperable, Reusable), is the backbone that keeps ocean data usable across the loop
• Dedicated tools, shared standards, and reliable dissemination protocols are what turn raw model output into fit-for-purpose services
• Validation and model intercomparison assess reliability and feed improvements back into the system
• User needs, capacity development, and outreach close the loop — making it virtuous, because each cycle improves on the last

References

• Wilkinson, M. D., Dumontier, M., Aalbersberg, I. J., et al. (2016). The FAIR Guiding Principles for scientific data management and stewardship. Scientific Data, 3, 160018. https://doi.org/10.1038/sdata.2016.18
• Tanhua, T., et al. (2024). An overview of the ocean data ecosystem. Ocean Science, 21, 3131–3164. https://doi.org/10.5194/os-21-3131-2025
• Pearlman, J., et al. (2019). Evolving and Sustaining Ocean Best Practices and Standards for the Next Decade. Frontiers in Marine Science, 6, 277. https://doi.org/10.3389/fmars.2019.00277
• Sotillo, M. G., et al. (2025). A description of validation processes and techniques for ocean forecasting. State of the Planet, 5-opsr, 16. https://doi.org/10.5194/sp-5-opsr-16-2025
• Hernandez, F., et al. (2025). A description of model intercomparison processes and techniques for ocean forecasting. State of the Planet, 5-opsr, 17. https://doi.org/10.5194/sp-5-opsr-17-2025
• Diarra, A., et al. (2025). Capacity development for the future of ocean prediction. State of the Planet, 5-opsr, 26. https://doi.org/10.5194/sp-5-opsr-26-2025