European Space Research Institute

Abstract Medicanes are cyclones with tropical characteristics in the Mediterranean basin. They typically show an eye with neighboring spiraling bands, but the surrounding region of maximum winds is not well documented, especially near the sea surface, due to limited available observational data. Estimating such characteristics accurately is essential for a better understanding of the specific fine‐scale wind structures and their relationship with the development mechanisms of medicanes. Here we first demonstrate the capacity of synthetic aperture radar (SAR) spaceborne instruments to diagnose important medicane structural characteristics. Building on an unprecedented collection of SAR high‐resolution observations and subsequent sea‐surface wind‐speed estimates, the eye patterns, heavy rainfall, convective cells, and boundary‐layer roll vortices are evidenced, with attributes comparable with those of tropical cyclones of similar intensities. Motivated by this resemblance, we estimate parameters characterizing the cyclone surface wind structure (the intensity, inner and outer size, and Rossby number) from high‐resolution observations of medicanes. We show that their inner size is of magnitude comparable with that of tropical cyclones of similar intensities. Conversely, typical outer sizes and Rossby numbers are found to be smaller. The misrepresentation of these characteristics in medium‐resolution global reanalyses brings into question their suitability and the application of diagnostic tools to these data to monitor the life cycles of medicanes. This invites us to develop dedicated observation‐based monitoring frameworks. Medium‐resolution scatterometers provide more reliable structural estimates despite the small medicane size. Along with SAR surface wind‐speed estimates, and complemented by upper‐level measurements, they may support future reanalysis efforts. Further investigation is needed to understand how the high‐resolution features reported in our work connect with kilometer‐ and subkilometer‐scale physical processes.

Abstract. The advent of big data in Earth Observation (EO), coupled with recent advances in Artificial Intelligence, has led to the development of geospatial embeddings that are compact, information-rich feature vectors designed to be ready-to-use in machine learning (ML) applications for a wide range of downstream tasks, including forest monitoring. Motivated by the limitations of existing global forest products and by policy requirements such as the EU Deforestation Regulation (EUDR), we assess whether lightweight classifiers applied to satellite embeddings from the Google DeepMind Alpha Earth Foundation (AEF) can accurately map global forest and tree crop extents. In this study, we introduce GEM-Forest, a global satellite embedding–based forest dataset in 10 m spatial resolution for 2020, and its associated products: GEM-FnF2020, a forest / non-forest (F/nF) classification, and GEM-TC2020, which further distinguishes non-forest areas containing tree crops. Using ∼47,000 globally distributed training samples covering all major biomes, collected through an automated approach combining multiple forest-related, land cover and tree crop datasets, we compared multiple ML approaches ranging from linear models to neural networks. Accuracy assessment on a global F/nF dataset with ∼21,000 samples showed similar performance across classifiers, with overall accuracies of 90–92 % and macro F1-scores of 0.89–0.90, while linear models often outperformed more complex approaches. The validation of the tree crop subclass across 10 datasets showed larger differences among different ML models, with the highest accuracies achieved mostly by linear models. This consistency indicates that the embeddings encode highly informative and linearly separable structure for global F/nF discrimination, including tree-crop separation. A linear Support Vector Machine was therefore used to generate GEM-FnF2020 that achieves a 91 % overall accuracy, a macro F1-score of 0.90, with balanced omission and commission error rates for forests (15 % and 13 %, respectively). These results match or exceed existing global products, with most errors occurring in open forests and forest–shrubland transition zones. Residual misclassifications of tree crops as forests in GEM-FnF2020 ranged from 0.5 % to 14.8 %, which demonstrates the importance of including the tree crop subclass in the GEM-TC2020 map. The GEM-TC2020 enables distinction of agricultural tree crops with an overall accuracy higher than 85 % for most tree crops, while the classification of European tree crops remains the most challenging. The classified tree crop class significantly improves the commission error rates in the main GEM-FnF2020 product (0.5–14.8 %). Our proposed approach demonstrates strong potential for temporal transferability across the 2017–2025 period covered by AEF embeddings. This capability allows multi-year applications and change detection based on models trained for a single year and represents a key next step in our research. Overall, the findings demonstrate that AEF embeddings combined with simple ML approaches support accurate, transferable, and computationally-efficient global forest mapping, with remaining limitations related to temporal resolution and feature interpretability. These results and the presented approach can support policy and regulatory decisions, including the EUDR, while the open-access release of the GEM-Forest datasets and trained models facilitates global use, further testing, and methodological development by the EO and forest monitoring communities.

The recent launches of Sentinel-4 on MTG-S and Sentinel-5 on EPS-SG-A1 significantly strengthen the Copernicus contribution to the international constellation of low-Earth orbiting (LEO) and geostationary (GEO) satellites dedicated to atmospheric ozone and air quality monitoring. Together with the already operational Sentinel-5P LEO and GEMS (East Asia) and TEMPO (North America) GEO missions, these nadir sounders form an unprecedented observing system coordinated to ensure complementary spatial and temporal sampling and long-term commitment. As a prerequisite to a truly integrated exploitation of this constellation, the present contribution reports on two complementary ozone data validation activities undertaken to build confidence in their traceability to community-agreed standards and in the mutual consistency across missions.As a first step, we summarize the global, in-depth, and recurrent validation of Sentinel-5P TROPOMI total ozone column data, conducted by the ESA/Copernicus Atmospheric Mission Performance Cluster (ATM-MPC). The ATM-MPC validation service assesses TROPOMI’s traceability to the well-established ground-based Brewer, Dobson, and ZSL-DOAS measurements contributing to WMO’s Global Atmosphere Watch (GAW) and the Network for the Detection of Atmospheric Composition Change (NDACC), and to measurements from the more recent Pandonia Global Network (PGN). The permanent QA/QC of TROPOMI demonstrates the value of this mission as the initial reference sounder for the constellation. As a second step, we present pioneering regional validation and evaluation results for the GEO missions GEMS and TEMPO performed within ESA’s PEGASOS project. This latter activity integrates independent ground-based validation and the use of TROPOMI data as a LEO transfer standard for cross-mission consistency assessment.The ATM-MPC results confirm the high quality of TROPOMI total ozone observations, characterized by low bias, small random uncertainty, and temporal stability meeting community requirements, both with respect to ground-based reference measurements and relative to the historical benchmark provided by the Aura OMI mission. Building on these resources – both the ground-based validation infrastructure and TROPOMI as well characterized and accurate measurement system- we assess the quality of the total ozone products from GEMS and TEMPO. The analysis demonstrates overall good data quality across both GEO missions, while also identifying dependences of bias and dispersion on measurement parameters and influence quantities. These findings highlight specific areas where further algorithm refinements may enhance consistency and performance within the emerging global GEO–LEO ozone observing system.We would like to acknowledge explicitly the long-term dedication of all the ground-based instrument teams (WMO-GAW, NDACC, PGN) to acquire high quality data and make them available to the satellite community.

Accurately characterizing atmospheric aerosols remains a primary challenge in achieving high precision in satellite retrievals of XCO2 and XCH4. While Shortwave Infrared (SWIR) spectral bands provide optimal sensitivity to greenhouse gas concentrations, Multi-Angular Polarimetric (MAP) measurements represent the most advanced approach for constraining aerosol and surface properties. Consequently, upcoming Copernicus missions, such as CO2M and MetOp-SG (hosting Sentinel-5/UVNS spectrometer and 3MI polarimeter), are equipped to exploit the synergy between these measurement types.In this context, we present an open-source, full-physics retrieval algorithm for GHG retrievals to demonstrate its capabilities through the synergistic inversion of SWIR spectrometric (i.e., S5-UVNS and CO2M/CO2I spectrometers) and multi-angle polarimetric data (3MI and CO2M/MAP). The core of this retrieval approach is the Generalized Retrieval of Atmosphere and Surface Properties (GRASP) algorithm (Dubovik et al., 2021). GRASP relies on rigorous radiative transfer modeling and statistically optimized fitting (Multi-Term Least Squares) to process a wide range of optical instruments. While originally designed for MAP applications (e.g., POLDER, 3MI; Li et al., 2019; Chen et al., 2020), the approach has been extended to combine MAP and SWIR spectrometric measurements, accounting for gas absorption, offering a simultaneous retrieval of detailed aerosol microphysics and surface properties alongside with columnar XCO2 and XCH4.In this work the GRASP algorithm has been successfully adapted to perform gas retrievals, allowing for the robust simultaneous inversion of aerosol properties and XCO2 and XCH4. We demonstrate the capabilities of this new GRASP inversion scheme through synthetic datasets representing the Sentinel-5+3MI and CO2M synergies. These synthetic experiments highlight that incorporating MAP measurements yields significantly improved accuracy for XCO2 and XCH4 compared to standalone spectrometer retrievals. This project has been funded through the OPERA-S5 project (OPEn platform for the Retrieval of Aerosol and CO2 from S5), an ESA-funded initiative developed by GRASP SAS company and SRON. This development lays the groundwork for the operational processing of future multi-sensor satellite missions, offering a powerful tool for enhanced global monitoring of greenhouse gases and aerosol interactions. ReferencesChen, C., Dubovik, O., Fuertes, D., Litvinov, P., Lapyonok, T., Lopatin, A., ... & Federspiel, C. (2020). Validation of GRASP algorithm product from POLDER/PARASOL data and assessment of multi-angular polarimetry potential for aerosol monitoring. Earth System Science Data Discussions, 2020, 1-108.Dubovik, O., D. Fuertes, P. Litvinov, et al., “A Comprehensive Description of Multi- Term LSM for Applying Multiple a Priori Constraints in Problems of Atmospheric Remote Sensing: GRASP Algorithm, Concept, and Ap-plications”, Front. Remote Sens. 2:706851. doi: 10.3389/frsen.2021.706851, 2021.Li, Lei; Derimian, Yevgeny; Chen, Cheng; Zhang, Xindan; Che, Huizheng; Schuster, Gregory L.; Fuertes, David; Litvinov, Pavel; Lapyonok, Tatyana; Lopatin, Anton; Matar, Christian; Ducos, Fabrice; Karol, Yana; Torres, Benjamin; Gui, Ke; Zheng, Yu; Liang, Yuanxin; Lei, Yadong; Zhu, Jibiao; Zhang, Lei; Zhong, Junting; Zhang, Xiaoye; Dubovik, Oleg Climatology of aerosol component concentrations derived from multi-angular polarimetric POLDER-3 observations using GRASP algorithm. Earth Syst. Sci. Data, vol. 14, no. 7, pp. 3439–3469, 2022, ISSN: 1866-3516.

Oceans remain severely under-sampled in aerosol optical properties, limiting our ability to assess marine aerosol–climate feedbacks, long-range transport and to validate satellite products over two-thirds of Earth’s surface. The Sun–sky–lunar photometer CIMEL CE318-T has been successfully adapted for autonomous shipborne operation through the PHOTONS Observation Service (ACTRIS, Univ. Lille, CNRS) in collaboration with CIMEL and within ACTRIS-CARS activities. The system is fully compatible with AERONET processing. Here we summarize recent developments and multi-year results supporting the deployment of a global network of automatic ship-based photometers.Since 2021, a CE318-T has operated continuously on the research vessel (R.V.) Marion Dufresne under the framework of MAP-IO program. Between July 2021 and June 2024, it collected >25,000 Level 1.5 measurements in the southwestern Indian Ocean, with mean AOD (Aerosol Optical Depth) of 0.09 ± 0.07 (440 nm) and 0.05 ± 0.03 (870 nm), and EAE (Extinction Angstrom Exponent) of 0.7 ± 0.4, typical of clean marine conditions. A biomass-burning aerosol (BBA) event allowed retrieval of microphysical properties (size distribution, refractive index), showing the capability of the system under real ship-motion.During TRANSAMA campaign (La Réunion–Barbados, Apr–May 2023), two CE318-T photometers combined with a micropulse lidar aboard R.V. Marion Dufresne provided complementary column and vertical information. Despite low AOD (0.08 ± 0.04 at 440 nm) in the remote South Atlantic, the lidar detected transported continental layers not evident from column-integrated data alone. Photometer intercomparison showed excellent agreement (R > 0.96; RMSE = 0.005–0.008). Motion-induced degradation on data quality highlighted the need for the ongoing instrumental tests, using a motion-simulation hexapode platform.A second permanent site aboard the R.V. Gaia Blu (CNR, Italy) has collected >20,000 measurements mainly in the Mediterranean since Feb 2024. Mean AOD (0.19 ± 0.14 at 440 nm) and EAE (1 ± 0.4) reflect more variable aerosol regimes, including BBA and Saharan dust. Comparisons with nearby AERONET ground-based stations validated retrieval quality. In addition, the installation of a 3D scanning lidar (Sep 2025) further enhances observations of aerosol vertical structure and air–sea interactions.These first results demonstrate the capability of ship-based photometers, and their synergy with lidar, to fill critical observational gaps over the oceans. Improved data acquisition strategies addressing vessel structure and motion are ongoing. To date, five ship-photometers operate across major maritime regions: R.V. Marion Dufresne-France (Indian Ocean), R.V. Gaia Blu-Italy (Mediterranean Sea), R.V. Sarmiento de Gamboa-Spain (Atlantic Ocean), R.V. Perseverance-France (currently Pacific Ocean), and MS Richard With-Norway (Arctic/Norwegian coast). These installations represent a major European step toward the first global network of automatic shipborne photometers to improve aerosol characterization over remote oceans and support satellite CAL/VAL.

The increasing volume and diversity of Earth Observation (EO) and climate data produced by Copernicus missions and the Destination Earth (DestinE) initiative pose a major challenge for interoperability and large-scale analysis. Today, global datasets are distributed on heterogeneous spatial grids, forcing users to repeatedly perform ad-hoc regridding steps, which are costly, error-prone, and difficult to reproduce.The GRID4EARTH project addresses this issue by promoting a common Discrete Global Grid System (DGGS) as a foundation for analysis-ready Earth data. In this context, HEALPix emerges as a strong candidate due to its equal-area property, hierarchical structure, and long-standing adoption in global modelling and large-scale data analysis. These properties enable efficient multi-resolution workflows, scalable data access, and natural integration with modern cloud-native formats such as Zarr.However, a key limitation remains: HEALPix is formally defined on the sphere, whereas EO are naturally referenced to the WGS84 ellipsoid. While often ignored at coarse resolutions, this mismatch introduces non-negligible area distortions at Copernicus resolutions and may bias zonal or regional analyses.To overcome this limitation, GRID4EARTH explores anextension of HEALPix to the WGS84 ellipsoid using the authalic sphere associated with the ellipsoid. By preserving equal-area properties on the ellipsoid, this approach provides a consistent spatial framework bridging spherical climate models and ellipsoidal EO. It enables a unified representation for DestinE model outputs and Copernicus satellite data, while remaining compatible with existing HEALPix-based tools and workflows.This contribution presents the motivation, principles, and expected benefits of ellipsoidal HEALPix within GRID4EARTH, and discusses its role as a practical and scalable DGGS for next-generation Earth system data infrastructures.

In the ESA project of DECSO-NRT-Europe, an operational system has been set up to derive daily NOx emissions in near-real time (NRT) for Europe based on observations of Sentinel 5P based on the offline inversion algorithm DECSO v6.5. This system has been developed to an NRT version for emissions on a spatial resolution of 0.1 x 0.1 degree (about 10x10 km). The inversion algorithm DECSO has been developed at KNMI for the purpose of deriving emissions for short-lived gases. It is using a Kalman Filter implementation for assimilating satellite column observations, optimising the emissions. DECSO is built on top of the regional chemistry-transport model (CTM) CHIMERE, which converts the analysed emissions into 3D concentration fields. The emission forecast model is based on persistency, predicting that the emissions remain constant since the previous analysis. This has the important advantage that the derived NOx emissions do not depend on a-priori (bottom-up) information on the expected locations and source strengths. Only a single CTM forward run is needed which makes the system computationally efficient. In the operational system of DECSO-NRT, each day the European NOX emissions are automatically derived and visualized, while on the third of each month the monthly emissions are calculated for the previous month and available on the data portal of DECSO-NRT. An archive of NOx emissions is available from the year 2019 till last month, consisting of NOx emissions for the anthropogenic, agricultural soil and forest soil emission sectors. When Sentinel 4 observations of NO2 become available, the service will be extended with hourly NRT NOx emissions during daytime.

Ammonia (NH3), nitrogen oxides (NOx) and sulfur dioxide (SO2) are key precursors of secondary inorganic aerosols and strongly influence air quality, nitrogen deposition and ecosystem health. Yet bottom-up emission inventories and the temporal profiles for these species remain highly uncertain and often inconsistent across regions and sectors. Here we present a global dataset of anthropogenic NOx, SO2 and NH3 emissions over land, providing estimates of the daily 10-day mean emission fields since 2019 at 1.27° × 2.5° resolution. Emissions are derived from an atmospheric transport and chemistry inverse modelling system based on the global chemistry transport model LMDZ-INCA and a finite-difference mass-balance (FDMB) inversion approach. We account for satellite retrieval operators by consistently applying averaging kernels to the modeled NO2, SO2 and NH3 fields prior to model–observation comparison and emissions inversion. For NOx and SO2, we assimilate tropospheric NO2 columns from TROPOMI, with OMI-based NOx inversions used for consistency checks. For NH3, we use IASI total columns. The TROPOMI- and OMI-based NOx inversions show similar large-scale spatial patterns but differ regionally in magnitude, and generally indicate higher NOx emissions than bottom-up inventories over major source regions such as China and India. The TROPOMI-based SO2 inversions suggest lower anthropogenic SO2 emissions than bottom-up inventories at the global scale and across most major source regions, with global totals remaining relatively stable over 2019–2023. For NH3, the IASI-based inversion reveals persistent hotspots over South and East Asia—especially India and China—where inferred emissions exceed estimates from inventories, with pronounced seasonal peaks in high-emitting regions. Our dataset provides retrieval-consistent, time-resolved constraints on major aerosol precursors and implies systematic discrepancies between bottom-up inventories and satellite-constrained emissions over major source regions. The presentation details the methodological choices ensuring the routine estimates of such global maps of emissions, the relevance of their relatively high resolution, and investigations for a joint inversion of the three species to strengthen the consistency of the overall dataset.

The Global Land Evaporation Amsterdam Model (GLEAM) estimates daily land evaporation using a wide range Earth observation forcing datasets. In the project GLEAM-HR funded by the European Space Agency (ESA), we aim to create a global high-resolution daily evaporation dataset at 1 km for a period of eight years (2016–2023). To produce high-resolution evaporation estimates, all forcing data must be processed at 1 km resolution, requiring substantial computational resources. As the complete high-resolution forcing data no longer fits within the memory capacity of single HPC nodes, parallelization tools are necessary. To achieve this parallelization in a seamless way, a workflow orchestration ecosystem is designed that leverages the use of Zarr, Apptainer and Nextflow.The Zarr ecosystem allows for easily writing to a dataset in parallel. Nextflow is an orchestration tool that allows dynamic job submissions, where the configuration of jobs can depend on the outcome of earlier jobs, such as the spatial domain to be processed. Apptainer is a containerization tool developed for HPC environments, allowing a “build once, deploy anywhere” approach. Combining these tools allows building a workflow orchestration environment that enables the automation of these parallel workflows while optimizing the job sizes for a given HPC environment.The use of containers allows this workflow to be ported to different hardware without the need to set up all the environments again, making the designed workflow fully reproducible independent of the computing environment. Combining this with Continuous Integration and Continuous Delivery (CI/CD) practices to automate the container building and deployment, code development and workflow execution can be cleanly separated.In a first test case, this processing workflow is used to produce global datasets of LAI, FPAR and vegetation cover fractions at 1 km resolution. Future work focuses on the extension of this workflow to the other forcing datasets and the entire pipeline execution.

This study presents daily, long-term (2003−2022) satellite-based estimates of river discharge and gridded runoff at 0.25° × 0.25° spatial resolution across the continental Pan-Arctic region. Satellite observations of precipitation from the Integrated Multi-satellite Retrievals (IMERG) for Global Precipitation Measurement (GPM), soil moisture and snow cover fraction from the European Space Agency's Climate Change Initiative (ESA CCI), and Terrestrial Water Storage Anomalies from satellite gravimetry are integrated into a conceptual hydrological model, STREAM, specifically adapted to simulate the hydrological regime of Pan-Arctic regions. The model is calibrated and validated using data from the 15 largest and best-monitored Arctic basins. It demonstrates high performance in reproducing daily river discharge observations, with a median Kling-Gupta Efficiency (KGE) of 0.83. To extend the model to ungauged regions, a regionalization approach based on similarities in basin climatic attributes is proposed, enabling the transfer of calibrated STREAM model parameters. When validated over 26 independent gauging stations, the regionalized river discharge estimates maintained acceptable accuracy, with a median KGE of 0.51. Using this framework, STREAM is applied across the continental Pan-Arctic region covering an area of 16.5 × 10 6 km 2 . The resulting spatially and temporally consistent runoff dataset, which can be used to complement more detailed process-based model estimates, enables the quantification of freshwater fluxes to the Arctic Ocean, estimated at 4760 ± 619 km 3 yr −1 . The results highlight the potential of satellite observations—particularly GRACE data—in supporting large-scale hydrological modelling and in reconstructing the Arctic freshwater cycle in data-scarce environments. • A satellite-based hydrological model (STREAM) is applied to the Pan-Arctic region. • STREAM performs well in 15 Arctic basins with a median KGE of 0.83. • Regionalization enables STREAM application in ungauged Arctic basins. • Development of a satellite-based daily runoff dataset for the Pan-Arctic region.