A serious obstacle of a total uptake of open Earth Observation data (Copernicus Sentinel images, NASA’s Landsat and similar) in daily lives is the steep data analysis curve to get from raw images to Analysis-Ready, Decision-Ready/Relevant, not to mention Forensics ready data. The combined complexity of high data volumes, atmospheric disturbances (clouds, haze), inconsistent coverage and diverse and complex signal physics (e.g. radar images vs optical images; sudden land-use changes due to deforestation or urbanisation) has kept the application of Earth Observation (EO) data relatively marginal. For example, in Europe, only a small fraction of farmers, spatial planners, civil engineers, and forest managers are estimated to use Sentinel images for decision-making. The recently generated Google DeepMind AlphaEarth (10 m global for 2017–2025) and Tessera embeddings being complete, consistent and ARD, provide an opportunity to decrease the steep data processing curve and enable thousands of applications. In our work, we have also consistently focused on making EO data more ARD and more usable, primarily by aggregating Landsat 1997–2025 16-day images (Potapov et al., 2020; https://doi.org/10.3390/rs12030426) into bi-monthly global mosaics (Consoli et al., 2024; https://doi.org/10.7717/peerj.18585). We now aim to significantly improve the usability and Analysis-Readiness of the global monthly, seasonal and annual mosaics (from the previous projects 2018–2022 and 2022–2026): by making them more complete (>99% pixels with values), consistent, artifact-free, fully cloud-optimized and compressed, and available via professional open geospatial solutions such as STAC/S3, generic API and Virtual Earth viewers. To achieve such a level of analysis-readiness, we plan to build a robust automated mosaicking tool. This tool uses cutting-edge aggregation, gap-filling and artifact removal algorithms, along with High-Performance Computing. Most of the code we run in C++, data IO operations run via the Infiniband infrastructure and most processing is implemented in RAM (18 servers with 1 to 3TB of RAM). This makes it possible to process over 1PB of data within weeks on our own infrastructure, producing next-generation open Earth Observation data for everyone and serving thousands of applications aimed at environmental monitoring and supporting the green transition.
Tomislav Hengl (Wed,) studied this question.