Search

Machine learning algorithms demonstrated promising results for hyperspectral semantic segmentation. However, they strongly rely on the quality of training datasets. As far as the annotation of hyperspectral images is often expensive and time-consuming, only a few thousand pixels can be labeled. In this context, active learning algorithms select the most informative pixels to be labeled. In the machine learning community, recent active learning methods have overcome the performance of conventional algorithms but do not always scale to large remote sensing images. Therefore, we introduce in this paper a preprocessing method that allows the use of computationally intensive active learning algorithms without significant impacts on their effectiveness.

International audience; The Soil Moisture and Ocean Salinity (SMOS) satellite has been providing global data for more than 11 years. The algorithm to retrieve surface soil moisture (SM) and vegetation optical depth (VOD) from Brightness Temperatures (TB) has constantly evolved as our understanding of the instrument and of the responses to various ecosystems increased. The evolution of the radiative transfer modelling used in the inversion process is in part based upon ground measurements done on experimental sites. We present the particular case of an L-band radiometer placed on top of a 800- meter-high cliff. This design allows us to study large and various scenes when compared to classical experiments where a radiometer is placed 10–20 m above a surface. With our set up, the footprints cover areas larger than 300 m in diameter, covering thus several land use types and the footprint can be directed towards different dominant land classes (forest, crops, urban etc…). This paper presents the whole set up consisting of in situ measurements (soil moisture and surface temperatures) and L-band radiometry acquired by the LEWIS instrument at high incidence angles. We perform SM/VOD retrievals from brightness temperatures using various configurations which are compared to the in situ observations. We test in particular a 2-VOD retrieval approach that consists in deriving one SM for the entire scene but two VOD, one for agricultural/low vegetation and a VOD for the forest parts, which improves the performance of the SM/VOD retrievals.

Black carbon aerosol emissions are recognized as contributors to global warming and air pollution. There remains, however, a lack of in-situ techniques to remotely quantify black carbon aerosol particles with high range and time resolution. This article presents for the first time, to our knowledge, a direct and contact-free remote measurement of black carbon aerosol number and mass concentration less than ten of meters from the emission source. This is done with a novel picosecond short-range elastic backscatter lidar (PSR-EBL) technique. To address the complexity of retrieving lidar products at short measurement ranges, we apply a forward inversion method featuring radiometric lidar calibration. Our method is based on an extension of a well-established light-scattering model, the Rayleigh-Debye-Gans for Fractal-Aggregates (RDG-FA) theory, which computes an analytical expression for lidar parameters. These parameters are the backscattering cross-sections and the lidar ratio for black carbon fractal aggregates. Using a small-scale Jet A-1 kerosene pool fire, it is shown that our technique can quantify the aerosol number and mass concentration with centimetre range-resolution and millisecond time-resolution.; Les émissions d'aérosols de carbone noir sont reconnues comme contribuant au réchauffement climatique et à la pollution de l'air. Il reste cependant un manque de techniques pour quantifier à distance les particules d'aérosol de noir de carbone avec une plage et une résolution temporelle élevées. Cet article présente une technique à distance directe et sans contact pour quantifier le nombre d'aérosols de carbone noir et la concentration massique à quelques mètres de la source d'émission. Cela se fait à l'aide de l'instrument Colibri basé sur une nouvelle technique, appelée ici Picosecond Short-Range Elastic Backscatter Lidar (PSR-EBL). Pour faire face à la complexité de la récupération des produits lidar à de courtes plages de mesure, nous appliquons une méthode d'inversion directe avec étalonnage radiométrique du lidar. Notre méthode est basée sur une extension d'un modèle de diffusion de la lumière bien établi, la théorie Rayleigh-Debye-Gans for Fractal-Aggregates (RDG-FA), qui calcule une expression analytique des paramètres lidar. Ces paramètres sont les sections efficaces de rétrodiffusion et le rapport lidar pour les agrégats fractals de carbone noir. À l'aide d'un feu de nappe de kérosène Jet A-1 à petite échelle, nous démontrons la capacité de la technique à quantifier le nombre d'aérosols et la concentration en masse avec une résolution centimétrique et une résolution temporelle en millisecondes.

International audience; As the volume of marine transportation increases, it becomes increasingly important to monitor ships for efficient coastal monitoring and management. To this end, high-resolution satellite images can be utilized to surveil oceanic environments synoptically. In this study, high-resolution optical satellite image was used to detect ships and estimate the size of each ship in the Korean coastal region. All the pixels in an image were first classified into ship, ship shadow, wake, sea, and land by applying a maximum likelihood classifier. The positions corresponding to the boundary of the ship were obtained from the magnitude of the 2-dimensional gradient on the classified ship pixels, and then the length and width of the ship were estimated by applying an ellipse fitting method to the ship periphery. This method resulted, in slight overestimations of the sizes of the ships. In order to improve the accuracy of the estimated ship sizes, a correction formula was developed by investigating the errors of the estimated values and their potential relationships to the variables representing the spatial shape of the vessels, such as eccentricity, kurtosis. Applying the suggested formulation for ship size estimation improved accuracy by 54.41% compared to the estimated sizes obtained through ellipse fitting. We anticipate that our method of estimating the lengths of the vessels will contribute to identifying missing ships using high-resolution satellite images.

International audience; An improved version of the MARMIT model is proposed. Wet soil is represented as a dry soil layer surmounted by a layer of a mixture of water and soil particles.

International audience; The detection and the quantification of greenhouse gases is essential to climate change studies, avoid leakages in industrial site, prevent accidental explosions. Because of these properties, methane (CH 4) is an important target gas in remote sensing quantification. We focused on industrial plume quantification with data obtained during airborne campaign with HySpex-Neo hyperspectral camera. The 1.4 m of spatial resolution allows comparing quantification methods on real data combined with full or semi-synthetic plume case. A linear method largely used in the literature is compared with a quantification method based on the rebuilt background image and the estimation of plume transmission (PTE method). We have developed a hybrid approach using intermediate results of two previous methods. The hybrid method is based on the optimal estimation (OE) formalism and is providing uncertainty estimates. We show that the linear method underestimates the concentration of plumes for concentration above 5000 ppm.m. For low reflectance pixels, the hybrid method is more robust than the PTE method. The uncertainty of the hybrid method is about 30% for pixels with concentrations above 5000 ppm.m. For a HySpex-Neo image, the total mass of the plume is underestimated by 30% with the linear method compared to the hybrid method.; La détection et la quantification des gaz à effet de serre sont essentielles pour étudier le changement climatique, éviter les fuites sur les sites industriels et prévenir les explosions accidentelles. En raison de ces propriétés, le méthane (CH4) est un gaz cible important pour la quantification par télédétection. Nous nous sommes concentrés sur la quantification des panaches industriels avec les données obtenues lors de la campagne aéroportée avec la caméra hyperspectrale HySpex-Neo. La résolution spatiale de 1,4 m permet de comparer les méthodes de quantification sur des données réelles combinées à un cas de panache complet ou semi-synthétique. Une méthode linéaire largement utilisée dans la littérature est comparée à une méthode de quantification basée sur la reconstruction de l'image de fond et l'estimation de la transmission du panache (méthode PTE). Une méthode hybride, utilisant les résultats intermédiaires des deux méthodes précédentes, est basée sur une estimation optimale et fournit une étude de la propagation des erreurs. Nous montrons que la méthode linéaire sous-estime la concentration des panaches pour une concentration supérieure à 15 000 ppm.m. Les méthodes PTE et hybride retrouvent une concentration en bon accord avec la concentration synthétique. Pour les pixels de faible réflectance, la méthode hybride est plus robuste que la méthode PTE.

International audience; In the context of climate change and rising frequency of extreme hydro-meteorological events around the world, flood risk management and mapping of heavy rainfall-related damages represent an ongoing critical challenge. For decades now, remote sensing has been largely used to investigate spatial and temporal changes in land use and water resources. Today, different satellite products provide fast and crucial knowledge for the study of hydrological disasters over large areas, possibly in remote regions, with high spatial resolution and high revisit frequency. Yet, until now, few works have sought to detect the full range of extreme rainfall-related damages with optical imagery, especially those caused by intense rainwater runoff beyond the direct vicinity of major waterways. The work presented in this paper focuses on the Aude severe weather event of October 15th, 2018, in the South of France, for which more than a thousand claims for agricultural disaster were registered, both related to river overflowing and rainwater runoff. The full resources of ground truths, contextual information, land use as well as digital elevation model (DEM) combined to high resolution and high frequency optical imagery (Sentinel-2, Pléiades) are used to develop an automatic damage detection method based on supervised classification algorithms. Through the combination of several indicators characterizing heterogeneous spectral variations among agricultural plots following the event, a Gaussian process classifier achieved various classification accuracies up to 90% on a large comparable and independent photo-interpreted validation sample. This work builds great expectations for applications in other areas with contrasted climate, topography and land cover.; Dans le contexte du changement climatique et de la fréquence croissante des événements hydrométéorologiques extrêmes dans le monde, la gestion des risques d'inondation et la cartographie des dommages liés aux fortes pluies représentent un défi critique permanent. Depuis des décennies, la télédétection a été largement utilisée pour étudier les changements spatiaux et temporels de l'utilisation des terres et des ressources en eau. Aujourd'hui, différents produits satellitaires fournissent des connaissances rapides et cruciales pour l'étude des catastrophes hydrologiques sur de vastes zones, éventuellement dans des régions éloignées, avec une résolution spatiale élevée et une fréquence de revisite élevée. Pourtant, jusqu'à présent, peu de travaux ont cherché à détecter la gamme complète des dommages liés aux pluies extrêmes par imagerie optique, en particulier ceux causés par un ruissellement intense des eaux de pluie au-delà du voisinage direct des principaux cours d'eau. Les travaux présentés dans cet article se concentrent sur l'épisode de temps violent de l'Aude du 15 octobre 2018, dans le sud de la France, pour lequel plus d'un millier de sinistres agricoles ont été enregistrés, tous deux liés au débordement des rivières et au ruissellement des eaux de pluie.L'ensemble des ressources de vérités de terrain, d'informations contextuelles, d'occupation des sols ainsi que du modèle numérique d'élévation (DEM) combiné à l'imagerie optique haute résolution et haute fréquence (Sentinel-2, Pléiades) sont utilisés pour développer une méthode de détection automatique des dommages basée sur une classification supervisée. algorithmes. Grâce à la combinaison de plusieurs indicateurs caractérisant les variations spectrales hétérogènes entre les parcelles agricoles à la suite de l'événement, un classificateur de processus gaussien a obtenu diverses précisions de classification allant jusqu'à 90% sur un grand échantillon de validation photo-interprété comparable et indépendant. Ce travail suscite de grandes attentes pour des applications dans d'autres domaines au climat, à la topographie et à la couverture du sol contrasté.

Characterization and seasonal (periodic) monitoring of plant species distribution in the context of former industrial activity is crucial to assess long-term anthropogenic footprint on vegetated area. Species discrimination has shown promising results using both HyperSpectral (HS) and MultiSpectral (MS) images. Airborne HS instruments enable high spatial and spectral resolution imagery while time series of satellite MS images provide high temporal resolution and phenological information. This paper aims to compare supervised classification results obtained with non-parametric (Random Forest, RF, Support Vector Machine, SVM) and parametric methods (Regularized Logistic Regression, RLR) applied on both kinds of images acquired on an industrial brownfield. The studied site is a complex vegetated environment due to species diversity: 8 dominant species are retained. The performance obtained by preliminary feature selection based on principal component analysis and vegetation indices, to improve separability of spectral or temporal information according to species, is analysed. The best performance is obtained by RLR method applied on HS data without feature selection (global accuracy of 93 %). Feature selection is found to be a necessary step to perform classification with time series of MS images. Species that are difficult to distinguish from the HS image, namely Salix and Populus, are well separated using Sentinel-2 images (precision around 70%).

International audience; Cryogenic cameras are an innovative alternative in the design of miniaturized infrared cameras using cryogenic detectors. In this presentation, we will apply this technology to design a snapshot multispectral camera for gas leak detection. A UAV compatible demonstrator in a commercial Detector Dewar Cooler Assembly (DDCA), called SIMAGAZ, has been made and tested in the TOTAL Anomaly Detection Initiatives (TADI) platform and the Esperse site of ONERA. The TADI infrastructure manages monitored gas leaks at flowrates from 0.1 g/s to 300 g/s and hosts remote sensors to test them in three scenarios: crisis-management, safety monitoring, and environmental monitoring. In Esperse site, first UAV flights with SIMAGAZ were performed. We demonstrate the ability to detect and quantify in real time the origin of methane gas leak, the flowrate and the volume of the plume with SIMAGAZ on ground or from a UAV. The core camera weights around 1kg, for around 1L footprint and a power consumption of 10W at the cooling steady state. Results from TADI and Esperse campaigns will be presented.

International audience; Elastic backscatter lidar is an established method useful to characterize the particles forming an atmospheric aerosol. Fundamental to lidar measurements are the extinction and backscattering properties of the aerosol particles in the lidar beam. This review aims to give the reader an understanding for how lidar measurements are made, modeled, and applied. Specific emphasis is placed on extinction, which manifests attenuation of the lidar beam by the aerosol, and the backscattering that ultimately constitutes the measured signal. The so-called lidar equation, which quantifies this signal in terms of aerosol-particle properties, is derived from radiative transfer theory. Extinction is examined in terms of the redistribution of energy from the lidar beam by the particles and is shown to be inherently linked to interference between the incident and scattered light. Ways to measure aerosol-particle extinction are reviewed, including a recent method where the wave interference constitutes a digital hologram that may be processed to yield both the cross section and an image of the particle. The semi-graphical method of phasor analysis is also presented and used to reveal connections between the scattering characteristics of a particle and the distribution of electric field within it. Finally, a recently developed elastic backscatter lidar system, Colibri, is described and an example of its measurement capabilities are presented.; Le lidar à rétrodiffusion élastique est une méthode établie utile pour caractériser les particules formant un aérosol atmosphérique et a de nombreuses applications en télédétection. Les propriétés d'extinction et de rétrodiffusion des particules d'aérosol dans un faisceau lidar sont fondamentales pour les mesures lidar élastique. Cet article de revue vise à donner au lecteur une compréhension de la façon dont les mesures lidar sont faites, modélisées et appliquées à la caractérisation des aérosols. Un accent particulier est mis sur l'extinction, qui manifeste l'atténuation du faisceau lidar par l'aérosol, et la rétrodiffusion qui constitue finalement le signal mesuré. L'équation dite lidar, qui quantifie ce signal en termes de propriétés des particules d'aérosol, est dérivée de la théorie du transfert radiatif. L'extinction est décrite en termes de conservation de l'énergie, car une particule absorbe et diffuse la lumière. Il est démontré que la redistribution de l'énergie du faisceau lidar par les particules est intrinsèquement liée à l'interférence entre la lumière incidente sur les particules et celle diffusée par celles-ci. Les méthodes de mesure de l'extinction des particules d'aérosol sont passées en revue, y compris une méthode récente où l'interférence des ondes impliquées dans l'extinction constitue un hologramme qui peut être traité pour donner à la fois la section transversale et une image d'une particule. La méthode semi-graphique d'analyse de phase est présentée et utilisée pour révéler le lien entre la diffusion d'une particule et la distribution du champ électrique en son sein. Enfin, un nouveau système de lidar élastique de rétrodiffusion, Colibri, est décrit dans le contexte des sujets abordés et un exemple de ses capacités de mesure est présenté.

Jose Antonio Benavent-Oltra is funded by the University of Granada through "Plan Propio. Programa 7, Convocatoria 2019". Roberto Roman is funded by MINECO under the postdoctoral programme Juan de la Cierva-Incorporacion (IJCI2016-30007). Juan Andres Casquero-Vera is funded by MINECO under the predoctoral programme FPI (BES-2017-080015). Maria J. Granados-Munoz received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 796539. Oleg Dubovik was supported by the Labex CaPPA project, which is funded by the French National Research Agency under contract "ANR-11-LABX0005-01". This work was also supported by the Spanish Ministry of Economy and Competitiveness (projects CMT2015-66742-R, CGL2016-81092-R, CGL2017-85344-R, RTI2018-097864-B-I00 and CGL2017-90884-REDT), by the Andalusia Regional Government through project P18-RT-3820 and the Unity of Excellence "Maria de Maeztu" (project MDM-2016-0600) financed by the Spanish State Research Agency (AEI). The authors thankfully ac-knowledge the FEDER programme for the instrumentation used in this work, the University of Granada, which supported this study through the Excellence Units Program, and the Sierra Nevada National Park. We also thank Dr. Grisa Monick who provided the Aethalometer AVIO AE-33 installed on the aircraft. Thanks to AERONET and ACTRIS/AERONET Europe for the scientific and technical support. Finally, the authors would like to acknowledge the use of the GRASP inversion algorithm software (http://www.grasp-open.com, last access: 1 April 2020), in this work., The Sierra Nevada Lidar aerOsol Profiling Experiment I and II (SLOPE I and II) campaigns were intended to determine the vertical structure of aerosols by remote sensing instruments and test the various retrieval schemes for obtaining aerosol microphysical and optical properties with in situ measurements. The SLOPE I and II campaigns were developed during the summers of 2016 and 2017, respectively, combining active and passive remote sensing with in situ measurements at stations belonging to the AGORA observatory (Andalusian Global ObseRvatory of the Atmosphere) in the Granada area (Spain). In this work, we use the in situ measurements of these campaigns to evaluate aerosol properties retrieved by the GRASP code (Generalized Retrieval of Atmosphere and Surface Properties) combining lidar and sun-sky photometer measurements. We show an overview of aerosol properties retrieved by GRASP during the SLOPE I and II campaigns. In addition, we evaluate the GRASP retrievals of total aerosol volume concentration (discerning between fine and coarse modes), extinction and scattering coefficients, and for the first time we present an evaluation of the absorption coefficient. The statistical analysis of aerosol optical and microphysical properties, both column-integrated and vertically resolved, from May to July 2016 and 2017 shows a large variability in aerosol load and types. The results show a strong predominance of desert dust particles due to North African intrusions. The vertically resolved analysis denotes a decay of the atmospheric aerosols with an altitude up to 5 km a.s.l. Finally, desert dust and biomass burning events were chosen to show the high potential of GRASP to retrieve vertical profiles of aerosol properties (e.g. absorption coefficient and single scattering albedo) for different aerosol types. The aerosol properties retrieved by GRASP show good agreement with simultaneous in situ measurements (nephelometer, aethalometer, scanning mobility particle sizer, and aerodynamic particle sizer) performed at the Sierra Nevada Station (SNS) in Granada. In general, GRASP overestimates the in situ data at the SNS with a mean difference lower than 6 mu m(3) cm(-3) for volume concentration, and 11 and 2 Mm(-1)for the scattering and absorption coefficients. On the other hand, the comparison of GRASP with airborne measurements also shows an overestimation with mean absolute differences of 14 +/- 10 and 1.2 +/- 1.2 Mm(-1) for the scattering and absorption coefficients, showing a better agreement for the absorption (scattering) coefficient with higher (lower) aerosol optical depth. The potential of GRASP shown in this study will contribute to enhancing the representativeness of the aerosol vertical distribution and provide information for satellite and global model evaluation., University of Granada, MINECO under the postdoctoral programme Juan de la Cierva-Incorporacion IJCI2016-30007, MINECO under the predoctoral programme FPI BES-2017-080015, European Commission 796539, French National Research Agency (ANR) ANR-11-LABX0005-01, Spanish Ministry of Economy and Competitiveness CMT2015-66742-R CGL2016-81092-R CGL2017-85344-R RTI2018-097864-B-I00 CGL2017-90884-REDT, Andalusia Regional Government P18-RT-3820, Spanish State Research Agency (AEI) MDM-2016-0600, Excellence Units Program, Sierra Nevada National Park

International audience; Combination of numerous satellite data has lately become available to cover large areas with very high spatial resolution (VHR) and high revisit frequency. Little studies have yet made use of these images to assess and map damages following an extreme rainfall event, in particular those caused by rainwater runoff. We therefore investigate a specific approach to detect these damages as exhaustively as possible from VHR and HR satellite data acquired as closely as possible before and after an intense rainfall event. To do so, we used Pléiades (0.7 m) and Sentinel-2 (10 m) images taken over the Aude region (France) which was heavily affected by a severe storm on October 15th, 2018. We chose to focus on agricultural lands as 1119 claims for agricultural disaster were registered and certified following this event. Post-event VHR Pléiades images were used to identify claimed damages over a subset of agricultural plots as well as to determine contextual information such as the type of damage (erosion, deposit, uprooting). Several indices and spectral filters were then applied to the selected areas from pre and post-event Sentinel-2 images. This exploratory work reveals that certain types of agricultural damage are well detected while others, albeit clearly visible on Pleiades images, are harder to identify with the selected indices and filters on Sentinel-2 images. It also shows the potential of this approach to discriminate the extent of damage that was declared over agricultural areas. This study emphasizes how relevant the use of combined spectral, temporal and contextual information is to detect damages following an extreme rainfall event, in particular those caused by rainwater runoff, thanks to optical imagery, as spectral knowledge alone does not appear to be sufficient. This preliminary work paves the way for further work based on the development of more advanced change detection methods, clues and filters as well as artificial intelligence.; De nombreuses données satellites peuvent aujourd'hui être combinées afin de couvrir des surfaces très importantes avec une très haute résolution spatiale (THR) ainsi qu'une haute fréquence de revisite. Le potentiel de ces images pour évaluer et cartographier les dommages engendrés par des pluies extrêmes, en particulier ceux causés par le ruissellement pluvial, a été jusqu'à présent peu étudié. Cette étude propose une méthode pour détecter de la manière la plus exhaustive possible ces dommages à partir de données satellitaires THR et HR acquises au plus près, avant et après, d'un événement de pluie intense. Pour ce faire, nous avons utilisé des images Pléiades (0,7 m) et Sentinel-2 (10 m) acquises au-dessus de la région de l'Aude (France), fortement touchée par des intempéries le 15 octobre 2018. Notre intérêt a porté sur les zones agricoles qui ont fait l'objet de 1119 demandes d'indemnisation en calamités agricoles pour cet événement. Plusieurs indices et filtres spectraux ont été appliqués sur un échantillon d'images Sentinel-2 sélectionnées avant et après l'épisode orageux. Ce travail exploratoire révèle que certains types de dommages agricoles sont bien détectés alors que d'autres, même clairement visibles sur les images Pléiades, sont plus difficiles à distinguer avec les indices et filtres sélectionnés sur les images Sentinel-2. Il démontre également le potentiel de ces méthodes pour discriminer les différents degrés de dégâts relevés sur les parcelles agricoles. Cette étude confirme l'importance de combiner information spectrale, temporelle et contextuelle pour détecter à l'aide de l'imagerie optique les dommages engendrés par des pluies extrêmes, en particulier ceux causés par le ruissellement pluvial. Ces travaux préliminaires ouvrent la voie au développement de nouvelles méthodes de détection, l'utilisation de nouveaux indices ainsi que sur l'intelligence artificielle.

International audience; Hazardous Noxious Substance (HNS) is transported entirely through large vessels, so there is always a potential risk of marine HNS spills. In the event of an HNS accident, it can cause enormous human and property damage, so prompt detection is required. However, there is a limit to human access by ship, we need to use remote sensing data. In this study, ground experiments using hyperspectral cameras were performed to construct a spectral library of HNS. We classified the HNS and non-HNS by applying the hyperspectral mixture algorithm, and presented the HNS detection probability for every pixel by calculating the spectrum-based abundance fraction. The results of this study are expected to be used to estimate the extent of HNS spill in the event of a marine HNS accident.

International audience; Gas leaks are a major issue for Oil & Gas companies, either for environment or safety purposes. Remote sensing technologies can be applied for a wide range of gas leak flowrates and in three main cases: major leaks in crisis management; medium size leaks in safety monitoring; small leaks in environmental monitoring. In 2019, TOTAL completed a five-year Research & Development collaborative project NAOMI with ONERA, The French Aerospace Lab, to develop technologies for remote detection, identification, visualization and quantification of gas leaks with applications relating to safety and environmental protection. TOTAL organized several gas test campaigns on Lacq Pilot Platform in France, called TADI (Total Anomaly Detection Initiatives) to evaluate and select relevant technologies.

International audience; Earth observation at a local scale requires images having both high spatial and spectral resolutions. As sensors cannot simultaneously provide such characteristics, a solution is combining images jointly acquired by two different optical instruments. Notably, hyperspectral pansharpening methods combine a panchromatic image, providing a high spatial resolution, with a hyperspectral image, providing a high spectral resolution, to generate an image with both high spatial and spectral resolutions. Nevertheless, these methods suffer from some limitations, including managing mixed pixels. This article introduces a new hyperspectral pansharpening method designed to deal with mixed pixels, which is called Spatially Organized Spectral Unmixing (SOSU). The performance of this method is measured on synthetic then real data (simulated from airborne acquisitions), using spatial, spectral and global criteria, to evaluate the contributions of the SOSU algorithm to mixed pixel processing. In particular, this contribution is confirmed in the case of a peri-urban area via a nearly ten percent increase in the rate of mixed pixels improved with SOSU, in comparison with the method used as a reference.; L'observation de la Terre à l'échelle locale nécessite de disposer d'images hautement résolues spatialement et spectralement. Les capteurs ne pouvant offrir simultanément de telles résolutions, une solution consiste à exploiter des images acquises par deux instruments optiques différents. Les méthodes de pansharpening hyperspectral, notamment, permettent de combiner une image panchromatique, à haute résolution spatiale, avec une image hyperspectrale, à haute résolution spectrale, afin de générer une nouvelle image hautement résolue spatialement et spectralement. De telles méthodes présentent cependant certaines limitations, en particulier le traitement des pixels mixtes. Cet article présente une nouvelle méthode de pansharpening hyperspectral appelée Spatially Organized Spectral Unmixing (SOSU) et adaptée au traitement de tels pixels. Les performances de cette méthode sont mesurées sur des données synthétiques puis réelles (simulées à partir d'acquisitions aéroportées), à l'aide de critères spatiaux, spectraux et globaux, afin d'évaluer l'apport du traitement des pixels mixtes. En particulier, cet apport est confirmé dans le cas d'un paysage périurbain par une augmentation de près de dix pourcents du taux de pixels mixtes améliorés avec SOSU, par rapport à la méthode utilisée comme référence.

International audience; The control and monitoring of greenhouse gases is an important issue for the study of climate change, for the impact in terms of public health or for the risks related to industry. An algorithm has been developped dedicated to commercial airborne hyperspectral camera as HySpex-NEO for the detection of industrial methane plume and the quantification of the emission source. HySpex-NEO is an imager used in airborne campaign with a spatial resolution of 1.4 m at flight altitude of 2 km, a swath of 650 m and a spectral resolution of 6 nm. This algorithm has been validated over a controlled release less than 100 g/s during an airborne campaign over Lacq (France) industry. It has also been applied to the Aliso Canyon leakage data acquired with AVIRIS JPL (Airborne Visible Infrared Spectrometer) with a spatial resolution of 6.8 m at flight altitude of 6 km, a swath of 5.6 km and a spectral resolution of 10 nm. Application to satellite hyperspectral data is shown on artificial data derived from airborne hyperspectral acquisitions.; Le contrôle et la surveillance des gaz à effet de serre est une question importante pour l'étude du changement climatique, pour l'impact en termes de santé publique ou pour les risques liés à l'industrie. Un algorithme a été développé pour les caméras hyperspectrales aériennes commerciales comme HySpex-NEO pour la détection du panache de méthane industriel et la quantification de la source d'émission. HySpex-NEO est un imageur utilisé dans les campagnes aéroportées avec une résolution spatiale de 1,4 m à une altitude de vol de 2 km, une fauchée de 650 m et une résolution spectrale de 6 nm. Cet algorithme a été validé sur un rejet contrôlé inférieur à 100 g/s lors d'une campagne aérienne au-dessus de l'industrie de Lacq (France). Il a également été appliqué aux données de fuite de l'Aliso Canyon acquises avec l'AVIRIS JPL (Airborne Visible Infrared Spectrometer) avec une résolution spatiale de 6,8 m à une altitude de vol de 6 km, une fauchée de 5,6 km et une résolution spectrale de 10 nm. L'application aux données hyperspectrales des satellites est présentée sur des données artificielles dérivées d'acquisitions hyperspectrales aéroportées.

In this paper, we focus on the retrieval of microphysical and optical properties of industrial aerosol plumes through a process using airborne hyperspectral and Sentinel-2 multi-spectral images. The process allows first to perform atmospheric correction and second to determine background aerosols thanks to a comparison between hyperspectral and Sentinel-2 reflectances. Hyperspectral methodologies use the radiance differential between the measurement in the plume and the corresponding measurements out of the plume to estimate plume properties. To retrieve the surface reflectance under the plume, a principal component analysis coupling hyperspectral and multispectral data class by class is achieved. The developed method aims to compare the difference between measured and estimated reflectance with a radiative transfer model accounting for plume properties (particle radius and aerosol optical thickness of the plume). We have applied the method to a steel plant in the south of France. The retrieved plume show an aerosol mean radius between 0.05 and 0.2 µm with a mean aerosol optical thickness about 0.05 along the plume.

Monitoring of coastal areas by remote sensing is an important issue. The interest of using an unmixing method to determine the seabed composition from hyperspectral aerial images of coastal areas is investigated. Unmixing provides both seabed abundances and endmember reflectances. A sub-surface mixing model is presented, based on a recently proposed oceanic radiative transfer model that accounts for seabed adjacency effects in the water column. Two original non-negative matrix factorization ( N M F )-based unmixing algorithms, referred to as W A D J U M (Water ADJacency UnMixing) and W U M (Water UnMixing, no adjacency effects) are developed, assuming as known the water column bio-optical properties. Simulations show that W A D J U M algorithm achieves performance close to that of the N M F -based unmixing of the seabed without any water column, up to 10 m depth. W U M performance is lower and decreases with the depth. The robustness of the algorithms when using erroneous information about the water column bio-optical properties is evaluated. The results show that the abundance estimation is more reliable using W A D J U M approach. W A D J U M is applied to real data acquired along the French coast; the derived abundance maps of the benthic habitats are discussed and compared to the maps obtained using a fixed spectral library and a least-square ( L S ) estimation of the seabed mixing coefficients. The results show the relevance of the W A D J U M algorithm for the local analysis of the benthic habitats.

International audience; Surface soil moisture content (SMC) is known to impact soil reflectance at all wavelengths of the solar spectrum. As a consequence, many semi-empirical methods aim at inferring SMC from soil reflectance, but very few rely on physically-based models. This article presents a multilayer radiative transfer model of soil reflectance called MARMIT (multilayer radiative transfer model of soil reflectance) as a function of SMC given on a mass basis and a method called MARMITforSMC to estimate it from soil reflectance spectra. This model depicts a wet soil as a dry soil covered with a thin film of water. It is used to assess SMC over seven independent laboratory datasets gathered from the literature. A learning phase is required to link the thickness of the water film with the SMC. For that purpose, a sigmoid function, the parameters of which are related to soil physical and chemical properties such as porosity, grain size and mineralogy composition, is fitted. SMC can be inferred with good accuracy (RMSE ≈ 3%) if the learning step is applied soil by soil. The link between SMC and water thickness actually depends on soil texture and chemical composition. If the soils are divided into classes and if the learning phase is applied to a class, the RMSE slightly increases up to 5%. Finally, MARMITforSMC provides lower RMSE than any other existing semi-empirical or physically-based method.

International audience; For remote-sensing applications such as spectra classification or identification, atmospheric correction constitutes a very important pre-processing step, especially in complex urban environments where a lot of phenomenons alter the shape of the signal. The objective of this article is to compare the efficiency of two atmo-spheric correction algorithms, COCHISE (atmospheric COrrection Code for Hyperspectral Images of remote-sensing SEnsors) and an empirical method, on hyperspectral data and for classification applications. Classification is carried out on several simulated spaceborne data sets with different spatial resolutions (from 1.6 to 9.6 m). Four classifiers are considered in the study: a k-means, a Support Vector Machine (SVM), and a sun/shadow version of each of them, which processes sunlit and shadowed pixels separately. Results show that the most relevant atmospheric method for classification depends on the spatial resolution of the processed data set. Indeed, if the empirical method performs better on high-resolution data sets (up to 4%), its superiority fades out as the spatial resolution decreases, especially with the lower spatial resolution where COCHISE can be 10% more accurate than the empirical method.

International audience; This brief review focuses on the use of optical remote sensing for oil prospecting and monitoring in vegetated regions. Major advances have been achieved in this field during the last decade. Recent studies demonstrated that oil can be detected and quantified by exploiting vegetation reflectance from airborne-and satellite-embedded sensors. Various methods coupling vegetation indices or radiative transfer models with classification or regression algorithms have been proposed for this purpose. High spatial and spectral resolutions greatly increase the accuracy of oil detection and quantification. However, important limits remain in regions with sparse or highly diversified vegetation. Oil spills become also more difficult to map when other environmental stressors occur. Hence, an important effort remains to move toward operational applications, especially on upcoming drone-and satellite-embedded hyperspectral sensors.

Remote sensing or remote diagnostic techniques meet many dual needs, such as the detection of threats (eg biocontaminants), the characterization of the optical properties of materials (eg paints) or the study of the environment (ex. problem of aerosols and effluents). The usual techniques such as remote sensing by LIDAR or the measurement of optical properties by scatterometer provide partial solutions to the problem of the optical diagnosis of diffusing media. These methods generally require an a priori knowledge of the environment. For example, hypotheses are formulated on the nature of particles (eg spherical form instead of arbitrary shapes, type and index ...), simple diffusion instead of multiple and dependent diffusion, and so on. The emergence of atmospheric LIDARs in the middle infrared range dates back several years. Several institutes and research laboratories, such as the National Institute of Optics (Quebec, Canada) or the Swedish Defense Research Agency (Sweden), have proposed LIDAR concepts in the infrared tunable OPO lasers for the purpose of probing the atmosphere for the detection of bioaerosols. These laser sources are advantageous because they provide high peak powers to probe the atmosphere on several kilometers of distance. However, new needs are emerging for in-situ and short-range measurements. These needs include adapting systems to make them more compact and compatible with embedded and easily deployable solutions in the fiel Over the past ten years, the use of hyperspectral lasers of the 'supercontinuum of white light' type has been anticipated to meet these new needs in the field of LIDAR. Several laboratories, such as the Finnish Geodetic Institute (Masala, Finland) or Air Force Research Lab (Ohio, USA), have been offering the first prototypes of hyperspectral LIDAR 'supercontinuum' for nearly 5 years. However, up to now, no solution makes it possible to trace all the microphysical properties of the particles simultaneously. For more than 10 years, LEUKOS has been developing supercontinuum laser sources. Long time oriented towards the generation of UV wavelengths, LEUKOS has recently begun to study the widening of the other side of the spectrum towards the infrared medium. These supercontinuum laser sources appear today as multi-application tools. LEUKOS wishes to focus on industrial products that integrate these new sources in close collaboration with French research centers in the fields of optics and the atmosphere. ONERA, through its work previously carried out in the Optics Department, positions itself as a proposal force in the fusion of spectral and polarimetric and directional data for the optical diagnosis of diffusing media. Research benches have been developed to measure hyperspectral and polarimetric optical signatures of materials and, more generally, of any scattering media. The LOA is particularly interested in the optical properties of diffusers such as aerosols and clouds to constrain radiative transfer models, interpret observations and better estimate the radiative radiative balance. One of the major interests of the proposed work lies in the determination of these properties over a wide spectral range. The measurement of aerosols (hydrometeors, marine salts, pollens, etc.) and the use of μLIDAR data from visible to thermal IR is innovative in the field of atmospheric sciences. The fusion of spectral, polarimetric and multi-directional information makes it possible to simultaneously find all the microphysical properties of interest of a diffusing medium such as size, complex optical index and particle concentration while reducing the number of hypotheses physical effects on the environment. This approach is, however, limited to particles of comparable size to the covered spectral domain. Access to diffusion measurements (eg backscatter coefficients, phase function) over a wide spectral range (typically between 0.4 μm and 4.7 μm) will make it possible to extract certain microphysical parameters without any simplifying hypothesis and to increase the size range accessible by this method (dust, marine aerosols, agglomerated soot, biocontaminants ...). Several methods are already envisaged (eg methods of differential diffusion, fusion of information ...). This approach makes it possible to improve the characterization of the atmosphere, which is an important need in the environment (study of pollution ...). ONERA, LOA and LEUKOS wish to join forces to extend this optical diagnostic technique which combines both the spectral aspect and the polarimetric aspect of light in the field of the middle infrared., Le lidar est un instrument de télédétection active similaire au radar opérant dans la partie optique du spectre électromagnétique. Le principe est d’illuminer un milieu diffusant, soit solide (ex. végétation, peinture) soit dilué (ex. atmosphère, panache de fumée), et d’en collecter la lumière rétrodiffusée. Dans notre étude, deux processus d’interaction lumière-matière mis en jeu au cours de la propagation d’une impulsion laser sont étudiés : la diffusion élastique et l’absorption de la lumière. Les méthodes de diagnostic par méthode lidar donnent des informations sur les paramètres optiques du milieu (coefficients de diffusion et d’absorption). Ces grandeurs dépendent des propriétés de la lumière (longueur d’onde, polarisation) et sont ensuite utilisées pour remonter aux propriétés microscopiques du milieu (i.e. distribution en taille, morphologie, indice optique, concentrations). Les techniques lidar actuelles présentent néanmoins des limitations. Les sources laser utilisées sont limitées à quelques longueurs d’onde visible ou proche infrarouge. Les propriétés optiques des particules composant le milieu diffusant sont donc relatives à ces longueurs d’onde, ce qui ne permet pas aujourd’hui d’identifier complètement les propriétés microphysiques du milieu sans en avoir une connaissance a priori. Des hypothèses sont formulées pour contraindre les méthodes d’inversion et remonter aux informations d’intérêt (i.e. nature, forme ou indice complexe des aérosols) des aérosols naturels et anthropiques (particules carbonées, poussières désertiques, cendres volcaniques, etc.). En outre, les lidars actuels sont souvent destinés à des portées kilométriques alors que de nouveaux besoins tendent vers des mesures à courtes distances : réaliser des mesures in situ en sortie de tuyère, contrôler localement les panaches de particules industrielles à haute résolution spatiale… Les sources laser supercontinuum s’étendent du visible à l’infrarouge et offrent de nouvelles possibilités de mesures des propriétés optiques des aérosols sur un large domaine spectral continu. Leur utilisation dans les systèmes lidar est donc envisagée pour obtenir les paramètres optiques du milieu sur toute une gamme continue de longueurs d’onde. L’un des objectifs majeurs de cette thèse est de démontrer la viabilité des lidars supercontinuum pour la caractérisation spectrale de cibles surfaciques et volumiques à haute résolution spatiale.Trois axes de recherche ont été menés. Le premier concerne l’étude numérique du système. Un simulateur lidar hyperspectral, nommé PERFALIS a été développé pour prédire les performances de l’instrument et montrer d’un point de vue numérique la faisabilité des mesures. Le second axe porte sur la conception instrumentale d’un lidar appelé COLIBRIS, développé pour réaliser des mesures lidar à courtes portées (> 10 m) à haute résolution spatiale (submétrique). Une première version a été proposée en analyse monochromatique et une seconde avec une analyse hyperspectral de la lumière rétrodiffusée. Le dernier axe concerne le développement d’une nouvelle méthode d’inversion lidar, nommée ST-LIM. Cette méthode a permis d’identifier les propriétés optiques d’un panache d’huile sans hypothèse a priori sur le lidar ratio (paramètre optique caractérisant le milieu sondé). La comparaison de mesures expérimentales avec les résultats de simulations a permis de valider le simulateur lidar dans ses deux modes de fonctionnement (monochromatique et spectral). Nous montrons entre autres qu’il est important de considérer la forme réelle des impulsions laser lors de l’interaction lumière-matière dans le cas de sondages de panaches de faible épaisseur à haute résolution spatiale. À l’avenir, le simulateur lidar pourra être utilisé pour dimensionner de nouveaux instruments lidar supercontinuum ou monochromatique dédiés à des applications spécifiques ou encore d’étudier les performances de nouveaux prototypes COLIBRIS pour en faire une version aéroportée.

In this paper, a new elastic lidar inversion equation is presented. It is based on the backscattering signal from a surface reference target (SRT) rather than that from a volumetric layer of reference (Rayleigh molecular scatterer) as is usually done. The method presented can be used when the optical properties of such a layer are not available, e.g., in the case of airborne elastic lidar measurements or when the lidar–target line is horizontal Also, a new algorithm is described to retrieve the lidar ratio and the backscattering coefficient of an aerosol plume without any a priori assumptions about the plume. In addition, our algorithm allows a determination of the instrumental constant. This algorithm is theoretically tested, viz. by means of simulated lidar profiles and then using real measurements. Good agreement with available data in the literature has been found.

We report on a direct detection differential absorption lidar (DIAL), designed for remote detection of CH4 and CO2. The system is based on a single-frequency optical parametric oscillator/amplifier system, tunable in the 1.57-1.65 µm range. The DIAL system, called NAOMI GAZL, was tested on a controlled gas release facility in October 2018.

International audience; This chapter aims at presenting thematic applications of hyperspectral imagery related to human activities. This chapter is a companion of Chapter 3.1 in which four applications related to monitoring of natural resources or inspection of scenes were presented.

International audience; The monitoring of soil contamination deriving from oil and gas industry remains difficult in vegetated areas. Over the last decade, optical remote sensing has proved helpful for this purpose. By tracking alterations in vegetation biochemistry through its optical properties, multi- and hyperspectral remote sensing allow detecting and quantifying crude oil and petroleum products leaked following accidental leakages or bad cessation practices. Recent advances in this field have led to the development of various methods that can be applied either in the field using portable spectroradiometers or at large scale on airborne and satellite images. Experiments carried out under controlled conditions have largely contributed to identifying the most important factors influencing the detection of oil (plant species, mixture composition, etc.). In a perspective of operational use, an important effort is still required to make optical remote sensing a reliable tool for oil and gas companies. The current methods used on imagery should extend their scope to a wide range of contexts and their application to upcoming satellite-embedded hyperspectral sensors should be considered in future studies.

International audience; During five years, in the frame of the NAOMI (New Advanced Observation Method Integration) research project, Total and ONERA have worked on radar and optical imagery to detect, characterize and quantify slicks at sea. Laboratory and pool measurements, physical modelling and offshore experiments have been combined to fully understand the signal collected over slick-covered area. As the measured signal is analytically expressed according to the geophysical parameters of the imaged slick, it enables to fully monitor the ocean surface: is a slick present? What kind of slick is it (extremely thin or not)? Is it a known product (existing in the data base)? Can the thickness be probed by the used of optical or radar device? What is the slick volume? In the Health Security and Environment (HSE) context, an exhaustive measurement campaign can be done in order to create a data base with hydrocarbon or hydrocarbon emulsion signatures, extinction coefficients, skin depths, minimum thicknesses perceptible thanks to extinction and thickness values. Thus, it offers more processing options in the optic branch of the tool to monitor the slick. Depending on the available data, optical and/or radar imagery, the capability of slick detection, characterization and quantification will be presented. After a recall of the HSE specificity, the paper will give an overview of the main features of the input data that is to say SAR and optical images. Then, based on modelling results, the optimal observation conditions for radar and optical imagery will be introduced. Afterwards, capability of detection will be described and illustrated for both the radar and the optical case. In the optical domain, the process will distinguish at least two classes: thin and thick. In the HSE context, a database can be used to identify some detected products. The last step is quantification. A sophisticated method, relying on L band radar imagery, will be used to identify pixels covered by a film, meaning presence of oil at the surface, and the ones for which the oil may be as droplets in the volume. The traditional use of SAR data is also extended to the estimation of the oil concentration within an oil and seawater mixture. For optical data, the most direct quantification process relies on automatic Bonn code classification. The code links a class with a range of thickness and computes a minimum and a maximum volume of product in each class. If the product is in the data base a more suited classification and volume assessment can be done. If the thickness is too thin (spectral signature due to absorption is too weak) or too thick (only the upper part of the product layer contributes to the signal), a thickness estimated thanks to pool experiment is associated to each class enabling to compute a volume per class and a global volume. In the other cases, in a near future, modelling would enable to assess the thickness. Concerning hydrocarbon emulsions, modelling in the optical domain is in progress in order to predict skin depth and to derive water content.; Pendant cinq ans, dans le cadre du projet de recherche NAOMI (New Advanced Observation Method Integration), Total et l'ONERA ont travaillé dans le domaine de l'imagerie radar et optique pour détecter, caractériser et quantifier les nappes d'hydrocarbures en mer. Des mesures en laboratoire et en bassin, de la modélisation physique et des expérimentations en mer ont été associées pour comprendre pleinement le signal mesuré au dessus d'une nappe. L'expression analytique du signal en fonction des paramètres géophysiques permet d'analyser et de suivre la surface marine : Une nappe est-elle présente ? Quel type de nappe est-ce (très fine ou pas) ? Le produit est-il connu (dans la base de données existante) ? L'épaisseur de la nappe peut-elle être mesurée par imagerie radar ou optique ? Quel est le volume de la nappe ?Dans le contexte Hygiène Sécurité et Environnement (HSE), une campagne de mesure exhaustive peut être faite pour créer une base de données avec les signatures des hydrocarbures ou des émulsions, les coefficients d'extinction, les épaisseurs de peau, les épaisseurs minimales perceptibles par extinction et les épaisseurs.Selon les données disponibles, imagerie radar et/ou optique, le potentiel de détection, caractérisation et quantification de nappe sera présenté. Après un rappel des spécificités HSE, le papier passera en revue les principales caractéristiques des données d'entrée, à savoir les images SAR et optiques. Ensuite, à partir des résultats de modélisation, on introduira les conditions d'observation optimales pour le radar et pour l'optique. Après cela, la capacité de détection sera présenté et illustrée pour le radar et pour l'optique. En optique, le traitement distingue a minima deux classes : fin et épais. Dans le contexte HSE, une base de données peut être utilisée pour identifier les produits détectés. La dernière étape est la quantification. Une méthode sophitiquée, s'appuyant sur la bande L en imagerie radar, sera utilisée pour identifier les pixels couverts par un film, ce qui signifie une présence l'huile en surface, et les pixels pour lesquels l'huile est sous forme de gouttelettes dans le volume. L'utilisation classique du SAR est aussi étendue à l'estimation de la concentration en huileau sein d'un mélange huile et eau de mer. Pour les données optiques, le traitement de quanticication le plus direct repose sur une classification automatique de type code de Bonn. Le code lie une classe avec une plage d'épaisseur et calcule un volume minimum et un volume maximum de produit dans chaque classe. Si le produit se trouve dans la base de données, une classification et une estimation de volume plus adaptées peuvent être faites. Si l'épaisseur est trop fine (signature spectrale due à l'absorption trop faible) ou trop épaisse (seule la partie supérieure de la couche de produit contribue au signal), une épaisseur estimée grâce à l'expérimentation en bassin est associée à chaque classe permettant de calculer un volume par classe et un volume global. Dans les autres cas, à court terme, la modélisation devrait permettre d'estimer l'épaisseur. Concernant les émulsions d'hydrocarbure, la modélisation en optique est en cours pour prédire l'épaisseur de peau et le contenu en eau.

International audience; Most satellite images are polluted by the presence of semi-transparent clouds, called cirrus, returning them unusable. Gao and Li (2017) have developed an empirical method for the correction of thin cirrus for Landsat-8, allowing to return to a clear sky apparent reflectance. Validated qualitatively, we propose here a quantitative validation method using Sentinel-2 data by comparing the corrected image with a clear sky reference image. Their method shows good results on dark surfaces, like water, with an apparent reflectance found close to 0.02. On the other hand, it becomes less accurate for thicker cirrus and on more reflective surfaces. In addition, the data analysis shows that pixels located in the shadow of the cirrus are over-corrected. The downward path should therefore be taken into account when correcting the signal.; Dans le domaine réflectif [0,4 - 2,5 μm], la télédétection par spectroscopie depuis l'espace permet d'accéder aux propriétés intrinsèques des surfaces, la réflectance, une fois les mesures corrigées des effets de l'atmosphère. Pour les scènes plates et sous l'hypothèse de réflectance lambertienne, cette correction atmosphérique est aujourd'hui très bien contrôlée en ciel clair. Mais statistiquement, les nuages couvrent environ les 2/3 de la surface Terrestre et réduisent donc considérablement le nombre d'images exploitables. Bien que des nuages épais comme le stratocumulus masquent complètement le paysage à acquérir, des cirrus fins, semi-transparents et situes à haute altitude, peuvent laisser entrevoir la surface sous-jacente mais induisent certains effets radiatifs qu'il convient de corriger pour retrouver la réflectance.Ce travail présentera une validation d'une méthode de correction des cirrus fins de la littérature.Dans un premier temps, une méthode de détection est développée par Gao et al. (1993). Elle est base sur l'utilisation simple d'une bande située près de 1.38 μm, qui est une bande d'absorption de vapeur d'eau saturée. Ensuite, une méthode de correction est proposée par Gao et al. (1995) dans la gamme spectrale [0,4 - 1,0 μm], ou ils supposent que le seul impact radiatif de tels nuages est du à la diffusion, qui ne dépend pas de la longueur d’onde (pas d’absorption). La méthode repose sur la correction d’une augmentation de la réflectance apparente, cause par la présence de cirrus. Cette correction est obtenue en utilisant un coefficient de corrélation calcule à partir de 2 bandes spectrales (1.38 vs 1.24 μm pour les surfaces d’eau et 1.38 vs 0.66 μm pour les surfaces mixtes). Plus récemment, Gao et al. (2017) ont amélioré la méthode en corrigeant chaque bande avec un coefficient distinct par bande n, calcule à partir de la corrélation linéaire entre la bande de cirrus (1.38 μm) et la bande n, et l'ont étendue au SWIR. Cette méthode est utilisée depuis de nombreuses années, mais sans validation quantitative. Aujourd'hui, avec l'avènement de la famille multispectrale Sentinel-2, grâce à sa haute revisite, nous avons la possibilité de réaliser cette validation en sélectionnant des couples d'images acquises sur la même scène, l'une avec cirrus et l'autre sous un ciel dégagé.L’étude a été initialement réalisée sur les surfaces d’eau et a donné de bons résultats avec des valeurs trouvées à 0,015 près, sur l’ensemble du spectre. Les cirrus plus épais (ρ * (1.38 μm)> 0.04) sont corriges avec la même précision. Cependant, la présence de nuages d'eau élevés ou de contrails perturbe l'estimation du coefficient de corrélation, ce qui entraine des erreurs plus importantes.Pour les surfaces continentales, la précision de la correction est plus difficile à analyser car celle-ci dépend du type de surface. La correction echoue dans le SWIR et aléatoirement entre 0.7 et 0.9 μm. En effet, pour une surface donnée, la présence de cirrus ne provoque pas nécessairement un décalage positif, entrainant une sur-correction de la méthode de Gao. Cela peut s’expliquer à partir de nos simulations, la méthode de Gao supposant un cirrus uniforme et homogène alors que le cirrus peut n'être présent que localement sur le trajet soleil-sol ou sur le trajet capteur-surface.Cette compréhension du phénomène conduira à proposer une amélioration du modèle de Gao et Li.

International audience; In the frame of the monitoring of Sentinel 2 MSI sensors in orbit, a method for inter-band calibration validation, using Antarctic site, has been developed and tested. The method relies on the directional hemispheric reflectance found in ASTER data base. In order to consider that the spectral shape of the directional directional reflectance is the same as the directional hemispheric one, wavelengths have to be limited to the spectral range below 0.9 µm. Thus this method cannot perform the inter-band calibration for the whole set of spectral bands of MSI but only from B1 to B8A. Results for B9 seem also correct although the central wavelength is a little bit beyond 0.9 µm. Given a clear day acquisition on DomeC site, a first set of computations of Top Of Atmosphere (TOA) reflectances for MSI spectral bands is performed using the three ASTER reflectances corresponding respectively to small, medium and coarse grain size for snow. The spectral reflectances are normalized by the B2 value. The comparison of the three normalized TOA reflectances to the actual one given by MSI enables to select the right snow grain size. Then the second step is the atmospheric parameters adjustment. Water content is assessed thanks to B9 band centered at 0.945 µm and ozone is assessed thanks to mainly B3 band centered at 0.56 µm. For a clear day, the aerosol optical thickness is weak and thus the TOA reflectance is not very sensitive to the more or less low abundance of continental aerosols. The ratio of normalized reflectance given by MSI to the normalized reflectance given by the computation is assessed for each spectral band between B1 and B9 and is compared to the inter-band calibration validations performed with other methods.; Dans le cadre du suivi en orbite du capteur MSI de Sentinel 2, une méthode de validation de l'étalonnage inter-bande a été développée et testée. La méthode s'appuie sur la réflectance hémisphérique - directionnelle que l'on trouve dans la base de données ASTER. Pour pouvoir considérer que la réflectance directionnelle - directionnelle est similaire à la réflectance hémisphérique - directionnelle, il faut se limiter aux longueurs d'onde inférieures à 0,9 µm. De ce fait, la méthode ne peut pas assurer la validation de l'étalonnage inter-bande pour l'ensemble des bandes spectrales de MSI mais uniquement pour les bandes B1 à B8A. Les résultats pour la bande B9 semblent également corrects bien que la longueur d'onde centrale soit un peu au delà de 0,9 µm.Pour une acquisition en ciel clair sur Dome C, un premier ensemble de réflectance en haut de l'atmosphère (TOA pour Top Of Atmosphère) pour les bandes spectrales de MSI est calculé en utilisant les trois réflectances ASTER correspondant à une taille de grain de neige respectivement petite, moyenne ou grosse. Les réflectances spectrales sont normalisées par leur valeur en B2. La comparaison des trois réflectances ainsi obtenues avec la réflectance réelle donnée par MSI permet de sélectionner la bonne taille de grain de neige. La seconde étape correspond à l'ajustement des paramètres atmosphériques. Le contenu en vapeur d'eau est estimé à partir de la bande B9 centrée à 0,945 µm et l'ozone est estimée à partir de la bande B3 centrée à 0,56 µm. En condition de ciel clair, l'épaisseur optique des aérosol est faible et de ce fait, la reflectance TOA n'est pas très sensible à l'abondance plus ou moins grande des aérosols continentaux. Le rapport de la réflectance normalisée fournie par MSI à la réflectance normalisée calculée est estimé pour chaque bande spectrale entre B1 et B9 et est comparé aux validations de l'étalonnage inter-bande réalisées avec d'autres méthodes.

Large-scale mapping of coastal oil spills and their monitoring over time is a major issue that can be adressed by using hyperspectral images and dedicated processing. Previous researches have shown that it is possible to map the polluted coastline caused by the explosion of the Deepwater Horizon (DwH) platform from AVIRIS images (AVIRIS: Airborne Visible/InfraRed Imaging Spectrometer). But the detection processes required either ground truth or laboratory spectra of hydrocarbons or were not fully automatic. In this paper we focused on an AVIRIS image which covers The Bay Jimmy, located south of New Orleans, and particularly impacted by oil pollution. Two automatic methods were developed to detect oiled coasts. In the first one, we have developed a new spectral index able to detect directly hydrocarbon and less sensitive to noise than indices proposed in previous works. The second one extracts endmembers via Orthogonal Subspace Projection, and then sorts the endmembers in terms of hydrocarbon indices scores, in descending order. Then, the detection map or the abundance map corresponding to the best endmember is used to map oiled areas. Both approaches give results consistent with those of studies previously conducted on the same image, and with maps built from field observations., La cartographie à grande échelle des marées noires côtières et leur suivi dans le temps est un problème majeur qui peut être résolu en utilisant des images hyperspectrales et un traitement dédié. Des recherches antérieures ont montré qu'il était possible de cartographier le littoral pollué par l'explosion de la plate-forme Deepwater Horizon (DwH) à partir d'images AVIRIS (AVIRIS : Airborne Visible/InfraRed Imaging Spectrometer). Mais les processus de détection nécessitaient soit l’utilisation de mesures in-situ, soit des spectres d'hydrocarbures en laboratoire, ou bien n'étaient pas entièrement automatiques.Dans cet article, nous nous sommes concentrés sur une image hyperspectrale AVIRIS qui couvre La Baie Jimmy, située au sud de la Nouvelle-Orléans, et particulièrement affectée par la pollution. Deux méthodes automatiques ont été mises au point pour détecter les côtes mazoutées. Dans la première, nous avons développé un nouvel indice spectral capable de détecter directement les hydrocarbures et moins sensible au bruit que les indices proposés dans les travaux précédents. La seconde utilise une méthode de démélange spectral, puis trie les pôles de mélange en termes d'indices d'hydrocarbure par ordre décroissant. Les cartes de détection ou d'abondance correspondant au meilleur pôle de mélange sont alors utilisées pour cartographier les zones mazoutées. Les deux approches donnent des résultats cohérents avec ceux d'études réalisées précédemment sur la même image et avec des cartes établies à partir d'observations sur le terrain.

International audience; To date, estimating oil thickness on the sea surface remains a challenge in most cases. When oil thickness estimation using optical data is limited by the absorption properties of the target, a solution consists in combining experimental and airborne hyperspectral data. We developed a method to identify thickness classes from hyperspectral data which, combined with realistic thickness values derived from a pool experiment, allows to estimate slick volume. Hyperspectral images of the same oil emulsion were acquired over a pool and at sea, under real conditions. From the pool data, we derived two classes: the sheen and the thick pixels, along with their respective thickness. These classes are then identified on the airborne images acquired during the NOFO campaign by generating a detection mask and using two classification approaches based on spectral indices. The proposed method allows to correctly identify the two thickness classes and, combined with the data from the pool experiment, provides a total slick volume larger than the one derived for the Bonn Agreement Oil Appearance Code.

International audience; Land surface emissivity is required for deriving surface temperature from thermal infrared radiances. When using single-channel or two-channel thermal infrared sensors, information on emissivity may be derived from spectral reflectance measurements through regression models. In this study, we present relationships derived over bare soils for Landsat 7 – ETM+ sensor. Reflectances in ETM+ channels were obtained from soil spectra (between 0.4 and 13 μm) extracted from the ASTER spectral library and the dataset acquired by Lesaignoux et al. (2013). The best relations were obtained between reflectances in the mid-infrared channels (ETM5 and ETM7) and the thermal infrared channel (ETM6) with correlation coefficients of 0.63 and 0.72 respectively. The relations were mostly generated by the variations of soil reflectances due to changes in soil moisture. Correlations were lower when considering the variations due to soil type.

To ensure confidence, measurements carried out by imaging radiometers mounted on satellites require robust validation using “fiducial quality” measurements of the same in situ parameter. For surface temperature measurements this is optimally carried out by radiometers measuring radiation emitted in the infrared region of the spectrum, collocated to that of a satellite overpass. For ocean surface temperatures the radiometers are usually on board ships to sample large areas but for land and ice they are typically deployed at defined geographical sites. It is of course critical that the validation measurements and associated instrumentation are internationally consistent and traceable to international standards. The Committee on Earth Observation Satellites (CEOS) facilitates this process and over the last two decades has organized a series of comparisons, initially to develop and share best practice, but now to assess metrological uncertainties and degree of consistency of all the participants. The fourth CEOS comparison of validation instrumentation: blackbodies and infrared radiometers, was held at the National Physical Laboratory (NPL) during June and July 2016, sponsored by the European Space Agency (ESA). The 2016 campaign was completed over a period of three weeks and included not only laboratory-based measurements but also representative measurements carried out in field conditions, over land and water. This paper is one of a series and reports the results obtained when radiometers participating in this comparison were used to measure the radiance temperature of the NPL ammonia heat-pipe blackbody during the 2016 comparison activities (i.e., an assessment of radiometer performance compared to international standards). This comparison showed that the differences between the participating radiometer readings and the corresponding temperature of the reference blackbody were within the uncertainty of the measurements, but there were a few exceptions, particularly for a reference blackbody temperature of −30°C. Reasons that give rise to the discrepancies observed at the low blackbody temperatures were identified.

International audience; The future space joint-mission TRISHNA (Thermal infraRed Imaging Satellite for High-resolution Natural resource Assessment, CNES and ISRO) will allow to retrieve LST at 60 m spatial resolution every 3 days, improving the urban environment monitoring capacities. Two methods can be used to derive LST from remote sensing TIR data, Split-Window (SW) or TES (Temperature Emissivity Separation). Even if proven efficient, the complexity and heterogeneity of urban areas limit the performance of those methods. This paper focuses on the TES algorithm. For TES, errors can be generated by the impact of the surface roughness on the data. In order to investigate the efficiency of the future TRISHNA mission, research was conducted to evaluate the TES algorithm. To do so, the impacts of using a representative material database when using TES were explored and validated based on airborne or satellite data, and the impact of the surface geometry on the TES estimations was evaluated using 3D thermo-radiative model simulations. This study shows that TES can be used for TRISHNA images with a mean difference of 1.6 K between validation measurements and the retrieved LST for both daytime and nighttime images and that neglecting the impact of the surface geometry while performing TES could lead to errors up to 4 degrees C in LST estimations.

The climate change context, along with the increasing scarcity and deteriorating quality of the water resource leads to monitor different components of the water cycle. A particular attention has to be paid to some areas, particularly, agricultural lands which represent about 70% of the water consumption, and coastal areas subject to strong interactions with land. Thermal infrared (TIR) data from space are well adapted to these purposes, but the spatial variability of the surface requires that the complexity of both physical and biological processes involved must be assessed at a smaller scale which corresponds to the scale at which decisions concerning water management or implementation of policies devoted to the mitigation of climate change effect are effective. In addition, surface fluxes show short-time scale variability, which requires frequent observations to be done. The need of space borne systems combining both high spatial resolution and high revisit frequency in thermal infrared (TIR), which do not exist today, is now largely recognized by the scientific community and end-users, especially as SENTINEL and RESOURCESAT missions now provide high quality complementary data in the optical domain. To fill this gap a project, TRISHNA (Thermal infraRed Imaging Satellite for High-resolution Natural resource Assessment), is currently in the feasibility assessment phase, conducted by the French Space Agency (CNES) and the Indian Space Research Organization (ISRO).Two scientific objectives drive the mission specifications, (i) monitoring of ecosystem stress of the continental biosphere and of water use with applications to agriculture and hydrology, and (ii) monitoring of coastal and continental waters. Four complementary goals enlarge the community aggregated around the project: (iii) urban microclimates monitoring (urban heat islands, mitigation of heat waves effects…), (iv) applications to Solid Earth/geology (detection of thermal anomalies, volcanology, peat fires…), (v) cryosphere monitoring (glaciers, polar regions…), and (vi) applications to atmosphere (water content, clouds…). These objectives and the expected applications will first be briefly reviewed at the symposium.The main mission specifications will then be presented. Additionally to previous work conducted to consolidate the revisit and resolution specifications, emphasis will be put on recent studies made for inventorying factors possibly perturbing surface temperature measurements. Significant advances obtained for characterizing and modelling TIR directional anisotropy and thermal hot spot effects will first be described and their impact on the choice of orbit discussed. Similarly, an original study of the impact of atmospheric turbulence on the accuracy of LST -largely ignored by the community till date- will be presented; it reinforces the need of high revisit, and allows to estimate the errors on LST measurements. An overpass time in the early afternoon, around 13:00 LST has been carefully justified to cope with the different objectives of the mission and to optimize the accuracy on retrieved fluxes. A baseline spectral configuration of 4 TIR channels within the range 8 - 12 µm is under study, which makes possible implementing both split-window and temperature-emissivity separation algorithms. The main specifications of VNIR instrument embarked aboard the same platform are presented, and the need of 6 bands, 4 in the VNIR (blue, green, red, near infrared at 485, 555, 670, 860 nm) and 2 in the SWIR (1.38 and 1.61 µm) justified. To cope with the requirement of global coverage at Equator with a single satellite, a revisit of 3 days is selected. The nadir resolution of 50 m is binned at 1 km over open ocean. The instrumental studies undertaken will be briefly presented.The programmatic context of existing missions will be analyzed, only TRISHNA and the ESA LSTM mission providing high spatio-temporal capacities. The forthcoming phases of the CNES-ISRO TRISHNA project will finally be briefly mentioned, for a launch date foreseen at the 2024-2025 horizon.

International audience; In-flight assessment of the radiometric performances of space-borne instruments can be achieved by means of vicarious calibration over Pseudo-Invariant Calibration Sites (PICS). PICS are chosen for the high temporal stability of their surface optical properties combined with a high spatial homogeneity. A first list of the main desert PIC sites was identified 20 years ago for the calibration of medium/coarse spatial resolution instruments in the solar spectral range (400–2500 nm). They are located in the Saharan desert and in the Arabian Peninsula. Six of them have since been endorsed by the CEOS/WGCV/IVOS as reference Calibration/Validation test sites. In this study, we have revisited the list of desert PIC sites at the global scale with the aim of (1) assessing if these twenty PICS are still “optimal”, in terms of temporal stability and spatial uniformity, and using up-to-date multi-spectral remote sensing data, and (2) identifying new calibration sites distributed over other areas of the world. We verified that the original sites remain very relevant, although alternate locations in their close vicinity have slightly better characteristics. We proposed four additional targets with similar characteristics, some of which may offer easier logistical access. In order to support radiative transfer simulations of satellite sensor measurements over the sites, we assessed the abilities of several semi-empirical models to reproduce the spectro-directional signatures of six IVOS sites and the four new candidate sites, and we derived climatologies of the main atmospheric properties (trace gas column load and aerosol optical depth).

Timely processing of observations from multi-spectral imagers, such as SEVIRI (Spinning Enhanced Visible and Infrared Imager), largely depends on fast radiative transfer calculations. This paper mostly concerns the development and implementation of a new forward model for SEVIRI to be applied to real time processing of infrared radiances. The new radiative transfer model improves computational time by a factor of &asymp, 7 compared to the previous versions and makes it possible to process SEVIRI data at nearly real time. The new forward model has been applied for the retrieval of surface parameters. Although the scheme can be applied for the simultaneous retrieval of temperature and emissivity, the paper mostly focuses on emissivity. The inverse scheme relies on a Kalman filter approach, which allows us to exploit a sequential processing of SEVIRI observations. Based on the new forward model, the paper also presents a validation retrieval performed with in situ observations acquired during a field experiment carried out in 2017 at Gobabeb (Namib desert) validation station. Furthermore, a comparison with IASI (Infrared Atmospheric Sounder Interferometer) emissivity retrievals has been performed as well. It has been found that the retrieved emissivities are in good agreement with each other and with in situ observations, i.e., average differences are generally well below 0.01.

The Indian and French Space Agencies, ISRO and CNES, have conceptualized a space-borne Thermal Infrared Reflectance (TIR) mission, TRISHNA (Thermal infRared Imaging Satellite for High-resolution Natural Resource Assessment). The primary design drivers of TRISHNA are the monitoring of (i) terrestrial water stress and use, and of (ii) coastal and continental water. A suit of four TIR bands and six optical bands is planned. The TIR bands will be centred at 8.6 μm, 9.1 μm, 10.3 μm and 11.5 μm to provide noon-night global observations at 57m nadir resolution over land and coastal regions. The field of view (FOV) is ±34° and the orbit of 761 km altitude was designed to allow 3 sub-cycle acquisitions during the 8-day cycle. The optical bands correspond to blue, green, red, and NIR plus two SWIR bands at 1.38 μm and 1.61 μm. The green, red, NIR and the 1.61 μm SWIR bands will have better radiometry quality than those of AWiFS. ISRO and CNES will develop optical and TIR payloads, respectively. Assessing evapotranspiration and furthermore Gross and Net Primary Productivity (GPP and NPP) will in turn assist in quantifying water use in rainfed and irrigated agriculture, water stress and water use efficiency, with expected applications to agricultural drought and early warning, crop yield prediction, water allocation, implementation of water rights, crop insurance business and agro-advisories to farmers. The other scientific objectives of TRISHNA are also briefly described. TRISHNA instrument will fly aboard a ISRO spacecraft scheduled to be launched from 2024 for a minimum period of 5 years’ mission lifetime.

International audience; Ceilometers are widely spread instruments mainly used for cloud base height determination. However, recent models are more sensitive to aerosols; hence, nowadays there is an increasing interest in these instruments to retrieve aerosol optical and microphysical properties.The high number of them distributed around the globe compared to the robust lidar systems and the fact that ceilometers work under unattended conditions are the main reasons behind this interest. In this paper a new methodology is proposed to retrieve aerosol vertical extinction and backscatter profiles from a Vaisala ceilometer CL51. This methodology is based in two parts: first, a signal pre-processing with a suppression of the dark current and background noises, and a correction of the water vapor absorption using near-real time temperature and absolute humidity (AH) profiles from a co-located Microwave radiometer (MWR). Then an iterative Klett-based algorithm that uses AERONET (AErosol RObotic NETwork) AOD (Aerosol Optical Depth) as an input to retrieve the extinction and backscatter profiles. The error analysis of the aerosol retrievals has been performed between MWR measurements and modelled radiosondes from HYSPLIT. The absolute errors found in temperature and AH are leading to errors in the pre-processed RCS signals up to 9%, and then in particle backscatter and particle extinction coefficients up to 2:2% and 25%, respectively.; Les ceilomètres sont des instruments très répandus, principalement utilisés pour déterminer la hauteur de la base des nuages. Cependant, des modèles récents sont plus sensibles aux aérosols ; c'est pourquoi, de nos jours, on s'intéresse de plus en plus à ces instruments pour mesurer des données optiques des aérosols et leurs propriétés microphysiques. Le nombre élevé de ces derniers, répartis dans le monde entier, par rapport aux systèmes lidar robustes et le fait que les plafonds fonctionnent dans des conditions non surveillées sont les principales raisons de cet intérêt. Dans ce document, une nouvelle méthodologie est proposée pour déterminer par inversion les profils d'extinction verticale et de rétrodiffusion des aérosols à partir d'un modèle CL51 de ceilomètre de Vaisala. Cette méthodologie est basée sur deux parties : premièrement, un prétraitement du signal avec une suppression du courant d'obscurité et du fond et une correction de l'absorption de la vapeur d'eau en utilisant des profils de température et d'humidité absolue (AH) en temps quasi réel d'un radiomètre à micro-ondes (MWR) colocalisé. Ensuite, un algorithme itératif basé sur Klett qui utilise AERONET (AErosol RObotic NETwork) AOD (Aerosol Optical Depth) comme entrée est appliquée pour récupérer les profils d'extinction et de rétrodiffusion. Les erreurs absolues trouvées dans la température et l'AH entraînent des erreurs dans les signaux RCS prétraités jusqu'à 9 %, puis dans les coefficients de rétrodiffusion et d'extinction des particules jusqu'à 2,2 % et 25 %, respectivement.

In urban areas, information concerning very high resolution land cover and especially material maps are necessary for several city modelling or monitoring applications. That is to say, knowledge concerning the roofing materials or the different kinds of ground areas is required. Airborne remote sensing techniques appear to be convenient for providing such information at a large scale. However, results obtained using most traditional processing methods based on usual red-green-blue-near infrared multispectral images remain limited for such applications. A possible way to improve classification results is to enhance the imagery spectral resolution using superspectral or hyperspectral sensors. In this study, it is intended to design a superspectral sensor dedicated to urban materials classification and this work particularly focused on the selection of the optimal spectral band subsets for such sensor. First, reflectance spectral signatures of urban materials were collected from 7 spectral libraires. Then, spectral optimization was performed using this data set. The band selection workflow included two steps, optimising first the number of spectral bands using an incremental method and then examining several possible optimised band subsets using a stochastic algorithm. The same wrapper relevance criterion relying on a confidence measure of Random Forests classifier was used at both steps. To cope with the limited number of available spectra for several classes, additional synthetic spectra were generated from the collection of reference spectra: intra-class variability was simulated by multiplying reference spectra by a random coefficient. At the end, selected band subsets were evaluated considering the classification quality reached using a rbf svm classifier. It was confirmed that a limited band subset was sufficient to classify common urban materials. The important contribution of bands from the Short Wave Infra-Red (SWIR) spectral domain (1000–2400 nm) to material classification was also shown.

International audience; As marine transportation has increased in coastal regions, maritime accidents associated with vessels have steadily increased. Remotely sensed satellite or airborne images can aid rapid vessel monitoring over wide areas at high resolutions. In this study, airborne hyperspectral experiments were performed to detect marine vessels mainly including fishing boat and yacht by applying pixel-based mixture techniques and to estimate the size of the vessels through an objective ellipse fitting method. Various spectral libraries of marine objects and seawaters were constructed through in-situ experiments for spectral analysis of the internal structures of vessels. The hyperspectral images were dimensionally reduced through principal component analysis. Several hyperspectral mixture algorithms, such as N-FINDR, pixel purity index (PPI), independent component analysis (ICA), and vertex component analysis (VCA), were used for the detection of vessels. The N-FINDR and VCA techniques presented a total of 14 vessels, the ICA technique detected seven vessels, and the PPI technique detected two vessels. The pixel-based probability of detection (POD) and false alarm ratio (FAR) for all 14 vessels were 96.40% and 4.30%, respectively. The sizes of the vessels were estimated by extracting the boundaries of the vessels through a two-dimensional gradient and applying the ellipse fitting method. Compared with the digital mapping camera (DMC) images with resolutions of 0.10 m, the root-mean-square errors of the length and width of the vessels were approximately 1.19 m and 0.81 m, respectively. The application of spectral mixing methods provided a high probability of detecting the objects, as well as the overall structures of the decks of the vessels.

International audience; The use of hyperspectral spectroscopy for oil detection recently sparked a growing interest for risk assessment over vegetated areas. In a perspective of image applications, we conducted a greenhouse experiment on a brownfield-established species, Rubus fruticosus L. (bramble), to evaluate the potential of vegetation reflectance to detect and discriminate among various oil-contaminated soils. The species was grown for 32 days on four different soils with mixtures of petroleum hydrocarbons and heavy metals. Additional plants were grown on either uncontaminated control or water-deficient soils for comparison. Repeated reflectance measurements indicated modified spectral signatures under both oil and water-deficit exposure, from leaf to multi-plant scales. The amplitude of the response varied with mixture composition, exposure time, acquisition scale and spectrum region. Reflectance changes were linked to alterations in chlorophyll, carotenoid and water contents using vegetation indices. These indices were used to catch spectral similarities among acquisition scales and to discriminate among treatments using Kendall’s coefficient of concordance (W) and regularized logistic regression. Of the 33 vegetation indices tested, 14 were concordant from leaf to multi-plant scales (W > 0.75, p < 0.05) and strongly related to leaf biochemistry (R2 > 0.7). The 14 indices allowed discriminating between each mixture and the control treatment with no or minor confusions (≤ 5 %) at all acquisition scales, depending on exposure time. Some of the mixtures remained difficult to discriminate among them and from the water-deficit treatment. The approach was tested at the canopy scale under natural conditions and performed well for identifying bramble exposed to either one of the experimentally-tested mixtures (90 % accuracy) or to uncontaminated soil (83 % accuracy). This study provided better understanding of vegetation spectral response to oil mixtures and opens up promising perspectives for future applications.

International audience; In this paper we focus on airborne hyperspectral imaging methodology to characterize PM (Particulate Matter) near industrial emission sources. Two short-term intensive campaigns were carried out in the vicinity of a refinery in the South of France, in September 2015 and February 2016. Different protocols of in-situ PM measurements were performed, at stack measurements (flow rate and offline chemical analysis) and on-line measurement at the refinery border (size distribution, concentration and chemistry of aerosols). A multi temporal methodology to retrieve aerosol type, to map the aerosol concentration and to quantify mass flow rate from airborne hyperspectral data is described in this paper. This method applied to the refinery detected plume from the main stack yields a black carbon to sulfate ratio of 10/90 in mass inside the plume, with an average size distribution smaller than 100 nm. These results are in a good agreement with on-line analysis of aerosols at refinery border. The resulting quantitative map with a metric spatial resolution leads to a flow rate estimated of about 1g/s and is in a good agreement with in-situ stack measurements and modelling.; Ce papier propose une méthodologie de caractérisation des PM (Particulate Matter) proches des sources industrielles par imagerie hyperspectrale aéroportée. Il s’appuie sur deux campagnes de mesures réalisées autour d’une raffinerie dans le sud de la France en Septembre 2015 et Février 2016. Différents protocoles de mesures in-situ des PM ont été réalisées dans ce cadre, mesures au niveau de l’émissaire principal (debit et analyse chimique) et mesures on-line au voisinage de la raffinerie (distribution en taille, concentration et analyse chimique). Une méthodologie multi-temporelle dédiée à l’inversion des types d’aérosols présents dans le panache permettant de fournir une carte en concentration d’aérosols et d’estimer le débit est décrite dans ce papier. Appliquée à ce cas d’étude, les resultats hyperspectraux ont permis d’estimer un rapport suie/sulfate de 10/90 en masse dans le panache avec une distribution en taille inférieure à 100nm. Ces résultats sont en bon accord avec les mesures in-situ. Les cartes en concentration obtenue à une resolution métrique permettent d’estimer un debit de l’ordre de 1g/s en bon accord avec les mesures in-situ au niveau de l’émissaire et avec les résultats de modélisation de dispersion des aerosols.

International audience; The estimation of the bathymetry and the detection of targets located on the seabed of shallow waters using remote sensing techniques is of great interest for many environmental applications in coastal areas such as benthic habitat mapping, monitoring of seabed aquatic plants and the subsequent management of littoral zones. For that purpose, knowledge of the optical effects induced by the neighborhood of a given seabed target and by the water column itself is required to better interpret the subsurface upward radiance measured by satellite or shipborne radiometers. In this paper, the various sources of photons that contribute to the subsurface upward radiance are analyzed. In particular, the adjacency effects caused by the neighborhood of a given seabed target are quantified for three water turbidity conditions, namely clear, moderately turbid and turbid waters. Firstly, an analytical expression of the subsurface radiance is proposed in order to make explicit the radiance terms corresponding to these effects. Secondly, a sensitivity study is performed using radiative transfer modeling to determine the influence of the seabed adjacency effects on the upward signal with respect to various parameters such as the bathymetry or the bottom brightness. The results show that the highest contributions of the adjacency effects induced by the neighborhood of a seabed target to the subsurface radiance could reach 26%, 18% and 9% for clear, moderately turbid and turbid water conditions respectively. Therefore, the detection of a seabed target could be significantly biased if the seabed adjacency effects are ignored in the analysis of remote sensing measurements. Our results could be further used to improve the performance of inverse algorithms dedicated to the retrieval of bottom composition, water optical properties and/or bathymetry.

International audience; The monitoring of the water cycle at the Earth surface which tightly interacts with the climate change processes as well as a number of practical applications (agriculture, soil and water quality assessment, irrigation and water resource management, etc...) requires surface temperature measurements at local scale. Such is the goal of the Indian-French high spatio-temporal TRISHNA mission (Thermal infraRed Imaging Satellite for High-resolution Natural resource Assessment). The scientific objectives of the mission and research work conducted to consolidate the mission specifications are presented. Progress in modelling of surface fluxes is then discussed. The main specifications of the mission such as the revisit, the spatial resolution, the overpass time, the spectral bands and the orbit are analyzed and justified. The resulting baseline of the mission is given.

International audience; In California, water is a perennial concern. As competition for water resources increases due to growth in population, California’s tree nut farmers are committed to improving the efficiency of water used for food production. There is an imminent need to have reliable methods that provide information about the temporal and spatial variability of crop water requirements, which allow farmers to make irrigation decisions at field scale. This study focuses on estimating the actual evapotranspiration and crop coefficients of an almond and pistachio orchard located in Central Valley (California) during an entire growing season by combining a simple crop evapotranspiration model with remote sensing data. A dataset of the vegetation index NDVI derived from Landsat-8 was used to facilitate the estimation of the basal crop coefficient (Kcb), or potential crop water use. The soil water evaporation coefficient (Ke) was measured from microlysimeters. The water stress coefficient (Ks) was derived from airborne remotely sensed canopy thermal-based methods, using seasonal regressions between the crop water stress index (CWSI) and stem water potential (Ψstem). These regressions were statistically-significant for both crops, indicating clear seasonal differences in pistachios, but not in almonds. In almonds, the estimated maximum Kcb values ranged between 1.05 to 0.90, while for pistachios, it ranged between 0.89 to 0.80. The model indicated a difference of 97 mm in transpiration over the season between both crops. Soil evaporation accounted for an average of 16% and 13% of the total actual evapotranspiration for almonds and pistachios, respectively. Verification of the model-based daily crop evapotranspiration estimates was done using eddy-covariance and surface renewal data collected in the same orchards, yielding an R2 ≥ 0.7 and average root mean square errors (RMSE) of 0.74 and 0.91 mm·day−1 for almond and pistachio, respectively. It is concluded that the combination of crop evapotranspiration models with remotely-sensed data is helpful for upscaling irrigation information from plant to field scale and thus may be used by farmers for making day-to-day irrigation management decisions.; En Californie, l’eau est une préoccupation constante. À mesure que la concurrence pour les ressources en eau augmente, la croissance de la population augmente. Il existe un besoin imminent de disposer de méthodes fiables pour fournir des informations sur la variabilité temporelle et spatiale des besoins en eau des cultures. Cette étude porte sur l'estimation des coefficients d'évaporation et des cultures d'un verger d'amandiers et de pistaches situé dans la Central Valley, en Californie, pendant toute une saison de croissance, en combinant un modèle simple d'évapotranspiration des cultures avec des données de télédétection. Un ensemble de données de l'indice de végétation NDVI dérivé de Landsat-8 a été utilisé pour faciliter l'estimation du coefficient de culture de base (Kcb) ou de l'utilisation potentielle de l'eau par la culture. Le coefficient d'évaporation de l'humidité du sol (Ke) a été mesuré à partir de microlysimètres. Le coefficient de stress hydrique (Ks) a été calculé à l'aide de méthodes thermiques basées sur le couvert forestier et télédétectées en vol, à l'aide de régressions saisonnières entre l'indice de stress hydrique des cultures (CWSI) et le potentiel hydrique de la tige (stem). Ces régressions étaient statistiquement significatives pour les deux cultures, indiquant de nettes différences saisonnières dans les pistaches, mais pas dans les amandes. Dans les amandes, les valeurs maximales estimées de Kcb se situaient entre 1,05 et 0,90, tandis que pour les pistaches, elles se situaient entre 0,89 et 0,80. Le modèle indiquait une différence de transpiration de 97 mm au cours de la saison entre les deux cultures. L'évaporation du sol représentait en moyenne 16% et 13% de l'évapotranspiration réelle totale d'amandes et de pistaches, respectivement. La vérification de l'évapotranspiration quotidienne des cultures basée sur un modèle a été effectuée à l'aide des données de covariance de Foucault et de renouvellement de la surface recueillies dans les mêmes vergers, conduisant à un R2 ≥ 0,7 et des erreurs moyennes de racine carrée (RMSE) de 0,74 et 0,91 mm · jour-1 pour amande et pistache, respectivement. Il est conclu que la combinaison de modèles d'évapotranspiration de cultures et de données de télédétection est utile pour augmenter l'information d'irrigation de l'échelle de la plante au champ et peut donc être utilisée par les agriculteurs pour prendre des décisions quotidiennes de gestion de l'irrigation.

As compared to near InfraRed passive instruments, lidars reduce the blases on the measurement of concentrations in the presence of diffusion by aerosols, but until recent days they were limited to one active wavelength at a time.CNES and ONERA conceived a novel type of lidar based on short frequency combs that provides the spectrum of a complete absorption line of the target molecule. The technique of Dual Comb Spectroscopy is adapted to short (up to 13 teeth) combs, as a compromise between spectral richness and Signal to Noise Ratio. The advantages of DCS remain: reduction of the impact of scene evolutions and spectral mis-calibration. The latter will make the high power laser emitters easier. Above all, the acquisition of a complete absorption line opens the path to the retrieval of two or more independent elements in the vertical profile of molecule concentration.We describe the technique of comb generation, the layout of the emitter section and the detection principles. Then we introduce the Short Comb Atmospheric Lidar Experiment (SCALE) project led by CNES in the frame of a new advance "phase 0" study. We expose the preliminary performances of this space lidar that could fit on a microsatellite platform in the case of CO2 sounding at 1.57 or 2 um.Results of CNES-ONERA four-year R&T activities are also presented that validate in the laboratory and in open space the short comb lidar.Finally first results of retrieval simulations are exposed., Les lidars réduisent l'imprécision des instruments passifs sur l'estimation de la distance sondée, donc sur la concentration de la molécule ciblée dans l'atmosphère. Mais jusqu'à récemment, les lidars étaient limités à une seule longueur d'onde active à la fois.Le CNES et l'ONERA ont conçu un nouveau type de lidar basé sur des mini- peignes de fréquence qui fournissent le spectre d'une ligne d'absorption complète de la molécule cible. La technique de la spectroscopie à double peigne (DCS) est adaptée aux mini-peignes (jusqu'à 13 dents), qui fait un compromis entre la richesse spectrale et le rapport signal/bruit. Les avantages du DCS sont : réduction de l'impact des évolutions temporelles de la scène, étalonnage auto spectral, étalement spectral de l'énergie par rapport au seuil de Brillouin. Ce dernier avantage facilitera la réalisation des émetteurs laser à haute puissance. L'acquisition d'une ligne d'absorption complète ouvre la voie à la récupération de deux ou plusieurs éléments indépendants (DOFS) dans le profil vertical de la concentration moléculaire.Nous décrivons la technique de génération de peigne, l'architecture de la section émettrice et les principes de détection. Nous présentons ensuite le projet SCALE (Short Comb Atmospheric Lidar Experiment) que le CNES étudiera prochainement dans le cadre d'une nouvelle étude avancée de "phase 0". Nous exposons les performances préliminaires de ce lidar spatial qui pourrait tenir sur une plate-forme microsatellite dans le cas d'un sondage CO2 à 1,5 µm.Enfin, les résultats des activités de R&T quadriennales du CNES-ONERA sont présentés pour valider en laboratoire et en espace libre le lidar a mini-peigne.

International audience; Detecting onshore hydrocarbon is a major topic for both environmental monitoring and exploration. In this work, a hyperspectral image acquired nearby an old oil extraction site is analyzed. The area of interest includes a pit filled with bio-degraded heavy oil, surrounded by herbaceous vegetation and many lagoons. First, we focused on methodologies that allow tracking hydrocarbon pollution in non-supervised setting. Based on the assumption that such pits are rare events in the image, statistical approach for anomalies detection, derived from the Reed-Xiaoli detector, is used for their detection. In order to decrease the false alarms rate, some a priori knowledge about the spectral signature of the pits and about the background is introduced. This approach succeeds in detecting the pit with very few false alarms.Hydrocarbon pollution can have an impact on vegetation. It can induce a stress, or conversely it can boost some vegetation species. In order to map the polluted area without any a priori knowledge, several un-supervised classification, including an original frame of automatic classification combining unmixing approach and SVM (support Vector Machine) are applied and compared. The results are compared with a partial “ground truth map” that has been derived from visual observations on the field, and with areas of stressed vegetation that have been mapped using combination of vegetation spectral indices. The classification results are consistent with the ground truth map and the retrieved the stressed vegetation areas.; La détection des hydrocarbures à terre est un sujet majeur tant pour la surveillance de l'environnement que pour l'exploration. Dans ce travail, une image hyperspectrale acquise à proximité d'un ancien site d'extraction pétrolière est analysée. La zone d'intérêt comprend un bourbier remplie de pétrole lourd biodégradé, entourée d'une végétation herbacée et de nombreuses lagunes. Tout d'abord, nous nous sommes concentrés sur les méthodologies qui permettent de détection la pollution par les hydrocarbures manière non supervisée. En partant de l'hypothèse que ces bourbiers sont des événements rares dans l'image, une approche statistique pour la détection des anomalies, dérivée de la méthode proposée par Reed-Xiaoli, est utilisée pour leur détection. Afin de diminuer le taux de fausses alarmes, des connaissances a priori sur la signature spectrale de ces bourbiers et sur la nature de l’environnement sont introduites. Cette approche réussit à détecter le bourbier avec très peu de fausses alarmes.La pollution par les hydrocarbures peut avoir un impact sur la végétation. Elle peut induire un stress, ou inversement, elle peut stimuler certaines espèces végétales. Afin de cartographier la zone polluée sans aucune connaissance a priori, plusieurs classifications non supervisées, y compris une approche originale de classification automatique combinant le dé-mélange spectral et le SVM (Support Vector Machine) sont appliquées et comparées. Les résultats sont comparés à une "carte de vérité terrain" partielle dérivée d'observations visuelles sur le terrain et à des zones de végétation stressée qui ont été cartographiées à l'aide d'indices spectraux de végétation. Les résultats de la classification sont cohérents avec la carte de vérité au sol et les zones de végétation stressée.

International audience; Hyperspectral remote sensing data offer unique opportunities for the characterization of land surface in urban areas. However, no hyperspectral-unmixing based studies have been conducted to automatically detect photovoltaic panels, which represent one of the important components of energy systems in such areas. In this paper, a hyperspectral-unmixing based method is proposed to detect photovoltaic panels and to estimate their areas. This approach is based on an original multiplicative nonnegative matrix factorization (NMF) algorithm with some known photovoltaic panel spectra. The proposed method can be considered as a partial/informed NMF approach. Experiments are conducted on realistic synthetic and real data to evaluate the performance of the proposed approach. In both cases, obtained results show that the proposed method yields much better overall performance than a method from the literature.

International audience; Hydrocarbon detection is important for both environment monitoring and hydrocarbon exploration. Hyperspectral imaging and derived spectral indices are used to detect hydrocarbons. With appropriate indices, light hydrocarbons on bare ground are detected. Heavier hydrocarbons are more difficult to detect. Plastic items are very well detected. Shadows and vegetation are generating some false alarms. Detection of hydrocarbon in urban environment, or on bare soils will be possible using spectral indices while detection of hydrocarbon in remote vegetated country areas will be difficult.