EarthServer-2 makes Agile Analytics on Big Earth Data Cubes of sensor, image, simulation, and statistics data a commodity for non-experts and experts alike through • navigation, extraction, aggregation, and recombining of any-size space/time data cubes; • easy to install & maintain value-adding services extending the existing portfolio of data and compute centers; • based on open standards, in particular: the OGC Big Data standards and the forthcoming ISO SQL/MDA (“Multi-Dimensional Arrays”) standard. In the Joint Research Activity, the project will advance the existing, world-leading rasdaman Array Database technology wrt. query functionality, inter-federation data processing with automatic data and query distribution, tape archive integration, and 3D/4D visualization based on NASA’s virtual globe technology. In the Services Activity, large data centers (ECMWF, PML, MEEO/ESA, GeoScience Australia, JacobsUni) will set up water, air, weather, and planetary services on 3D & 4D data cubes up to Petabyte-size with user-tailored clients for both visual and textual ad-hoc mix&match. In the Networking Activity, the project will advance open Big Data standards in OGC, RDA, and ISO (in particular: write ISO SQL/MDA). Further, all adequate channels will be used for strong dissemination & exploitation, specifically: writing a monograph explaining OGC Big Geo Data standards; scientific publications & active conference organization; Earth science data user workshops for each domain addressed; actively contributing technology & experience to GEO / GEOSS and further bodies; establish standardized Big Geo Data benchmark and run it against EarthServer-2 and further relevant systems. Altogether, EarthServer-2 will maintain and extend the lead in Big Earth Data services established in the highly successful EarthServer-1 project. Being already supported by ESA, rasdaman will form an enabling building block for COPERNICUS / Sentinel.

Foundation models represent a paradigm shift in AI, exhibiting remarkable capabilities across multiple tasks. Their true potential lies in generalizing across diverse domains and modalities, a largely untapped frontier. DVPS advances this frontier by focusing on multimodal foundation models (MMFM), aiming to harness their capabilities across various application domains. DVPS emphasizes three core benefits of MMFM: label efficiency, compute reusability, and engineering efficiency. However, achieving these benefits in multimodal settings presents challenges such as modality-specific architecture and cross-modal alignment. To overcome these, DVPS aims to develop generalizable methods that work across diverse modalities and domains, creating a unified framework for MMFM development and integrating new modalities into existing models. The project focuses on generating foundational knowledge, delivering tested methods, and creating algorithms to expand MMFM capabilities across domains like cardiology, geo-intelligence, and language communication. DVPS also includes two "surprise domains" to drive innovation by challenging initial assumptions. Key objectives include the development of AutoDVPS, a toolkit for automated MMFM design, and the creation of DVPSBench, a benchmarking suite for evaluating MMFM across tasks and domains. DVPS aims to foster a European ecosystem for MMFM research, promoting transparency, fairness, and ethical compliance in line with European values. Through collaboration and open-source contributions, DVPS seeks to standardize and advance MMFM as a scientifically rigorous discipline.

RELIANCE will extend the EOSC service offering with a set of industry-strong, innovative, interconnected services for the open, efficient, and cross-disciplinary management of the research lifecycle. In accordance with FAIR and Open Science principles, it will adopt a holistic approach to address different research activities, spanning: discovery of and access to research data, methods and materials; extraction of valuable information from scientific literature; structuring data-driven scientific investigations as semantically rich research objects; FAIR-ness self-assessment; dynamic collaboration, sharing and dissemination considering the whole lifecycle as a first-class citizen; and supporting reproducibility, validation, versioning and reuse of research results. Core to the service portfolio are Research Objects as the overarching information artefact to manage research, Data Cubes for efficient and scalable structured data access and discovery, and Text Mining for the extraction of information from scientific text as machine-readable metadata. RELIANCE will pilot the services in three different Earth Science communities, fostering the use of Copernicus data and demonstrating their efficacy in real-life vertical and multi-disciplinary scenarios, and will launch an Open Call to engage other communities. The high-TRL research-enabling services portfolio will be onboarded in ESOC following established procedures, ensuring the interplay with cross-cutting and added value services, and following a user-centric approach. RELIANCE will expose its services via APIs and libraries, enabling their use via different frontends and Jupyter notebooks. RELIANCE brings into EOSC more than 3,000 research objects from research communities like Earth Science, represented by three project partners, but also others like Astrophysics and Bioinformatics, which have embraced research objects concept and tools, and will become the best ambassadors to enlarge and engage new communities

Earth system models (ESM) show diverging estimates of ecosystem carbon (C) uptake which drives persistent uncertainty about future climate and limits the realization of policy mitigation goals. The CONCERTO project aims to strengthen the European research ecosystem by creating an innovative scientific collaborative framework that enhances our understanding, monitoring, and modelling of the terrestrial cycle (CC), and leads to reduced uncertainty and ESM convergence. This framework has three main elements: the exploitation of novel Earth Observation (EO) data, the innovation of process models fed by these data, and Data Assimilation (DA) and Machine Learning (ML) techniques. The consortium includes European experts and institutes active in these areas. CONCERTO will advance the representation of land cover, leaf area index (LAI) and management intensity through new maps of high resolution with layers relevant for the CC. The new management map will enhance the creation of a new biomass production map. Further, CONCERTO will prepare for the ingestion of FLEX data in land surface models (LSM) through DA, and exploit Sentinel-5P/TROPOMI HCHO indicative of biogenic volatile organic compound (VOC) emissions. As step towards moving away from prior (static) parameterizations, the novel P model will be used to ingest these new data sources. This will enhance understanding of ecosystems responses to climate extremes, fires, and vegetation recovery from them, and the parameterization of LSMs included in ESMs. Novel representations of the underrepresented processes of lateral carbon fluxes through outgassing of CO2 from rivers will be developed. Dedicated seasonal and climate experiments will demonstrate how improved representation of land surface processes benefit the accuracy and trustworthiness of global Earth System simulation. This will lead to better predictability of the influence of land management on the CC and underpin avenues towards carbon neutrality