Agriculture Cluster Webinar Discussion and Q&A

Thank you for attending the EarthCODE webinar and contributing to such a lively and insightful discussion!

The meeting focused on how to ensure long-term availability and findability of data produced by ESA-funded Agriculture Science Cluster projects. Attendees discussed challenges surrounding data sharing, storage, and the importance of FAIR (Findable, Accessible, Interoperable, Reusable) principles. The session went into detail about EarthCODE - ESA’s Earth Science Collaborative Open Development Environment, the benefits and possibilities of using integrated platforms.

The presentation covered the following topics:

  • Explore the EarthCODE and learn how to perform your research using various platforms and share research data, even during the early stages of your project.
  • Discover the tools you can access via the EarthCODE portal.
  • How to engage with the community and maximize EarthCODE’s resources.
  • Insight into future developments, including DOI assignment, documentation, and user contributions.
  • Share your feedback on how EarthCODE can better support your work and ask questions during open discussion with EarthCODE team!

On the back of this we created this Discourse topic for the Agriculture Science Cluster to continue the conversation!

Q&A

Q: Where can I deposit my CSV files +1+1?
A: If you are referring to the result files, you can store these on the ESA Project Results Repository and store metadata about these on the Open Science Catalog.


Q: With ESA access to other platforms, can we use the full potential of GEE?
A: GEE is not currently an EarthCODE platform, and there are no short-term plans to integrate it. However, you can still use GEE and then manually publish your datasets and workflows in EarthCODE. We’ve also noted that several requests to integrate GEE were raised during the webinar.


Q: Can we also add pretrained models? Or anonymised data (e.g. extracted spectral information together with yield) in cases where data cannot be shared as it is?
A: FAIR does not mean Open. So yes, pretrained or anonymised data can be added. But as mentioned above with GEE, this may not be fully streamlined as we favor FAIR and Open data sharing. Please consider adding some “sample datasets” if possible to enable others to easily reproduce and reuse (with the sample data).


Q: How does this initiative compare to GEO Knowledge Hub? How can we avoid that projects will need to put effort into increasingly more platforms to register their project, results, algorithms, …?
A: This is indeed another separate initiative. Once your data and workflows are FAIR, it is straightforward to expose the metadata in different ways (e.g. in different registries).

EarthCODE is meant to be used for all ESA-funded projects, and integration with EarthCODE platforms makes it easier to reproduce and reuse existing datasets and workflows. EarthCODE provides more than just a repository; it provides an ecosystem and integrated platforms for doing open science easily.


Q: When is the release planned? When can we schedule its integration into current or future projects?
A: The integration of the first “set” of platforms into EarthCODE will be available very soon (by the end of June). The interoperability between EarthCODE platforms will come a bit later (everything should be ready by September). Afterward, the EarthCODE project will regularly integrate new platforms, so additional releases will come later next year.

For updates, make sure to follow:


Q: How can we integrate a platform? Can we? What are the prerequisites?
A: Tenders are published regularly, and to integrate a new platform you would need to reply to one of these tenders. The previous best-practice procurement tender ended earlier in April (read about it here).

Integration of platforms includes several items (non-exhaustive list):

  • Implement Single Sign-On (SSO) using the EarthCODE identity system (GitHub-based)
  • Support automated publishing of workflows and data to the EarthCODE Open Science Catalog
  • Integrate with the ESA Project Results Repository (PRR) for long-term data storage
  • Ensure that experiments published in the Catalog can be executed within the platform
  • Register and onboard services in the NoR portfolio for eligibility and discoverability
  • Provide documentation for using the platform and its integration with EarthCODE
  • A Service Level Agreement is required

Q: I recently had a call with the APEx team and already raised the question. What is the relationship with these 2 projects (EarthCODE & APEx)? Could we say that they refer to projects with a different degree of maturity?
A: The connection is that once a workflow is published in EarthCODE, APEx can pick it up for scaling and operationalization. APEx is more about creating scalable production-ready services, while EarthCODE is more about cataloging results from projects.


Q: Will regional-scale analysis be easier through EarthCODE integration with various platforms? It’s really a challenge with Sentinel-2 data processing at large scale.
A: Yes, the platforms integrated in EarthCODE provide data access at scale in a cloud-native environment, where your algorithms run close to the data. Scalability is improved when using the platforms.


Q: Is there any way to get support when integrating our project into these platforms? +1
A: Yes, there are several options for this:

  1. The platforms are well-documented, and we encourage you to explore them.
  2. Platforms provide training and support; you can request this through the platform’s resources.
  3. The best way to start is to join our Discourse channel or get in touch with us directly if you need guidance.

Q: Regarding NOR and potential funding, should the project be ESA-funded or are other projects (from the cluster) also eligible?
A: It is not mandatory, but ESA-funded projects get higher priority.


Q: Will the EarthCODE ID be project-specific, or open to many users?
A: Published metadata on EarthCODE is open to the public.


Q: Would it be possible to move credits from one service to another? For example, the project got a certain amount of credit for DeepESDL, but during the process it looks like openEO would be good for certain tasks?
A: It is not possible at the moment, but we aim to enable it. It is a bit difficult because each platform is managed by different providers/companies.


Q: The tendency of ESA to cluster all important project meetings, cross-project meetings, tutorials, courses, demonstrations, trainings, etc. in one week of LPS is challenging for participants. I strongly encourage keeping such an in-depth training for a moment outside LPS.
A: Thank you for your comment. The training event on EarthCODE at LPS will take place on a separate day, before or at the beginning of the conference.