Anne Puissant

The study of the Earth system involves a wide range of disciplines that are increasingly collaborating to address global challenges like climate change. In France, Data Terra , the national Research e-Infrastructure (e-IR) supports this effort by managing data for five key Earth system components: Ocean, Atmosphere, Solid Earth, Land Surface, and Biodiversity. As part of Earth observation work, numerous themes and studies of the components of the Earth system produce both data derived from sensors installed on satellite or aerial platforms and data from in-situ measurements. It is particularly true with the Land Surface Data and services Hub (Theia) where data produced and used by the scientist enables to characterise the earth surface conditions and analyse their spatio-temporal dynamics over a long time to better understand and quantify the impacts and the resilience of territories faced to an increase of extreme events such as flooding or heat waves. To ensure efficient data management in repositories, the FAIR principles have been established. They are based on four key concepts: Findability, Accessibility, Interoperability, and Reusability. Among these principles, Interoperability remains a crucial yet challenging aspect, essential for enabling seamless data integration, interpretation, and use by both humans and machines. Achieving FAIR interoperability requires the use of controlled vocabularies and structured representation models that align with these principles. Data produced and used by the Theia hub is made available in data repositories and referenced in catalogues. To improve the Interoperability of this data, controlled vocabularies have been produced by experts. Several observatories from the French Critical Zone Research Infrastructure (OZCAR-RI) that produce the in-situ data documenting the continental surfaces use their own vocabulary, which means that the same observed property may have different variable names (e.g. ‘soil moisture’, ‘soil water content’, etc) which is an obstacle to cross-referencing data from these different sources. In order to harmonize this heterogeneous data managed by different communities and make it FAIR into a common system, the Theia/OZCAR thesaurus , a controlled vocabulary for variable names and objects of interest was created. It implements the SKOS and I-ADOPT framework ontologies (Coussot et al. 2024), enabling precise and FAIR-compliant description of variable names. Like an in-situ observation, one or more composite pixels within a satellite image can be considered as a proxy of a measured phenomenon with one or more variables (soil moisture, landcover, biomass, tree height, etc). Although the community experts in this spatial research area understand the specification of a variable, the definition is not the same for all scientists and it is really useful to cross-reference all the definitions in order to sort, map and merge all this without loss of information. Finally, to highlight them with the production of a dictionary that can be easily shared on the web with a common vocabulary. The Theia spatial vocabulary provides a first thesaurus for the main variables calculated as proxy from remote sensing data (satellite or aerial). Alongside distributed infrastructure services, Data Terra RI provides discovery, access, and dissemination tools to enable researchers to effectively conduct their scientific investigations. Among these resources, EarthPortal, a FAIR-compliant semantic artefact catalogue, promotes the use of SKOS controlled vocabularies and OWL ontologies gathered in thematic categories and groups to enhance semantic interoperability across Earth sciences disciplines (Pierkot et al. 2023). Beyond providing access to semantic artefacts, EarthPortal offers advanced tools to support third-party applications: the annotator , which suggests relevant terms based on textual or keyword input; the recommender , which identifies pertinent semantic artefacts corresponding to the provided input, and the mappings tool , which generates, stores, and visualizes relationships between different semantic artefacts. Through its REST API, EarthPortal facilitates seamless integration with external applications, enabling users to access semantic artefacts and leverage these tools directly within their workflows. A concrete application example with the semantic artefacts provided by the Theia Hub will serve to highlight EarthPortal's functionalities with its user interface and API. To make the both vocabularies developed by the land surfaces data and services hub Theia available and reusable in third-party applications such as the data catalogue, they have been integrated into the EarthPortal. However, with both Theia/Ozcar and Theia/Spatial thesaurus there might be some redundancies in their entry metadata. To facilitate their management, a third entry regroups both thesauri to centralize the metadata and visualize both at once. It will not however replace them, they will instead appear as “views”, a feature that allows to make EarthPortal entries appear as part of another entry. The demonstration will proceed by illustrating how the tool enhances the interoperability of data by integrating EarthPortal with EaSy Data, the French data repository for Earth and Environmental sciences. This connection enriches metadata with semantic annotations, improving discoverability and user experience. Additionally, we will explore recent advancements, such as the federation of EarthPortal with other OntoPortal Alliance platforms to facilitate discovery of semantic artefacts across domains and the future development of a generic connector for the GeoNetwork catalogue, which will allow EarthPortal's semantic artefacts and tools to be used directly, enhancing metadata editing, search and data processes.

In France, the national research and innovation program ‘OneWater - Eau Bien Commun’ (2022-2032) addresses a wide range of key scientific issues to help protect and manage water as a common good. This 10-year program is made up of several projects that will generate a large amount of highly heterogeneous data, ranging from sensor data to social science data, samples and model-based data. It will also draw on existing data documenting the state of water resources, produced by observatories and research infrastructures (e.g. the OZCAR network of critical zones or the Réseau des Zones Ateliers in France), but also by operational public monitoring services run by the French State for its national Water Information System and by regional stakeholders. All the data are not yet fully available according to the FAIR (Findable, Accessible, Interoperable, Reusable) principles (Wilkinson et al. 2016). To process, share and re-use heterogeneous data sets and ultimately generate new knowledge, a dedicated project has been set up within the OneWater program to establish the national ‘OneWater FAIR Water Data Platform’. The ambition of this project is to go beyond a simple data catalog by fostering a FAIR Water Data ecosystem based on international standards and implementing semantic web interoperability in order to enhance data discoverability and use. The proposed approach is to produce, as far as possible, data that is compliant with FAIR principles by DNA. The OneWater data platform will be fully integrated in the national research data ecosystem dealing with research data, which relies on the Data Terra Research Infrastructure and its data hubs. The land surfaces THEIA data hub with its remote and in situ data components (Theia/OZCAR Information System (IS), Braud et al. 2020) are particularly relevant to the OneWater project. Collaboration will also be organized with the services supporting national public policy data and associated monitoring networks (French Ministry of Environment). At the international level, connection with the international community is being established so that the OneWater initiative can contribute to and benefit from the FAIR Water community. This includes the Open Geospatial Consortium Hydrology Domain Working Group (OGC HydroDWG), the Research Data Alliance Global Water Information Interest Group (RDA GWIIG), the World Meteorological Organization (WMO), the Integrated European Long-Term Ecosystem, critical zone and socio-ecological Research Infrastructure (eLTER RI), the Terrestrial Environmental Observatories Network (TERENO) amongst others. The OneWater data project does not only address data and information technology needs, but is also committed to supporting the water community to achieve creating an ecosystem of standards, tools and resource persons. Actions will help identify and overcome barriers that hinder the streamlined exchange of FAIR Water Data and the adoption of best practices. This will be achieved by building up a national overview of water data use practices, tools and obstacles for both researchers and operational stakeholders, then by organizing training sessions and supporting the community so that the tools they traditionally use evolve towards FAIR practices. In order to gather the needs of the platform users, a survey was sent to the water community in order to gather their requirements in terms of the data products and services needed to facilitate their work (data collection, data analysis, model set up, calibration and evaluation). The results of the survey were analysed to define several use cases that will serve as examples to build the OneWater platform iteratively, regularly gathering users' feedback and taking it into account to ensure that the developed platform truly meets their needs. To facilitate interaction with the user community, a public gitlab has been set up. One working group addressed the definition of FAIRness within the OneWater Data community by specifying how the FAIR principles of Wilkinson et al. (2016) are interpreted for Water data in the light of the existing international standards and best practices. It will also propose FAIRness grids analysis for various numerical resources (data sets, software, catalogs, vocabularies) after an analysis of existing on-line tools that often work as black boxes and do not provide enough information on how they have interpreted and evaluated the FAIR principles. The proposed grids were tested with data sets from the national groundwater monitoring system and the Theia/OZCAR IS (Braud et al. 2020) and with the Theia/OZCAR thesaurus that implements the RDA I-ADOPT recommendation for the description of Observable Properties (Coussot et al. 2024). OneWater's FAIR Data platform architecture is defined around well-known and most recent interoperability standards and practices (OGC, W3C, RDA, INSPIRE, etc.). It also takes into account the latest FAIR Open Science prototyping exercises such as the use of platforms like GALAXY and the associated recommendations (Royaux et al. 2024) and the results of pilot projects such as the OGC Open Science Persistent Demonstrator. In order to gather the needs of the platform users, a survey was sent to the water community in order to gather their requirements in terms of the data products and services needed to facilitate their work (data collection, data analysis, model set up, calibration and evaluation). The results of the survey were analysed to define several use cases that will serve as examples to build the OneWater platform iteratively, regularly gathering users' feedback and taking it into account to ensure that the developed platform truly meets their needs. To facilitate interaction with the user community, a public gitlab has been set up. One working group addressed the definition of FAIRness within the OneWater Data community by specifying how the FAIR principles of Wilkinson et al. (2016) are interpreted for Water data in the light of the existing international standards and best practices. It will also propose FAIRness grids analysis for various numerical resources (data sets, software, catalogs, vocabularies) after an analysis of existing on-line tools that often work as black boxes and do not provide enough information on how they have interpreted and evaluated the FAIR principles. The proposed grids were tested with data sets from the national groundwater monitoring system and the Theia/OZCAR IS (Braud et al. 2020) and with the Theia/OZCAR thesaurus that implements the RDA I-ADOPT recommendation for the description of Observable Properties (Coussot et al. 2024). OneWater's FAIR Data platform architecture is defined around well-known and most recent interoperability standards and practices (OGC, W3C, RDA, INSPIRE, etc.). It also takes into account the latest FAIR Open Science prototyping exercises such as the use of platforms like GALAXY and the associated recommendations (Royaux et al. 2024) and the results of pilot projects such as the OGC Open Science Persistent Demonstrator. This communication will present the general approach that is being implemented in the OneWater data project and highlight first results obtained since the beginning of the project.

Trees provide essential ecosystemic services (e.g. stocking carbon or locally regulating temperatures) and play an important role in the resilience to climate change, especially in urban areas. Quantifying these services in cities is difficult because little information is known about each tree, and no allometric model yet exists for urban trees.The UrTrees project calls the citizen to the rescue to help collect measurements and increase our knowledge of urban tree key features. Using the mobile app that we have designed, only a short video around the tree is necessary to approximate three key measurements: the tree height, its diameter at breast height (1m30 above ground) and the crown volume. No expertise in trees is required to use UrTrees, which has even been tested on children from 6 years old.A 3-dimensional point cloud of the scanned tree is first derived from the video using Structure-from-Motion algorithms. Geometric models for the trunk and the crown are then fitted to the point cloud in order to estimate the key measurements.Additionally, Pl@ntNet can be used to identify tree species. All measurements are stored in a database providing data for urban tree studies and feedback to the mobile app user. Efforts have been put into the mobile app user experience, with a scoring system, daily quests and point cloud interactive visualisation. Individual tree information collected through the app will be made freely available to the general public.

While satellite time series are essential tools to derive phenometrics at unprecedented spatial and temporal scales, non-systematic acquisition or medium spatial resolution of available missions are potentially problematic. At the same time, low-cost observation networks bridge the gap between satellite and in-situ observations which considerably increases ground-based data and associated possibilities. Here we provide robust statistics about the reliability of satellite derived phenometrics of urban trees across phenophases. Environmental and acquisition factors influencing the quality of phenometrics estimates were analyzed. First, a multi-facet regression-based analysis was conducted to measure discrepancies between PlanetScope (and Sentinel-2) and ground-based measurements across phenophases. Then, we performed hierarchical partitioning to tackle the effects of biological parameters (canopy closure and colour leaf) for assessing phenometrics with satellite time series. Third, we ran Monte-Carlo simulations to propagate errors according to viewing angles in PlanetScope acquisition. Our results show that: (i) PlanetScope provide consistent phenometric estimates for different tree layouts belonging to the same species (average $R^{2}$ = 0.50 $\pm$ 0.18). Performances are higher than Sentinel-2 but duration-based phenometrics estimates were poorly reconstructed with both satellite missions, (ii) contributions of biological parameters in the vegetation signal above trees strongly vary between growth periods. While canopy closure drives the growing season signal (independent contribution $>$ 40%), colour leaf plays a major role in the senescence season, (iii) variable viewing angles in PlanetScope acquisitions showed only significant effects on duration-based metrics estimates. Our research opens new perspectives for monitoring urban trees which improves the measurement of ecosystem services for local inhabitants.

The size of urban settlements is rapidly increasing worldwide. This sprawl triggers changes in land cover with the consumption of natural areas and affects ecosystems with important ecological, climate, and social transformations. Detecting, mapping, and monitoring the growth and spread of urban areas is therefore important for urban planning, risk analysis, human health, and biodiversity conservation. Satellite images have long been used to map human settlements. The availability of the Sentinel constellation (S2) allows the monitoring of urban sprawl over large areas (e.g., countries) and at high frequency (with possible monthly updates). This massive data stream allows the proposal of new types of urban products at a spatial resolution of 10 meters. In this context, we developed a fully automated and supervised processing chain (URBA-OPT) using open-source libraries and optimized for rapid calculation on high-performance computing clusters. This processing chain has been applied to all of France. The objective is to propose an annual product at 10 m spatial resolution available through the THEIA Data and Services Centre. Our results demonstrate the feasibility and accuracy of producing an annual urban settlement map using this methodology, providing a valuable tool for urban planning and environmental monitoring.

Urban areas are increasing since several years as a result of development of built-up areas, network infrastructure, industrial areas or other built-up areas. This urban sprawl has a considerable impact on natural areas by changing the functioning of ecosystems. Mapping and monitoring Urban Fabrics (UF) is therefore relevant for urban planning and management, risk analysis, human health or biodiversity. For this research, Sentinel-2 (level 2A) single-date images of the East of France, with a high spatial resolution (10 m), are used to assess two semantic segmentation networks (U-Net) that we combined using feature fusion between a from scratch network and a pre-trained network on ImageNet. Moreover three spectral or textural indices have been added to the both networks in order to improve the classification results. The results showed a performance gain for the fusion methods in classifying several UF. However, there is a difference in performance depending on the urbanization gradient; highly urbanized areas provide a better distinction of some UF’s classes than rural areas.