Currently, more than two billion people live in or near-coastal zones at the ocean-land interface with almost a billion more living in adjacent low-lying coastal areas. These areas and populations are at risk from increasing storminess and longer-term sea level rise resulting in coastal erosion, water pollution, coastal inundation and ecosystems degradation. Earth Science Digital Twins, that is, the combination of data, models, and AI/ML technologies to simulate Earth system processes and enables short- and long-term forecasts, provide understanding and actionable information to reduce risks to humans, infrastructure and ecosystems. Satellite and in situ data are critical components of a coastal zone digital twin (CZDT) to provide timely and spatially relevant input data for models and serve as a check or validation of digital twin performance. This paper introduces a CZDT concept being developed as part the Space for Climate Observatory with joint participation of CNES, NASA and NOAA and shows initial use cases, satellite data and advanced technology components.

An Earth System Digital Twin (ESDT) is a dynamic, interactive, digital replica of the state and temporal evolution of Earth systems. It integrates multiple models along with observations, and connecting them with analysis, AI, and visualization tools. Together, these enable users to explore the current state of the Earth system, predict future conditions, and run hypothetical scenarios to understand how the system would evolve under various assumptions. Since 2021, NASA's Advanced Information Systems Technology (AIST) program has invested in two ESDT efforts to tackle the impacts of our changing climate. The establishment of ESDT for flood and air quality enabled our teams to formalize the software framework. The open-source framework is called the Integrated Digital Earth Analysis System (IDEAS). By working with the Apache Science Data Analytics Platform (SDAP) community, IDEAS is now a subproject of SDAP. The paper presents the ongoing development of IDEAS and its current applications.

As the severity and occurrence of flood events tend to intensify with climate change, the need for flood forecasting capability increases. In this regard, the Flood Detection, Alert and rapid Mapping (FloodDAM) project, funded by Space for Climate Observatory initiatives, was set out to develop pre-operational tools dedicated to enabling quick responses in flood-prone areas, and to improve the reactivity of decision support systems. This work focuses on the assimilation of 2D flood extent data (expressed in terms of wet surface ratios) and in-situ water level data to improve the representation of the flood plain dynamics with a Telemac-2D model and an Ensemble Kalman Filter (EnKF). The EnKF control vector was composed friction coefficients and corrective parameter to the input forcing. It is then augmented with the water level state averaged over several floodplain zones. This work was conducted in the context of Observing System Simulation Experiments (OSSE) based on a real flood event occurred in January-February 2021 on the Garonne Marmandaise catchment. This allows to validate the observation operator associated to the wet surface ratio observations as well as the dual state-parameter sequential correction implemented in this work. The merits of assimilating SAR- derived flood plain data complementary to in-situ water level observations are shown in the control parameter and observation spaces with 1D and 2D assessment metrics. It was also shown that the correction of the hydraulic state significantly improved the flood dynamics, especially during the recession. This proof-of-concept study paves the way towards near-real-time flood forecast, making the most of remote sensing-derived flood observations.

<p>The notion of digital twin can be ambiguous because it can be defined in various ways. These last months have seen the emergence of many global digital twin initiatives. The challenge of these global digital twins is to create a qualified digital replica model of our planet, making it possible to monitor, simulate and anticipate natural phenomena and human activities. The target users are either scientists or decision makers. Through the digital twin, they have access to a digital representation of an environment using all available spatial and non-spatial data accompanied with a set of physical and statistical models to calculate projections, replay past events or simulate future ones.</p> <p> </p> <p>Refining and evaluating the accuracy of these projections is a major challenge for digital twins. In addition to the knowledge of physical modeling, suitable data must also be available. Complementary to the global approach, the notion of local and dated digital twins appears then to be essential. Considering a digital representation of a restricted geographical area of interest (an urban area, watershed, coastline, etc.) allows to access to very high-resolution "fresh" data in 2D and 3D, in-situ data and small-mesh physical model. This user-centered and naturally thematic approach responds more finely and more pragmatically to the objectives presented. These local, dated and thematic digital twins are by essence ephemeral: a way to meet a specific need.</p> <p> </p> <p>The challenge is therefore to setup a Digital Twin Factory (DTF). This DTF relies on a data lake, a high computing capacity via clouds and/or HPC and has thematic algorithms and methodologies able to generate registered and coherent layers of information in order to enrich a datacube from which physical indicators can be computed spatially. Thanks to its thematic, local and on-demand characteristics, the DTF can mitigate the need to have an universal model of metadata. This datacube allows to apply local physical and artificial intelligence models. The overarching architecture of the DTF will be presented. Specific examples on coastal, urban and risk topics will also be presented. These digital twins rely on a large number of expertises in both data and modeling involving various French (CNES, IGN, SHOM, IRD, CEA, INRAE, METEOFRANCE, CERFACS, BRGM, etc.) or international organizations (ESA, NASA, NOAA,…).</p> <p>For coastal areas, the goal is to well describe the bathymetry topography continuum by taking into account the intertidal zones and the specialized dynamic models together with 3D coastal land cover characterisation. For urban areas, the ambition is first to automatically produce a qualified 3D map together with its additional layers of information: 3D objects and related semantics (land cover and land use) including temporal dynamic, thermal information. Then, for issues related to the management of natural risks (such as floods or fires) similar data layers can be used. Finally, new hypothesis can be injected in these digital replica and multiple scenarios can be applied to assess causal relationship between hypothesis and prediction. Very promising results will also be presented.</p>

3D Geospatial information plays a key role in many soaring sectors such as sustainable and smart cities, climate monitoring, ecological mobility, and economic intelligence. The availability of huge volumes of satellite, airborne and in-situ data now makes this production feasible at large scale. It needs nonetheless a certain level of manual intervention to secure the level of quality, which prevents mass production. This paper presents the AI4GEO program that aims at developing an end to end solution to produce automatically qualified 3D Digital model at scale together with multiple layers of information.

As the severity and occurrence of flood events tend to intensify with climate change, the need for flood forecasting capability increases. In this regard, the Flood Detection, Alert and rapid Mapping (FloodDAM) project, funded by Space for Climate Observatory initiatives, was set out to develop pre-operational tools dedicated to enabling quick responses in flood-prone areas, and to improve the reactivity of decision support systems. This work focuses on the assimilation of 2D flood extent data (expressed in terms of wet surface ratios) and in-situ water level data to improve the representation of the flood plain dynamics with a Telemac-2D model and an Ensemble Kalman Filter (EnKF). The EnKF control vector was composed friction coefficients and corrective parameter to the input forcing. It is then augmented with the water level state averaged over several floodplain zones. This work was conducted in the context of Observing System Simulation Experiments (OSSE) based on a real flood event occurred in January-February 2021 on the Garonne Marmandaise catchment. This allows to validate the observation operator associated to the wet surface ratio observations as well as the dual state-parameter sequential correction implemented in this work. The merits of assimilating SAR- derived flood plain data complementary to in-situ water level observations are shown in the control parameter and observation spaces with 1D and 2D assessment metrics. It was also shown that the correction of the hydraulic state significantly improved the flood dynamics, especially during the recession. This proof-of-concept study paves the way towards near-real-time flood forecast, making the most of remote sensing-derived flood observations.

An Earth System Digital Twin (ESDT) is a dynamic, interactive, digital replica of the state and temporal evolution of Earth systems. It integrates multiple models along with observation data, and connecting them with analysis, AI, and visualization tools. Together, these enable users to explore the current state of the Earth system, predict future conditions, and run hypothetical scenarios to understand how the system would evolve under various assumptions. The NASA's Advanced Information Systems Technology (AIST)'s Integrated Digital Earth Analysis System (IDEAS) project is to establish an extensible architectural solution to develop digital twins of our physical environment for Earth Science. IDEAS delivers a formal system architecture with mechanisms for the outputs of one model to feed into others; for driving models with observation data; and for harmonizing observation data and model outputs for analysis. To validate and demonstrate the IDEAS architecture, this project collaborates with the Space Climate Observatory (SCO)'s FloodDAM project and the Centre National d'Etudes Spatiales (CNES) to focus on floods detection, prediction and their impacts.

Flooding is one of the most devastating natural hazards to which our society worldwide must adapt, especially as its severity and occurrence tend to increase with climate changes. This research work focuses on the assimilation of two‐dimensional (2D) flood observations derived from remote‐sensing images acquired during overflowing events. To do so, the resulting binary wet/dry maps are expressed in terms of wet surface ratios (WSR) over a number of floodplain subdomains. This ratio is assimilated jointly with in‐situ water‐level gauge observations to improve the flow dynamics within the floodplain. An Ensemble Kalman Filter (EnKF) with a dual state‐parameter analysis approach is implemented on top of a TELEMAC‐2D hydrodynamic model. The EnKF control vector is composed of spatially‐distributed friction coefficients and a corrective parameter of the inflow discharge. It is extended with the hydraulic states within the floodplain subdomains. This data assimilation strategy was validated and evaluated over a reach of the Garonne river. The observation operator associated with the WSR observations, as well as the dual state‐parameter sequential correction, was first validated in the context of Observing System Simulation Experiments. It was then applied to two real flood events that occurred in 2019 and 2021. The merits of assimilating synthetic aperture radar‐derived WSR observations, in complement to the in‐situ water‐level observations, are shown in the parameter and observation spaces with assessment metrics computed over the entire flood events. It is also shown that the hydraulic state correction within the dual state‐parameter analysis approach significantly improves the flood dynamics, especially during the flood recess.

The availability of 3D Geospatial information is a key issue for many expanding sectors such as autonomous vehicles, business intelligence and urban planning. Its production is now possible thanks to the abundance of available data (Earth observation satellite constellations, insitu data, …) but manual interventions are still needed to guarantee a high level of quality, which prevents mass production. New artificial intelligence and big data technologies adapted to 3D imagery can help to remove these obstacles. The AI4GEO project aims at developing an automatic solution for producing 3D geospatial information and new added-value services. This paper will first introduce AI4GEO initiative, context and overall objectives. It will then present the current status of the project and in particular it will focus on the innovative platform put in place to handle big 3D datasets for analytics needs and it will present the first results of 3D semantic segmentations and associated perspectives.

Floods are the most common natural disasters all over the world. The space climate observatory (SCO)-FloodDAM project aims at utilizing the capabilities of new observing strategies in order to better alert, detect and map flood events globally. Leveraging from both aerial- and satellite-based platforms (Sentinel, TerraSar-X, SWOT) as well as in-situ based sensors, the main objective of the project is to develop an automatic system to better prevent flood events and assess their consequences. In this paper, we will demonstrate the strategy deployed for the selected test-sites in France.

Pixels covered by clouds in optical Earth Observation images are not usable for most applications. For this reason, only images delivered with reliable cloud masks are eligible for an automated or massive analysis. Current state of the art cloud detection algorithms, both physical models and machine learning models, are specific to a mission or a mission type, with limited transferability. A new model has to be developed every time a new mission is launched. Machine Learning may overcome this problem and, in turn obtain state of the art, or even better performances by training a same algorithm on datasets from different missions. However, simulating products for upcoming missions is not always possible and available actual products are not enough to create a training dataset until well after the launch. Furthermore, labelling data is time consuming. Therefore, even by the time when enough data is available, manually labelled data might not be available at all. To solve this bottleneck, we propose a transfer learning based method using the available products of the current generation of satellites. These existing products are gathered in a database that is used to train a deep convolutional neural network (CNN) solely on those products. The trained model is applied to images from other - unseen - sensors and the outputs are evaluated. We avoid labelling manually by automatically producing the ground data with existing algorithms. Only a few semi-manually labelled images are used for qualifying the model. Even those semi-manually labelled samples need very few user inputs. This drastic reduction of user input limits subjectivity and reduce the costs. We provide an example of such a process by training a model to detect clouds in Sentinel-2 images, using as ground-truth the masks of existing state-of-the-art processors. Then, we apply the trained network to detect clouds in previously unseen imagery of other sensors such as the SPOT family or the High-Resolution (HR) Pleiades imaging system, which provide a different feature space. The results demonstrate that the trained model is robust to variations within the individual bands resulting from different acquisition methods and spectral responses. Furthermore, the addition of geo-located auxiliary data that is independent from the platform, such as digital elevation models (DEMs), as well as simple synthetic bands such as the NDVI or NDSI, further improves the results. In the future, this approach opens up the possibility to be used on new CNES’ missions, such as Microcarb or CO3D.

The availability of 3D Geospatial information is a key stake for many expanding sectors such as autonomous vehicles, business intelligence and urban planning.The availability of huge volumes of satellite, airborne and in-situ data now makes this production feasible on a large scale. It needs nonetheless a certain level of skilled manual intervention to secure a certain level of quality, which prevents mass production.New artificial intelligence and big data technologies are key in lifting these obstacles.The AI4GEO project aims at developing an automatic solution for producing 3D geospatial information and offer new value-added services leveraging innovative methods adapted to 3D imagery.The AI4GEO consortium consists of institutional partners (CNES, IGN, ONERA) and industrial groups (CS-SI, AIRBUS, CLS, GEOSAT, QWANT, QUANTCUBE) covering the whole value chain of Geospatial Information.With a 4 years’ timeline, the project is structured around 2 R&D axes which will progress simultaneously and feed each other.The first axis consists in developing a set of technological bricks allowing the automatic production of qualified 3D maps composed of 3D objects and associated semantics. This collaborative work benefits from the latest research from all partners in the field of AI and Big Data technologies as well as from an unprecedented database (satellite and airborne data (optics, radars, lidars) combined with cartographic and in-situ data).The second axis consists in deriving from these technological bricks a variety of services for different fields: 3D semantic mapping of cities, macroeconomic indicators, decision support for water management, autonomous transport, consumer search engine.Started in 2019, the first axis of the project has already produced very promising results. A first version of the platform and technological bricks are now available.This paper will first introduce AI4GEO initiative: context and overall objectives.It will then present the current status of the project and in particular it will focus on the innovative approach to handle big 3D datasets for analytics needs and it will present the first results of 3D semantic segmentations on various test sites and associated perspectives.

Through various initiatives, CNES, the French space agency, has been involved in major disaster and environment monitoring from Space for many years and in particular in the International Charter which delivers satellite data to nations affected by major disasters and in the THEIA organization which promotes the use of satellite data to monitor human and climate impacts on environment. Thus, CNES developed in 2014 a reference framework to generate value-added products from satellite data. Since 2017, considering the increase of data and processing requests, CNES has decided to move to a new framework based on big data and cloud technologies and extended to data analytics. This paper will first introduce THEIA and Charter initiatives. It will present the current framework in operation and its limitations. It will then focus on the innovative approach to handle big data and analytics needs and finally presents the first results and perspectives.

En télédétection optique, les images satellites à très haute résolution spatiale sont sujettes au bruit instrumental, ce qui nécessite l’application de processus de restauration après acquisition. Afin d’améliorer la qualité de la restauration et donc du produit final, il est nécessaire de retirer ce bruit instrumental tout en conservant les informations importantes contenues dans l’image telles que les textures, les discontinuités ou encore les zones homogènes. Dans les cas d’application réels des chaînes opérationnelles développées par le Centre National d’Études Spatiales (CNES) pour les missions Pléiades [2] et Spot 5 THR [5], de nouvelles techniques innovantes de restauration ont été introduites et validées. Dans cette étude, nous nous sommes concentrés sur la partie débruitage de ce processus de restauration, entre l’étape de restauration du bruit instrumental et l’étape de déconvolution. Le bruit instrumental de ces données a une variance dépendante au signal. Une première étape de stabilisation de la variance grâce à la transformée de Anscombe [1] nous permet de transformer ce bruit instrumental en bruit gaussien de moyenne nulle et de variance égale à un. Dans ce contexte de suppression de bruit blanc, plusieurs méthodes issues de l’état de l’art ont été étudiées [11] et la meilleure méthode identifiée est la méthode Non Local Bayes développée par Lebrun [7] en 2013. Cette méthode est une évolution de deux autres méthodes issues de l’état de l’art que sont Non Local Means (NLM) [3] et Block Matching and 3D filtering (BM3D) [6]. Comme nous le verrons dans nos expérimentations, cette méthode s’applique bien au domaine des images satellites optiques. NLB fournit les meilleurs résultats en termes de qualité de débruitage mais également de performance calculatoire. Dans un contexte de production massive de données satellitaires, le temps de calcul est un facteur primordial. Ainsi, dans le cadre de la chaîne Pléiades mais également celle de l’initiative CNES Spot World Heritage (SWH) [13] visant à retraiter toutes les données Spot 1 à 5, le coût de calcul prohibitif d’une méthode peut empêcher son utilisation en milieu opérationnel. Il convient ainsi de sélectionner la meilleure méthode et ses améliorations afin d’obtenir des résultats de qualité tout en minimisant le temps de calcul.

Very high resolution optical remote sensing images (RSI) are often corrupted by noise. Among popular denoising methods in the state of the art, nonlocal Bayes (NLB) has led to successful results on real datasets, with high quality and reasonable computation time. However, its computation time remains prohibitive with respect to requirements of operational RSI pipelines, such as Pléiades one. In this paper, we tackle such an issue and introduce several optimizations aiming to significantly reduce the computation time required by NLB while keeping the best denoising quality (i.e., preserving edges, textures, and homogeneous areas). More precisely, our improvements consist of reducing multiple estimations of a same pixel with a masking technique and modifying the spatial extent of the similar patch search area (i.e., one of the main parts of nonlocal algorithms, such as NLB). We report several experiments and discuss optimal settings for these parameters, allowing a gain in computation time of 50% (resp. 15%) with optimized masking strategy (resp. spatial extent of the search area). When both contributions are combined, we achieve the same denoising quality as standard NLB while doubling the computation efficiency, the latter being increased fivefold if we accept a very small (lower than 0.1%) loss in quality.

In the context of the Spot World Heritage (SWH) initiative, the long term archive of Spot 1 to 5 satellite data will be reprocessed. This initiative gives the opportunity to increase Spot data quality and production efficiency with the introduction of new state-of-the-art processing methods. In this context, we consider the Spot 5 Supermode processing chain and the introduction of a new denoising method. Our objective is to propose the best denoising method in terms of efficiency and accuracy for the Spot 5 Supermode data production. We report two experimentations that lead to the choice of the Non Local Bayes (NLB) denoising method; it is fast, accurate and remains stable in terms of efficiency and quality for different landscapes and image sizes. We also introduce two optimized versions of the NLB algorithm that increase the computation efficiency by a factor up to 4.

Restoration of Very High Resolution (VHR) optical Remote Sensing Image (RSI) is critical and leads to the problem of removing instrumental noise while keeping integrity of relevant information. Improving denoising in an image processing chain implies increasing image quality and improving performance of all following tasks operated by experts (photo-interpretation, cartography, etc.) or by algorithms (land cover mapping, change detection, 3D reconstruction, etc.). In a context of large industrial VHR image production, the selected denoising method should optimized accuracy and robustness with relevant information and saliency conservation, and rapidity due to the huge amount of data acquired and/or archived. Very recent research in image processing leads to a fast and accurate algorithm called Non Local Bayes (NLB) that we propose to adapt and optimize for VHR RSIs. This method is well suited for mass production thanks to its best trade-off between accuracy and computational complexity compared to other state-of-the-art methods. NLB is based on a simple principle: similar structures in an image have similar noise distribution and thus can be denoised with the same noise estimation. In this paper, we describe in details algorithm operations and performances, and analyze parameter sensibilities on various typical real areas observed in VHR RSIs.

Pleiades-HR is a high resolution remote sensing system developed by the French Space Agency (CNES) for civil and military users. The constellation is composed of two identical satellites PHR1A launched on 2011, December 17th and PHR1B launched one year after, on 2012, December 2nd. More than 600 images can be daily acquired by each satellite in various viewing angles conditions: the satellites are able to target images with viewing angles greater than 47°. Since the launch of the first satellite, major improvements have been integrated in the ground processing to enhance the quality of products and offer new options. Starting from a Pleiades-HR system overview, this paper offers a description of the ground processing and presents an assessment of the different products available now, including image quality performances.

Launched on 2011, December 16th by the second Soyuz operated in French Guiana, PLEIADES-HR acquired its first high resolution images 3 days after. The PLEIADES program is a space Earth Observation system led by the French Space Agency (CNES), it provides the first european high resolution satellite which will simultaneously acquire in Panchromatic and Multi-Spectral modes 20 km wide images with a 70 cm nadir resolution.Imaging capabilities have been highly optimized in order to acquire, in the same pass, along-track mosaics, stereo pairs and triplets, and multi-targets. Ground segment processes automatically data to ensure operational requirements and quick access to images for the users. Since ground processing capabilities have been taken into account very early in the mission development, it has been possible to relax some costly on-board components requirements, in order to achieve a cost effective on-board/ground compromise. Starting from the PLEIADES system and on board characteristics, this paper first presents an overview of ground segment functional breakdown. Then it focuses more precisely on the different levels of products and associated processing performances. Finally the paper shows how appropriate ground processing systems allowed CNES Image Quality team to assess radiometric and geometric performances during the 6 first months of PLEIADES 1A.

In partnership with the European Commission and in the frame of the Global Monitoring for Environment and Security (GMES) program, the European Space Agency (ESA) is developing the Sentinel-2 optical imaging mission devoted to the operational monitoring of land and coastal areas. The Sentinel-2 mission is based on a satellites constellation deployed in polar sun-synchronous orbit. Sentinel-2 will offer wide improvements such as a unique combination of global coverage with a wide field of view (290km), a high revisit (5 days with two satellites), a high resolution (10m, 20m and 60m) and multi-spectral imagery (13 spectral bands in visible and shortwave infra-red domains). In this context, the Centre National d'Etudes Spatiales (CNES) supports ESA to define the system image products and to prototype the relevant image processing techniques.