EGU22-9101, updated on 28 Mar 2022
https://doi.org/10.5194/egusphere-egu22-9101
EGU General Assembly 2022
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

openEO Platform: Enabling analysis of large-scale Earth Observation data repositories with federated computational infrastructure 

Benjamin Schumacher1, Patrick Griffiths2, Edzer Pebesma3, Jeroen Dries4, Alexander Jacob5, Daniel Thiex6, Matthias Mohr3, and Christian Briese1
Benjamin Schumacher et al.
  • 1EODC, Wien, Austria (benjamin.schumacher@eodc.eu)
  • 2ESA, Frascati, Italy
  • 3University of Münster, Münster, Germany
  • 4VITO, Mol, Belgium
  • 5EURAC, Bozano, Italy
  • 6Sinergise, Ljubljana, Solvenia

The growing data stream from Earth Observation (EO) satellites has advanced scientific knowledge about the environmental status of planet earth and has enabled detailed environmental monitoring services. The openEO API developed in the Horizon 2020 project openEO (2017–2020, see https://openeo.org/) demonstrated that large-scale EO data processing needs can be expressed as a common set of analytic operators which are implemented in many GIS software or image analysis software products. The openEO Platform service implements the API into an operational, federated service currently running at back-ends at EODC and VITO with access to SentinelHub data to meet processing needs of a wide user community.

openEO Platform (https://openeo.cloud/) enables users to access a large collection of open EO data and perform scientific computations with intuitive client libraries simplifying underlying complexity. The platform is currently under construction with a strong focus on user co-creation and input from various disciplines incorporating a range of use-cases and a free-of-charge Early Adopter program that allows users to test the platform and to directly communicate with its developers. The use cases include CARD4L compliant ARD data creation with user defined parameterisation, forest dynamics mapping including time series fitting and prediction functionalities, crop type mapping including EO feature engineering supporting machine learning based crop mapping and forest canopy mapping supporting regression based fraction cover mapping.

The interaction with the platform includes multiple programming interfaces (R, Python, JavaScript) and a browser-based management console and model builder which allows a direct, interactive display and modification of processing workflows. The resulting processing graph is then forwarded via the openEO API to the federated back-ends.

In the future users will be able to process continental-scale EO data and create ready-to-use environmental monitoring services with analysis-ready data (ARD) and predefined available processes. This presentation will provide an overview of the current capabilities and the evolution roadmap of openEO Platform. It will demonstrate the utility of the platform to process large amounts of EO data into meaningful information products, supporting environmental monitoring, scientific research and political decision-makers.

How to cite: Schumacher, B., Griffiths, P., Pebesma, E., Dries, J., Jacob, A., Thiex, D., Mohr, M., and Briese, C.: openEO Platform: Enabling analysis of large-scale Earth Observation data repositories with federated computational infrastructure , EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9101, https://doi.org/10.5194/egusphere-egu22-9101, 2022.