Open Earth Observation and fediverse-type systems as a basis for innovation, competitiveness, permanent monitoring of the European environment and green transition

Open-Earth-Monitor Cyberinfrastructure (OEMC) | Grant agreement ID: 101059548 Open Earth Observation and fediverse-type systems as a basis for innovation, competitiveness, permanent monitoring of the European environment and green transition Open-Earth-Monitor Cyberinfrastructure project report and policy brief Authors: Tom Hengl1, Gregory Duveiller2, Johannes Reiche3, Steffen Fritz4, Martin Herold5, Edzer Pebesma1,6, Simone Sabbatini7, Luca Brocca8 Affiliations: 1 OpenGeoHub foundation, Doorwerth, the Netherlands 2 Max Planck Institute for Biogeochemistry, Jena, Germany 3 Wageningen University and Research, Wageningen, the Netherlands 4 International Institute for Applied Systems Analysis (IIASA), Laxenburg, Austria 5 GFZ Helmholtz Centre for Geosciences, Potsdam, Germany 6 IfGI University of Münster, Münster, Germany 7 CMCC Foundation (Euro-Mediterranean Center on Climate Change), Viterbo, Italy 8 National Research Council, Research Institute for Geo-Hydrological Protection (CNR-IRPI), Perugia, Italy Suggested citation for the full report: Hengl, T., Duveiller, G., Reiche, J., Fritz, S., Herold, M., Pebesma, E., Sabbatini, S., Brocca, L. (eds, 2025). Open Earth Observation and fediverse-type systems as a basis for innovation, competitiveness, permanent monitoring of the European environment and green transition. OpenGeoHub foundation, Open-Earth-Monitor Cyberinfrastructure project report and policy brief, DOI: 10.5281/zenodo.17952431. Funding: The Open-Earth-Monitor Cyberinfrastructure project has received funding from the European Union's Horizon Europe research and innovation programme under grant agreement No. 101059548. Disclaimer: Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or European Commission. Neither the European Union nor the granting authority can be held responsible for them. The information is provided “as is”. Open-Earth-Monitor Cyberinfrastructure (OEMC) project consortium and its suppliers and licensors hereby disclaim all warranties of any kind, express or implied, including, without limitation, the warranties of merchantability, fitness for a particular purpose and non-infringement. Use of AI in Writing This Report: Gemini Pro and Grok were used to run literature review and locate news sources and fact check data. Gemini was used to correct English and generate a list of abbreviations. Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org Scope This report and policy brief was produced by the OEMC project Executive Board and represents views and opinions of the authors. It aims to support Decision-makers especially within European organizations and international organisations looking for guidelines on: how to use open EO data to support new businesses especially Small Medium Enterprises (SME’s)? which cyberinfrastructures and data and software choices to follow, what are the current bottlenecks of upscaling open EO projects such as Copernicus Sentinel? We also try to answer practical and actual questions such as: what can AI technology do and what it cannot do to speed up land restoration and green transition? Should European decision-makers worry about AI technology controlled by a few large corporations from silicon valley? The policy brief is divided in the following sections: ● Open-Earth-Monitor Cyberinfrastructure (OEMC) project and key project outputs, ● Backgrounds: the rationale for open EO, ● How can open EO data and CDSE speed-up Green transition? ● Key modern technical challenges with processing (Petabytes) of EO data and potential solutions, ● Modern geospatial data solutions and modern state-of-the-art cyberinfrastructures for EO data, ● Emergence of the GeoAI technology and Earth Intelligence (and what to expect in the years to come), ● Next milestones for the OEMC project in 2026. A 4-page summary policy brief is provided as attachment to this report. Read more about the OEMC project from the CORDIS Results in brief: “Easy monitoring of Earth’s biophysical indicators at multiple scales”1. Activities and Impact ● OEMC organizes annual Open-Earth-Monitor Global Workshops (e.g., 2023 in Bolzano, Italy; 2024 with IIASA; 2025 planned in Barcelona and focusing on Green Deal monitoring). ● OEMC engages stakeholders through interviews, use case demonstrations (e.g., forest carbon monitoring, pasture productivity via Global Pasture Watch integration). ● OEMC emphasizes open-source solutions to address data silos, improve accessibility, and enable global-scale environmental tracking. Abbreviations and terms used in the text: ➔ AGB — Above Ground Biomass, usually measured in t per ha or kg per m2. ➔ AI — Artificial Intelligence. ➔ ARCO — Analysis Ready Cloud Optimized: used in the context of "Landsat archive 1997–2024 bimonthly mosaics ARCO added to CDSE". ➔ ARD — Analysis-Ready Data. ➔ CAP — Common Agricultural Policy. ➔ CDSE — Used throughout the text, referring to the infrastructure for Earth Observation data (e.g., Copernicus Data Space Ecosystem). ➔ CMCC — Centro Euro-Mediterraneo sui Cambiamenti Climatici (an author/partner affiliation). ➔ CNR — Consiglio Nazionale delle Ricerche (project partner). 1 https://cordis.europa.eu/article/id/457195-easy-monitoring-of-earth-s-biophysical-indicators-at-multiple-scales Page 1 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org ➔ Copernicus Sentinel — EU Space Programme is an EU programme with entrusted entities such as ESA, EUMETSAT, ECMWF that implement it. ➔ DANA — The name given to a severe storm event (DANA storm of October 29, 2024). ➔ DGGS — Discrete Global Grid Systems: a recursive partitioned mosaic that covers the entire Earth's surface. ➔ DRD — Decision-Ready, Decision-Relevant data. ➔ DTM — Digital Terrain Model (used in the context of "Global ensemble DTM"). ➔ ECMWF — European Centre for Medium-Range Weather Forecasts (an entrusted entity for the Copernicus Programme). ➔ EEA — European Environment Agency (mentioned in the context of a report). ➔ EO — Earth Observation technology, includes: space-based infrastructure (sensors), data (images, scans, scenes) and ground infrastructure for storing, processing and sharing output data. ➔ ESA — European Space Agency. ➔ ESDS — Earth Science Data Systems (NASA's Program). ➔ ETH — Swiss Federal Institute of Technology in Zurich (lead for World-reforestation monitor). ➔ EUDR — EU Deforestation Regulation. ➔ EU-INC — EU Innovation Capacity (used in the context of supporting young businesses, i.e., "28th Regime / EU-INC"). ➔ EUMETSAT — European Organisation for the Exploitation of Meteorological Satellites (an entrusted entity for the Copernicus Programme). ➔ FAIR — Findable, Accessible, Interoperable and Reusable. ➔ Fediverse — a decentralized cyberinfrastructure where new members are continuously added and have relative autonomy while following some minimum community rules and regulations. ➔ FRD — Forensic-Ready-Data: for example a map of illegal deforestation areas, flood damage areas and similar. ➔ FVC — Fractional Vegetation Cover; inverse of the Bare Surface Fraction. ➔ GeoAI — Geospatial Artificial Intelligence. ➔ GFZ — Helmholtz Centre Potsdam - German Research Centre for Geosciences (an author/partner affiliation). ➔ GHG — Greenhouse Gasses (carbon dioxide, methane, nitrous oxide, and water vapor). ➔ IoT — Internet of Things. ➔ JRC — Joint Research Centre (referred to as EC JRC). ➔ LAI — Leaf Area Index. ➔ ML — Machine Learning. ➔ NASA — National Aeronautics and Space Administration. ➔ OEMC — Open-Earth-Monitor Cyberinfrastructure project. ➔ OGH — OpenGeoHub (an author/partner affiliation). ➔ SME — Small and Medium Enterprises. ➔ TIB — Technical Information Library (channel hosting video-recordings). ➔ VLMs — Vision-Language Models. ➔ WMO — World Meteorological Organisation. ➔ WUR — Wageningen University & Research (an author/partner affiliation). ➔ XDGGS — Multi-array binary format (used in the context of data formats for massive usage). Page 2 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org Open-Earth-Monitor Cyberinfrastructure (OEMC) project Consortium and Funding: ● Lead: OpenGeoHub Foundation (Doorwerth, Netherlands). ● Partners: A consortium of 23 organizations, including research institutions, universities, and companies across Europe (e.g., IIASA, Eurac Research, Vizzuality, CMCC, GFZ, Sinergise, and others). ● Duration: 2022–2026 (4 years). ● Approach: Implemented in three phases: 1. Development of computing engines and data services. 2. Application to real-world use cases supporting EU Green Deal actions. 3. Dissemination via workshops, stakeholder engagement, and tool refinement. Open-Earth-Monitor Cyberinfrastructure project (OEMC) was funded under the HORIZON EUROPE (HORIZON-CL6-2021-GOVERNANCE-01-162) call and is focused on speeding up uptake of EO data for monitoring purposes, demonstrated through almost 30 use-cases (duly registered via the GeoKnowledge Hub3), and helping users of EO data make their workflows more FAIR (Findable, Accessible, Interoperable and Reusable), specifically to support the objectives of the European Green Deal, the EU Climate Adaptation Strategy, the European Data Strategy and the UN Sustainable Development Goals. The OEMC project has already produced over 60 research publications4, Terabytes of open geospatial data, number of software components and over 140 video-recordings of talks, training courses and similar. OEMC is closely contributing to GEO’s initiative (Intergovernmental Group on Earth Observations) through the EuroGEO5 initiative and various national initiatives. The OEMC project is largely inspired by similar past projects such as the Earth System Data Laboratory6, EarthCube7 and is building upon the openEO.org software and EcoDataCube.eu, OpenLandMap.org initiatives. In a nutshell, the OEMC project looks at open source Fediverse-type infrastructures such as Mastodon and asks: “how to organize open development communities into efficient decentralized open collaborative systems (fediverse) where computational notebooks are used seamlessly to overlay, process EO data and visualize data and patterns?” OEMC is based on the following six (6) generic research & innovation activities: 1) Semi-automated to fully automated mapping pipelines (based on cutting-edge ML and cloud-solutions for data access) so that production of environmental information can be accelerated and costs reduced, 2) Hybrid process-based ML-based modeling and knowledge-guided ML integrating spatiotemporal process-based modeling with training and calibration, trying to mimic major Earth System processes (and thereby support the construction of Digital Twin Earth systems), 3) Optimization of computing by advancing our programming to incorporate also data mining and model testing at multiple spatial and temporal scales, 4) Innovating how the monitoring networks are organised (in-situ) and how IoT (Internet of Things) technology is used to stream data directly into the automated analysis, 7 https://www.earthcube.org/info and the corresponding final report https://doi.org/10.6075/J04X5806. 6 https://earthsystemdatalab.net/ 5 https://www.eurogeosec.eu/ 4 https://www.zotero.org/groups/5705036/open-earth-monitor_cyberinfrastructure_project 3 https://gkhub.earthobservations.org/search?q=OEMC&l=list&p=1&s=10&sort=bestmatch 2 https://doi.org/10.3030/101059548 Page 3 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org 5) Innovating how scientific data is visualised and disseminated (including on mobile phones) to help users quickly understand risks of climate change and find environmental and climate solutions, 6) Forecasting the short-term and long-term future scenarios for land use / land cover by implementing re-analysis methods and scenario testing on existing forecasts (e.g. climate, energy-supply scenarios). Furthermore, we base our R&I on the following three key premises: 1) to increase usage of environmental information it should be distributed as what users consider as decision-ready data or (at least) analysis-ready data, 2) climate action or any similar large-scale environmental management/restoration will struggle until most citizens are aware of the financial benefits and co-benefits of ecosystem services, 3) users, i.e. people, should be central to co-designing a monitoring system and ought to be involved from the start of the implementation phase. The key OEMC project outputs in 2025 f The OEMC project is delivering hundreds of terabytes of new value-added data, software components, and workflows demonstrated using over 30 use cases with target stakeholders. Specific focus on the project was on contributing to the basic layers on CDSE and building global and pan-EU reference data sets to support monitoring of forests, soil, agriculture, biodiversity, water and climate resources. As mentioned above, the OEMC project has already produced numerous publications, novel data sets and software components. Instead of listing all of them, we provide a brief list of the key outputs the project has produced so far (note: OEMC is an official open data project meaning that all outputs are available under the CC-BY or similar open data license by default): ● Landsat archive 1997–2024 bimonthly mosaics8 ARCO added to CDSE (about 150TB of data); thanks to a dedicated support from ESA, DG-DEFIS and Cloudferro (lead OGH); ● Global ensemble DTM9 (30 m) and Global Ensemble Land Cover (17 classes; pending) also to be added to OEMC catalog & CDSE (lead OGH); ● Pantropical commodity product (10 m) upcoming (lead WUR); ● Africa drivers of deforestation10 (publication & data) (lead WUR); ● EU Deforestation alerts paper & product (10 m) upcoming (lead WUR; operational dataset to be released in Q1/2026); ● 50 million field boundaries11 for pan-EU + the crop classification provided as open data (lead Sinergise); ● Soil moisture products12 for pan-EU / collaboration between OEMC, EUMETSAT, JRC and Italian Civil Protection for drought monitoring (lead CNR); ● AGB reference data13 for global modeling submitted (lead GFZ); ● FLUXNET datasets14 updated release on global in-situ ecosystem GHG flux exchanges (datasets, new access tool and paper to be released in Dec 2025) (lead CMCC); 14 https://stac.earthmonitor.org/insitu_fluxnet/collection.json?.language=en 13 https://doi.org/10.5281/zenodo.15495068 12 https://doi.org/10.5194/hess-29-3865-2025 11 https://doi.org/10.5281/zenodo.14229032 10 https://doi.org/10.1038/s41598-024-52138-9 9 https://github.com/openlandmap/GEDTM30 8 https://browser.stac.dataspace.copernicus.eu/collections/opengeohub-landsat-bimonthly-mosaic-v1.0.1?.language=en Page 4 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org ● World-reforestation monitor: Global annual 20 m maps for LAIe / FAPAR / FCOVER for reforestation monitoring: Data, Preprint, GEE-app (lead ETH); ● About 140 videos available on 3 international Global Workshops / summer schools on EO data science (produced jointly); ● OEMC use-cases added to GeoKnowledge Hub15; These outputs are used by various research organizations, but also commercial companies for developing additional services and products. Total project outputs can be best followed via the official output channels and our STAC data catalog16. Fig. 1: Video-recordings of oral talks, keynotes, training sessions, workshops and interviews are hosted on TIB channel (videos registered via open license and with a DOI and hosted on European servers). Backgrounds: the rationale for open EO data and applications Opening EO data is a low cost investment considering the benefits to society, human health and saving lives. European citizens pay about 0.06% of the GDP to run the European Space Agency17, and similar small shares to pan-EU organisations for the costs of EO-based monitoring systems of natural resources and natural hazards such as storms, wildfires, floods, and droughts. The financial analysis shows that this is a modest cost in exchange for ensuring the safety of European citizens. Recent events such as the Valencia floods in Spain, specifically the DANA storm18 of October 29, 2024, that claimed the lives of over 200 people 18 https://en.wikipedia.org/wiki/2024_Spanish_floods 17 https://space-economy.esa.int/ see: “Creating Value for Europe” 16 https://stac.earthmonitor.org/ 15 https://gkhub.earthobservations.org/search?q=OEMC Page 5 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org and caused catastrophic damage estimated to be over €50 billion19 underscore the importance of such investments. This includes insured losses of around $4.2 billion and state-based catastrophe insurance claims of €3.5 billion, according to Commercial Risk. Similarly, the 2021 floods in Germany and Belgium caused total damage estimated at €46 billion (not to mention the 239 reported human casualties)20. This includes both insured and uninsured losses, as well as damage to infrastructure, businesses, and homes. The floods were among the most expensive extreme weather events in Europe since 1980. The annual cost of floods and droughts in Europe due to climate change is estimated to be in the billions of euros, with recent studies suggesting averages of around €15.5 billion to €26 billion per year. These do not include losses of human lives (1980–2021: 4,161 flood fatalities, 159,003 from heatwaves, and 15,169 from droughts, wildfires, or coldwaves) as well as permanent losses of soil and soil productivity. These costs are projected to increase significantly in the future due to more frequent and intense extreme weather events. EEA report “Economic losses and fatalities from weatherand climate-related extremes”21 (floods, heatwaves, droughts, forest fires and coldwaves, storms and landslides) found that economic losses from weatherand climate-related extremes amounted to just over €45 billion in 2023 for 38 European countries, including EU and other EEA members and cooperating countries. Total economic losses from weatherand climate-related events exceeded €790 billion across the EEA-38 member and collaborating countries (32 EEA members plus the six Western Balkan countries) between 1980 and 2023. While in general unhealthy/degraded soils cost over 50 billion annually22. On the other hand, EO can indeed be used to save lives. Some examples include using Sentinel-1/2 for alpine avalanche warnings, detecting snowpack instability, preventing fatal avalanche accidents every winter (Austria, Switzerland, France), for coastal flood evacuations to detect rapidly rising coastal waters (e.g. during the storm Xynthia; France, 2010), for detecting potential landslides (Aosta Valley, Dolomites) and volcanic early warnings (Sentinel-1 InSAR is used daily by INGV to detect ground uplift, fault movement and lava flow directions for Mount Etna, Campi Flegrei), for wildfire evacuations (Greece, Portugal, Spain) where a fusion of Sentinel-2, MODIS, VIIRS23 and Meteosat has been central to detecting new fire outbreaks within minutes, predicting fire spread and similar. Improvements in hurricane track forecasts have reduced the economic impact of hurricanes by about $2 billion per hurricane, demonstrating that observations pay for themselves24. In summary, the costs of developing and maintaining Earth Observation (EO) missions such as Copernicus Sentinel satellites are minor compared to the potential losses in goods and human lives due to poor planning and poor forecasting. The economic benefits of Copernicus are forecasted to generate 10 to 20 times the programme's total investment over the long term25. between 2017 and 2035, Copernicus is expected to generate €67 billion to €131 billion in benefits for European society. Companies such as FARMSTAR (France), VITO / WatchITgrow (Belgium), SatAgro (Poland), Aurea Imaging & NEO (Netherlands), eAgronom (Estonia), OneSoil (Belarus/Global focus), Agreena (Denmark) use Sentinel data for crop monitoring, yield prediction and subsidy monitoring; PINA EARTH uses Sentinel data for forest management i.e. to quantify and certify forest carbon capture; Sentinel Hub (Sinergise) and Copernicus Data Space Ecosystem (CDSE / Cloudferro) lead the data aggregation & cloud processing services; Sentinel images are also used for insurance claims verification, oil spill & ice detection, ship tracking & security i.e. for tracking illegal fishing. The European Commission now, for example, invests significantly in soil conservation with €1 billion alone 25 https://www.copernicus.eu/en/what-are-tangible-benefits-copernicus 24 https://player.vimeo.com/video/1140103431 23 https://www.earthdata.nasa.gov/data/instruments/viirs 22 https://doi.org/10.1016/j.landusepol.2025.107755 21 https://www.eea.europa.eu/ims/economic-losses-from-climate-related 20 https://www.preventionweb.net/quick/95596 19 https://www.theguardian.com/world/2025/sep/15/europes-summer-of-extreme-weather-caused-43bn-of-short-term-losses-analysis-finds Page 6 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org allocated to the Soil Mission (part of Horizon Europe) between 2021 and 2028. This investment supports research and on-the-ground actions to reverse soil degradation and puts Europe at the forefront of soil conservation activities in the world. In addition, the EU’s Common Agricultural Policy (CAP) supports sustainable soil management on farmland, while the recent Nature Restoration Law promotes improvements in the soil-water-nexus through ecosystem restoration in natural and agricultural areas. All these programmes and projects need Analysis-Ready (ARD), Decision-Ready, Decision-Relevant data (DRD) and/or Forensic-Ready-Data (FRD i.e. data that can be used in court) on the state of environment and efficient models capable of mapping risks and predicting the future with highest possible accuracy. Data producers such as ESA often lack the capacity and expertise to produce ARD, DRD and FRD data and do not see this as their core missions. This is where the HORIZON Europe and the EuroGEO initiative play a key role. How can open EO data and CDSE speed-up Green transition? “One of the things that I think is really going to affect our global society is — and one of the things that I think is going to propel open source and transparency and data sharing is — I believe that true open collaboration, which requires all those things, is going to be a tremendous economic engine. And I think that those organisations [...] that take an exclusively proprietary mindset are going to see themselves being left behind.” Ron Garan26 Open Earth Observation (EO) systems & programmes such as NASA’s/USGS’s Landsat, MODIS, ICESat, Earth Science Data Systems (ESDS), and EU Space Programme’s Copernicus Sentinel constellation have revolutionized EO science. Copernicus Sentinel is now the foundation of monitoring implementation of key EU agricultural and environmental policies, including Carbon Farming Initiative, Common Agriculture Policy, Soil Monitoring Law, and the EU Deforestation Regulation (EUDR)27. The open data license used for Copernicus Sentinel allows researchers, entrepreneurs and government agencies to develop procedures, products and services that can then be used to serve target customers, projects and users/stakeholders. Most importantly, an open access policy enables data optimization for massive usage through cloud-based coding infrastructures and web-mapping apps in a non-centralized way. Thanks to an “open science approach” and forward-thinking vision, the European Union and ESA have become the “world’s lighthouse of open EO”. Today, European (and global) citizens can track the state of the environment, including every farm, forest patch, and even individual trees across borders with weekly revisit times and at high spatial resolution of 10 meters (it is today difficult to “hide below the clouds”). From the data science perspective, open EO data has also resulted in a revolution in research and applications, especially those built on using open source programming data science languages such as Python, R, Julia, Javascript. f Considering the benefits of open EO data and considering the total costs of ESA missions such as Copernicus Sentinel (especially in terms of costs per capita per year) there is still room to increase investments into EO missions; there is still, however, also a large imbalance between investments into the ESA’s space segment and investments into exploitation and uptake of open EO data, this eventually making EO data largely under-utilized. The most important users — land owners, farmers, municipalities, forest managers, nature conservation agencies — on the other hand, are not even aware of the existence of open EO data, where to find it, how to use it and how to move from EO data to decisions (and they should not be blamed for this situation). 27 https://www.euspa.europa.eu/newsroom-events/news/eu-space-key-enabler-eu-deforestation-free-products-regulation 26 https://bigthink.com/videos/the-importance-of-international-data-sharing/ Page 7 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org Deep Learning (DL) systems, particularly Convolutional Neural Networks (CNNs) like U-Net54, are the backbone of modern LCC. Today various companies and groups show how 10–20 m resolution Sentinel images can be sharpened up to 1 m resolution with impressive accuracy. ● AI for citizen science: AI can significantly help filter citizen science data, then make this data become the dominant source of data, especially for training models and helping improve accuracy of decisions locally. Moreover, AI tutors could help citizens, become experts in biology, ecology and similar, without expensive scholarships. Another novel concept is the one of “Earth Intelligence”. In a nutshell, Earth Intelligence is a natural fusion of EO and AI technologies aiming at generating actionable insights about the Earth's physical, biological, and human systems. The earth is an interconnected system (atmosphere, ocean, terrestrial / land), but institutions and projects remain siloed. In the scheme below we extrapolate the concept of Earth Intelligence and combine it with Digital Earth simulations: a system of human-centric and biological systems combined with a game theory and optimization. Fig 2: Hypothetical hybrid global modeling system testing various scenarios and helping decision makers come to an international agreement. Imagine Paris-climate-type agreements that were moderated by AI. 54 https://doi.org/10.3390/rs13183600 Page 14 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org The government in Albania is already testing using AI for processing citizen applications and managing funds55, so why not also imagine a world where AI (trained jointly by local, regional, national and international government bodies) is asked to recommend the optimal path towards progress? It is difficult to predict how Earth Intelligence will evolve and whether it will ever serve all citizens of the planet (without exploiting or discriminating against any group), but it is certainly going to become increasingly possible and it could just help save us from making drastic planning mistakes. United Nations Environment Programme’s (UNEP) Global Environment Outlook 7 report ‘A future we choose’56, estimates the global macroeconomic benefits of transition to circular economy and significant change and food and energy cultures (sustainable diets, reducing waste, improving agricultural practices and expanding protected areas while restoring degraded ecosystems) will start to appear in 2050, and grow to $20 trillion (approximately €17.19 trillion) per year by 2070. The report also suggests strongly to move beyond GDP as a measure of economic well-being: the intersection between optimal human health, wellbeing, development index and ecosystem health & resilience and sustainability of resources on scales of 200+ years, is maybe narrow in the plot shown above, however we should not stop believing that it exists. f AI and especially Artificial General Intelligence could be soon used to help human societies find an optimal path to the future (without leaving anyone behind). Such a self-regulated, multicriteria optimized AI+governance system could be called “Earth Intelligence”. Next milestones for the OEMC project in 2026 The OEMC project is due to finish in December 2026. The remaining 12 months will be used to finish multiple monitors and release all data and software on public repositories and a central app (developed by Vizzuality and GILAB). Remaining milestones for the project include: - Produce V2 of the Landsat data set i.e. a monthly product in 16bit format; potentially produce annual embeddings for 2000–2024+ by combining Landsat, MODIS (250 m) and land surface parameters from Digital Terrain Modeling57; - Release all data in STAC data catalog and available for viewing using Terria.io digital Earth; - Complete a central Tier-1 web-GIS app where all users can access data and read about the project and find all key data sets and publications; - Provide guidelines (the summary report) for public release which should also include recommended standards and specifications for Open-Earth-Monitor type infrastructures, - Release all final output during the OEMC global workshop in Barcelona 7–9 October 2026; - Connect with the OEMC stakeholders committee and collect the final feedback on the key project outputs; - Register all project outputs with DOI and provide as online libraries via Zotero.org, Zenodo.org, Github.com / Codeberg.org and similar; These outputs would then be gradually turned into more operational products, several will be potentially used as commercial products; many outputs are already used for sister projects especially those focused on geospatial data production e.g. Global Pasture Watch, https://app.Landmetric.org and similar. The commercial partners on the project Sinergise, 57 https://github.com/openlandmap/GEDTM30 56 https://www.unep.org/resources/global-environment-outlook-7 55 https://www.forbes.com/sites/kolawolesamueladebayo/2025/10/23/albanias-ai-minister-is-a-real-world-test-for-algorithmic-governance/ Page 15 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org GILAB, Brockman Consult, Vizzuality and TerraSigna are already developing a number of solutions that could be brought to TRL >7. Summary conclusions “Rigorous and transparent carbon accounting is fundamental to progress. The partnership between the Greenhouse Gas Protocol (GHGP) and ISO is the foundation of the global carbon-accounting system.” Ani Dasgupta, President & CEO of the World Resources Institute (WRI)58 The OEMC project advocates for Open Earth Observation (EO) data and fediverse-type systems as the foundation for permanent environmental monitoring and the green transition in Europe. We highlight the pivotal role of Copernicus Sentinels in data-driven governance and potentially new environmental accounting systems. We especially focused on speeding up the uptake of FAIR EO data, producing numerous research publications, open geospatial data, and software components, and emphasize the importance of the Copernicus Data Space Ecosystem (CDSE) as a fully operational platform. After three years of OEMC project and four (4) public events and two (2) summer schools, and numerous discussion panels, we are providing some general summary recommendations and conclusions aiming at policymakers: Open Earth Observation is a foundation of data-driven governance; EO-based monitoring might soon become the basis of environmental accounting: Thanks to the Copernicus Sentinels, today the whole planet is scanned at a spatial resolution of 10 m and the revisit time on each point every 5–6 days (ESRI Wayback provides public access to 30 cm VHR images for 2014 to present). Satellite images are a foundation of data-driven governance. In 2025 we still only pay taxes on property, cars and similar. This system however is becoming outdated in not recognising the loss, due to anthropogenic activities, of essential services ecosystems provide to humankind, and its related costs. A tax-like system will be needed to compensate and prevent a loss of ecosystem services and environmental impacts, along the lines of the carbon tax59. All parties of society should pay their share of the costs of climate crisis, GHG emissions, land degradation (soil erosion, pollution, compaction, loss of biodiversity, deforestation etc) and these need to be accounted transparently and correctly i.e. taking an objective estimate of the proportion of their share of responsibilities. We anticipate that Copernicus Sentinels (i.e. open EO) might soon become the basis of a new accounting system across the European Union. Businesses should not miss out on the opportunity to provide cost-effective and robust services to support this accounting. A lot of testing and development is, however, still needed until these new accounting systems become fully operational. Copernicus Data Space Ecosystem60 (CDSE) is a fully-fledged infrastructure and a fully-operational platform — everyone is invited: CDSE funded and established by European Union, is now a fully operational platform for solving critical problems of data access, processing, and availability, hosting nearly 100 petabytes of data with tools that support AI and machine learning workflows. The platform is open to researchers, policymakers, and businesses, providing seamless integration and scalable compute resources. In addition to Sentinel 1 and 2 data, the OEMC project has recently also added a bimonthly Landsat at 30 m data set which covers 1997–2024 to CDSE so that users can also go beyond year 2016 in their analysis. A number of tutorials explaining how to access and use CDSE to a full capacity have been made available (see e.g. Pratichhya Sharma: Introduction to openEO within the Copernicus Data Space Ecosystem and/or Valentina Premier: Streamlining Snow Monitoring with openEO and CDSE). 60 https://dataspace.copernicus.eu/ 59 https://taxfoundation.org/data/all/eu/carbon-taxes-europe/ 58 https://www.tfs-initiative.com/news/catalysing-change-in-carbon-accounting Page 16 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org Key to success of many EU funded programmes is in the well-planned development cycles (see below): This applies to both policy adoption and implementation and design of technology. In addition, we believe that users should be central to designing a cyber-infrastructure and we have insisted on involving them in the Open-Earth-Monitor design from the start of the project. This is why we have insisted on opening the OEMC Global workshops with programmes focused on practical tutorials, hackathons and similar, where user communities can directly reflect, provide feedback and eventually use, re-use and extend produced components. Fig. 3: Uptake of the EO technology through policy cycle (above); recommended efficient evolution of technology from service evolution to ground and space segment evolution (below); source “EO for Policymaking in the EU” by Marc Dowell (EC JRC). The fast emerging AI technology, in the coming years, will likely boost the use of EO data by an order of magnitude, but there are still a lot of challenges in front of us: The new sub-field Page 17 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org of geospatial science GeoAI is fast emerging. The recent AlphaEarth61 (Google Earth embeddings) are the first complete consistent and ready-to-use open (distributed under CC-BY license) EO global coverage 10 m resolution annual products (2017–2024) currently available. Embeddings have a potential to revolutionize use of EO: (A) they significantly compresses the data volumes (up to few hundred times), (B) data is complete and consistent (every year produced using the same modeling) and available on a cloud-optimized infrastructure allowing for easy development of apps and services, (C) embeddings allow overcoming current satellite images limitations such as presence of clouds, data gaps and artifacts due to many independent scenes or scanning angles. Even though the data produced by Google Deepmind is open, the code and models seem to be closed source, hence any critical issues are invisible to the community. Fig. 4: A small number of US companies (Microsoft Azure, Amazon Web Services, Meta, OpenAI and Google) dominate 70–80% of the European market for cloud services. Globally, this is even more distinct. Image source: Financial Times. Europe is a strong partner with the USA, however the EU’s cloud-dependence on the USA weakens the sovereignty of many private and business entities: A small number of US companies (Microsoft Azure, Amazon Web Services, Meta, OpenAI and Google) dominate 70–80% of the European market for cloud services62. The silicon-valley tech-giants offer cutting-edge technology often at no cost (free), which is a well-established business model, but also largely an illusion that these services are “for free”. Customers are attracted to use these services as they are easy to set-up and require no initial transaction, but the total bill to pay is often much larger than if they would pay a local start-up the actual costs of service. It is really a problem of low digital literacy of users where most of the customers do not read any terms, do not think or understand how their data is used, and how do these tech giants actually make enormous profits. Most youths in Europe have most (>80%) of their data in the USA cloud (but have no right of vote in USA)63. True: this is because many of these services are high quality and difficult to match with European systems, but there are already plenty of 63 https://www.euronews.com/next/2025/08/05/most-european-companies-rely-on-us-tech-giants-to-operate-their-businesses-study-warns 62 https://www.clingendael.org/publication/netherlands-and-eu-prioritising-cloud-sovereignty 61 https://deepmind.google/discover/blog/alphaearth-foundations-helps-map-our-planet-in-unprecedented-detail/ Page 18 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org European solutions that are of comparable quality and even not-for-profit e.g. Codeberg.org vs Github.com (Microsoft), but still either relatively unknown or significantly under-utilized. Similarly, CDSE, EarthCode64, Copernicus Land Monitoring Services (CLMS)65 and openEO.cloud are now a fully operational cloud-service for accessing and processing EO data, and have comparable functionality as Google Earth Engine (currently world’s #1 infrastructure for processing EO data). Horizon Europe funded projects should all put an effort to prioritize using, extending and improving this infrastructure and insisting on at least 50% of cloud sovereignty within each organization. Self-hosting data using open source solutions such as Nextcloud66, Ceph67 and similar are viable options. The OEMC project has also advanced openEO software68 (openEO can now also run AI-workflows) and organized a series of tutorial and training sessions on how to access and use CDSE (also through OpenEO), so plenty of support for everyone. The European Union needs to embrace innovation and competitiveness and prioritize digital sovereignty and do everything possible to keep young, talented and creative people busy and engaged (and living in Europe). f A small number of US corporations (Microsoft Azure, Amazon Web Services, Meta, OpenAI and Google) dominate 70–80% of the European market for cloud services. To ensure at least minimum sovereignty, European organizations could insist on at least 50% of cloud sovereignty. Self-hosting data using open source solutions such as Nextcloud, Ceph, Geonode, TerriaJS and similar are viable options. Science and technology are basis for evidence-based decisions, however, efficient governance/NGO’s are basis for implementation and impact of these decisions: Science and technology alone can not solve problems such as the climate crisis, the loss of biodiversity and land degradation without strong governance (legislative, executive, and judicial) and civil society organizations (NGO’s) — a harmonic implementation of laws that are based on state-of-the-art knowledge and technology, operational laws and civil mechanisms that prevent corruption and misuse of information is needed to get a robust green transition moving. An example is the Soil Monitoring Law69 that has been recently approved by the European Parliament, the Directive 2024/1203 on the protection of the environment through criminal law70, also the EU Deforestation Regulation (although postponed to 2026) and similar. On the other hand, the Forest Monitoring Law has been recently rejected by the EU Parliament71 and this is somewhat confusing. Despite the technical feasibility and scientific capability of Copernicus satellites to monitor forests at high resolution and with transparency, political resistance — primarily from member states — has led to significant watering down of the legislation. This resistance reflects broader political reluctance to embrace open environmental data, jeopardizing progress on critical climate goals, not to mention active pushback against evidence-based climate policies. We underscore the urgency of improving forest carbon sink monitoring, given recent data showing a decline in Europe’s forest carbon sequestration72, which undermines climate targets. There is today a high interest in establishing robust monitoring mechanisms and infrastructures to track every pixel, every tree, every water stream, every soil (pedon) both at national, continental and global scales: Robust monitoring systems imply four basic principles: (I) there needs to be a baseline (i.e. the reference, year 0); (II) changes need to be registered using immutable, decentralized infrastructures to avoid any misuse or 72 https://www.peer.eu/news/detail/the-european-forest-carbon-sink-is-declining-can-we-reverse-the-trend 71 https://www.euractiv.com/news/european-parliament-rejects-law-to-monitor-forests/ 70 https://www.nortonrosefulbright.com/nl-nl/knowledge/publications/1ad9c021/the-new-eu-directive-2024 69 https://www.consilium.europa.eu/en/press/press-releases/2025/09/29/council-adopts-new-rules-for-healthier-and-more-resilient-european-soils/ 68 https://open-eo.github.io/openeo-python-client/ 67 https://ceph.io/en/ 66 https://nextcloud.com/ 65 https://land.copernicus.eu/en 64 https://earthcode.esa.int/ Page 19 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org manipulation (e.g. using block-chain solutions); (III) uncertainty of estimates / metrics needs to be assigned and communicated clearly to users73; (IV) tax-payers paying for the infrastructure and production of data should have free, fast and simple access to data (including from their mobile phones) and should be able to make fast decisions. By meeting these simple principles, one can create a robust system where basically every tree (see illustration below), every water stream and every pedon is tracked across the whole planet. On a positive side: open source offers several solutions for all of the four key principles, so implementation of robust monitoring can be also done cost-effectively, raising on the wings of open development communities. Fig. 5: “The Eyes of Earth” ESRI’s Wayback provides public access to Very High Resolution (VHR) images at cca 30 cm resolution and covers 2014 to 2024+. Today it is increasingly difficult to “hide from the public eye” any land use change, including cutting out a single tree. ESRI Wayback data is proprietary, however, ESRI allows users to digitize content and share digitized content as open data. The key role of in-situ measurements in validating satellite data should be recognised and brought to the attention of the public: Without “ground truth” we could not have usable EO datasets. Also, in-situ measurements are powerful tools per se, useful to characterise an area, an ecosystem or a biome under so many different perspectives. The challenge with in-situ measurements is in their high heterogeneity, in terms of data format (csv or other ASCII files, netCDF, geo-parquet, tiles, ...), type (point data, timeseries, vectorial data), origin (GHG flux data, biodiversity, citizen science, forest biomass, ocean and water systems, ...) not to mention metadata availability and readiness in terms of findability, accessibility, interoperability and reusability (FAIRness). In OEMC, we valued the importance of in-situ datasets from the beginning by dedicating a whole work package to their preparation — scouting, collection, FAIRification, adaptation to the monitors' needs. We then develop metadata catalogues based on STAC standards where the in-situ datasets used in the project are described, linked and associated to the corresponding monitors: the definition of the keywords, the requirements in terms of open access, and the classification based on their degree of readiness for the use in 73 https://doi.org/10.1038/s41467-022-32693-3 Page 20 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org the project have been key aspects to showcase both the reproducibility and the potential of the in-situ datasets. Citizen-science initiatives and data need to play a more prominent role in environmental monitoring: There is a lot of opportunity for including citizens and their networks into monitoring of environmental resources. In fact, for a society to be successful and democratic, it is crucial that NGO’s, voluntary organizations and individual citizens get fully involved and become a dominant source of up-to-date high quality information on the status of the environment. Systems that are fully reliable on government agencies and long-term funding are often inefficient and also there is no guarantee that they will continue after the programme finishes or political leadership changes. Self-organized communities of volunteers, enthusiasts and not-for-profit organisations should be leading innovation and fostering tangible real-world impact. This is not a trivial goal as citizen science data has many challenges such as data quality, engagement, motivation, digital divides, and ethical considerations around payment for data collection. AI could potentially play a significant role in the coming decade to help enhance citizen science data quality through image recognition and validation. The future is in the hands of the youth. Gamification and token systems are some strategies to get younger generations to sustain citizen engagement in data collection projects. Beyond modeling and monitoring biophysical variables representing vegetation, water, land use dynamics, it is becoming increasingly important to combine such approaches with socio-economic modeling: Within the OEMC project we have been testing developing models where key socio-economic variables are used as a target variable and EO data, climatic data as predictors. Colleagues from the Max Planck Institute in Jena74, seek to create synthetic indicators akin to medical diagnostics for a planet’s health, capturing systemic changes and risks at the intersection of environmental and socio-economic factors (key stakeholder: the European Central Bank, interested in assessing systemic financial risks related to climate change and biodiversity). The methodology now involves a three-way Canonical Correlation Analysis (CCA) framework, exploring bidirectional explanatory relationships among spheres (e.g., atmosphere explaining biosphere, and vice versa). Early results show moderate correlations and reveal complex but interpretable linkages between climate variables, biospheric indicators (like vegetation indices), and socio-economic variables (such as food and fuel prices). This area of research has a large potential and could help broader society better understand how the climate crisis is affecting their own health and wellbeing. Land owners and land managers that help decrease land degradation i.e. help restore ecosystems and its functions should be rewarded either through national schemes, tax reduction schemes and/or subsidies: There is clear consensus among the project participants and participants of our workshops that champions of land & ecosystem restoration should be celebrated and rewarded for their work75. The best way to speed up restoration of ecosystems is to motivate all land owners / land managers that there are tangible benefits of doing it fast and doing it right. The Restor.eco infrastructure (a partner/stakeholder of the OEMC consortium) demonstrates that indeed land owners are excited about being validated and listed on a public portal. We need such crowd-sourced, easy-to-use systems to reach millions of land owners and especially those that are unaware of the benefits of land restoration. The three main technical challenges that require urgent improvements for the full exploitation of Earth Observation (EO) data are: [1] making EO data more Analysis-Ready (ARD), Decision-Ready, Decision-Relevant (DRD) and/or Forensic-Ready-Data (FRD), [2].testing and implementing new more efficient data formats (e.g., Cloud-Optimized formats, 75 https://investinginregenerativeagriculture.com/2025/03/04/tom-hengl/ 74 https://www.bgc-jena.mpg.de/en/bgi/efeo Page 21 of 23 Open-Earth-Monitor Cyberinfrastructure (OEMC) | https://EarthMonitor.org multi-array binary formats like XDGGS), [3] helping European researchers develop and extend their own GeoAI solutions / transforming core Copernicus service outputs into specific policy-ready indicators. Modern research publications should come with fully-documented computational notebooks (including DockerHub scripts) and metadata explaining all inputs, processing steps and outputs promoting reproducible research as the basis of research quality. Platforms such as CDSE, EarthCode, Codeberg.org, could play the key role in reaching this. f Formal national and global carbon-accounting systems are inevitable for the success of green transition and for combating climate change. Open EO should be a foundation of transparent and robust carbon-accounting. All parties of society should pay their share of the costs of climate crisis, GHG emissions, land degradation, and these need to be accounted transparently and correctly by taking an objective estimate of the proportion of their share of responsibilities. We believe that ESA and the European Union could significantly increase investments in making EO data more usable, to boost exploitation and support growth of start-ups and Small & Medium Enterprises (SME’s) that can bring this data to diversity of users, most importantly to farmers, municipalities, forest managers and similar. ESA has invested over the last decade about €3.2 billion for period 2014–2021 and €5.4 billion for period 2021–2027 for the Copernicus programme76 (about €3 per EU citizen per year; total costs of ESA is about €14 per citizen per year) so why not spend an additional 10–15% of that amount on supporting open development communities that can boost usability of EO data? Note: NASA costs its citizens about 4–5 times more per capita than ESA costs its citizens77, reflecting the greater national emphasis and absolute funding on space exploration in the United States, so there is definitely room to improve EU’s investments in the uptake of open EO. Emerging EO companies are rapidly prototyping solutions. Traditional development cycles are becoming too slow for today's dynamic market demands. Open source here plays a key role: keeping EO data open and using open source solutions as a basis of back-end is critical for several reasons: ● Foundation for EU Policy and Governance: Open EO systems like Copernicus Sentinels are the foundation for monitoring key EU environmental and agricultural policies (e.g., Soil Monitoring Law, EU Deforestation Regulation) and are anticipated to become the basis for a new environmental accounting system. It is crucial that these systems stay as transparent as possible and as decentralized as possible. ● Safety and Economic Protection: The cost of maintaining open EO missions is minor compared to the catastrophic economic losses (billions of euros) and human casualties caused by extreme weather events like floods and droughts, which the data helps to monitor and mitigate. Even more funds should be put into making sure that EO data penetrates all potential aspects of life, especially to reach land managers, farmers, forest managers and similar. ● Enabling Innovation and Efficiency: An open access policy allows for data optimization for massive usage through non-centralized, cloud-based infrastructures and fosters innovation among researchers and businesses. It helps avoid the inefficiency of multiple groups repeatedly paying to process the same raw data on proprietary commercial cloud services. The OEMC project is finishing with the final closing conference in Barcelona in October 2026. For more information refer to: https://earthmonitor.org/global-workshop-2026/. 77 https://www.esa.int/About_Us/Corporate_news/ESA_facts 76 https://www.esa.int/Applications/Observing_the_Earth/Copernicus/New_financial_resources_for_Copernicus_space_component Page 22 of 23 Open-Earth-Monitor Cyberinfrastructure project policy brief December 2025 MONITOR OPEN EARTH Contact earthmonitor.org @EarthMonitorOrg @openearthmonitor.bsky.social This project has received funding from the European Union's Horizon Europe research and innovation programme under grant agreement No. 101059548. Events In October 2026, the Open-Earth-Monitor project will hold its Final Conference in Barcelona, Spain. Discover the event: https://earthmonitor.org/global-workshop-2026/.