• Rural Digital Europe
• 2013-2022close
• Hyper Article en Ligne close

We analyse the flux density variation associated to low frequency (LF) broadcasting transmitters observed by the ICE electric field experiment onboard DEMETER micro-satellite, observed from 01st Jan. to 09th Dec. 2010. We select five stations localised around the Mediterranean and the Black seas: Tipaza (252 kHz, 02°28’E, 36°33’N, Algeria), Roumoules (216 kHz, 06°08’E, 43°47’N, Monte Carlo), Polatli (180 kHz, 32°25’E, 39°45’N, Turkey), Nadour (171 kHz, 02°55’W, 35°02’N, Morocco) and Brasov (153 kHz, 25°36’E,45°40’, Romania). The detection of the LF transmitter signals by DEMETER micro-satellite is found to depend on the radiated power, the emitted frequency, and the orbit paths with regard to the location of the stations. This leads us to characterise the reception condition of the LF signals and to define time intervals where the detection probability is high. We show that LF signal are regularly recorded, each 12 days, when the satellite is above the broadcasting station. The signal intensity levels are principally significant during the solar activity. Hence we find that the solar and the geomagnetic activities are slightly correlated to the maxima of LF signal as recorded by DEMETER. Also we note a drop of the intensity level several days before the occurrence of earthquakes in/around the Mediterranean and Black seas. International audience

International audience; Researches of the last fifty years in the western South of France brought new evidences about the pre-colonial phase, the discovery of the Greek city of Béziers (Béziers I)-one of the two Rhòde of the sources-, considerable advances on the Greek Agde (Agàthe) and on native sites. They shed light on why and how Greek colonization took root and, thereafter, on a different colonial functioning than envisaged before, in which the Dorian component-seemingly extended to the Rhòde of Iberia-preceded the Phocean one and counterbalanced the Marseille's role for several centuries.

International audience

Aujourd’hui, les nombreuses applications de l’Internet des Objets (IoT : Internet of Things) peuvent significativement améliorer la vie quotidienne des utilisateurs. Grâce à ces applications, il est possible de commander à distance les différents appareils de la maison, surveiller les signes vitaux d’un patient et alerter automatiquement son médecin en cas de problème. Cependant, les problèmes de sécurité et de protection de la vie privée empêchent les utilisateurs de faire pleinement confiance à ces applications, ce qui peut avoir pour effet de ralentir l’adoption globale de ces technologies et leur large déploiement. Pour résoudre ces problèmes de sécurité et de protection de la vie privée, plusieurs solutions ont été proposées. Cependant, plusieurs défis restent encore à relever pour permettre une large adoption de ces applications. L’approche centrée sur l’utilisateur semble être très pertinente pour relever un grand nombre de ces défis. Pour offrir une sécurité et une protection de la vie privée centrées sur l’utilisateur et permettre la prise en charge de nombreuses applications IoT, les travaux de cette thèse proposent d’adapter la mise en œuvre des mécanismes de sécurité et de protection de la vie privée en fonction du contexte de l’utilisateur. Dans un premier temps, cette thèse présente l’architecture CASPaaS (Context-Aware Security and Privacy as a Service). Cette architecture de sécurité et de protection de la vie privée sensibles au contexte pour l’IoT est basée sur l’approche ‘as a service’. Elle garantit l’adaptation dynamique et personnalisée des services de sécurité et de protection de la vie privée en fonction du contexte de l’utilisateur. Grâce à la conception ‘as a service’, cette architecture se caractérise par une grande flexibilité qui lui permet de prendre en charge de nombreuses applications IoT. Dans un second temps, cette thèse présente un système permettant de gérer la sécurité et la fiabilité de l’architecture CASPaaS elle-même. Ce système, appelé SETUCOM (SEcure and TrUstworthy COntext Management) gère la sécurité des données contextuelles échangées au sein de l’architecture CASPaaS ainsi que la confiance des sources de données. Ceci permet de pallier un grand nombre d’attaques pouvant conduire au disfonctionnement de notre architecture CASPaaS. Dans un troisième temps, cette thèse présente un nouveau système de gestion décentralisée des autorisations sensibles au contexte pour l’IoT. Ce système, basé également sur l’approche ‘as a service’, offre à l’utilisateur une gestion dynamique, décentralisée et simple des autorisations. Dans un dernier temps, cette thèse s’intéresse au déploiement du service CASPaaS au plus proche des utilisateurs en se basant sur une infrastructure de type Edge Computing. Dans ce contexte, nous proposons une nouvelle stratégie de placement dynamique de ce service. Cette stratégie fait appel à des techniques de l’intelligence artificielle afin de garantir un placement efficace tout en optimisant les différentes performances (réseau, service, capacités des nœuds Edge, processus de placement lui-même, etc.). Today, the Internet of Things (IoT) applications can significantly improve the daily life of users. Thanks to these applications, it is possible to control the various devices in the home remotely, monitor a patient's vital signs, and automatically alert his doctor when problems occur. However, security and privacy issues hinder the users to trust fully these applications, what may have as effect to slow down the overall adoption of these technologies and their widespread deployment. Several solutions have been proposed to address these security and privacy issues. Despite this, several challenges still need to be overcome to enable the global adoption of this type of application. The user-centric approach seems to be very relevant to address a large number of these challenges. To provide user-centric security and privacy protection and to enable the numerous IoT applications support, this thesis proposes to adapt the implementation of security and privacy protection mechanisms based on the user context and enable support for many IoT applications. First, this thesis presents the CASPaaS (Context-Aware Security and Privacy as a Service) architecture. This context-aware security and privacy architecture for IoT is based on the 'as a service' approach. It ensures the dynamic, personalized adaptation of security and privacy services based on the user's context. Thanks to the 'as a service' design, this architecture is characterized by high flexibility that will make it able to support multiple IoT applications. In a second step, this thesis presents a system to manage the security and reliability of the CASPaaS architecture itself. This system, called SETUCOM (SEcure and TrUstworthy COntext Management), addresses the security of the contextual data exchanged within the CASPaaS architecture and the trustworthiness of these data sources. This allows mitigating a large number of attacks that can lead to the malfunctioning of our CASPaaS architecture. In a third step, this thesis presents a new decentralized context-aware authorization management system for the IoT. This system, also based on the 'as a service' approach, offers the user dynamic, decentralized, and simple authorization management. Finally, this thesis focuses on deploying the CASPaaS service closer to the users based on an Edge Computing infrastructure. In this context, we propose a new dynamic placement strategy for this service. This strategy uses artificial intelligence techniques to ensure an efficient placement while optimizing the different performances (network, service, nodes capacities, placement process, etc.).

The works focus on the identification of industrial robots that belongs to the field of the identification of continuous-time inverse dynamic models in closed loop. First, a generic instrumental approach relevant for the identification of rigid industrial robots is proposed. The set of instruments is the inverse dynamic model constructed from simulated data calculated from the simulation of the direct dynamic model. This algorithm termed the IDIM-IV method validates the inverse and direct dynamic models simultaneously, improves the noise immunity of estimates with respect to corrupted data in the observation matrix and has a rapid convergence. This new approach is experimentally validated and compared with other standard methods. Then, a statistical test able to assess the validity of the set of instruments as well as the consistency of the least-squares estimates is presented. This test is based on the use of the Two-Stage-Least-Squares method and the regressed Durbin-Wu-Hausman test that are commonly used in econometrics. Finally, the perspectives that the IDIM-IV method can offer to the communities of robotics and automatic control are enlightened Ce manuscrit résume les travaux menés sur l'identification des robots industriels conduits à l'ONERA. Dans le premier chapitre, les méthodes usuelles de modélisation et d'identification appliquées aux robots industriels rigides sont présentées. Le deuxième chapitre introduit une nouvelle méthode d'identification basée sur l'utilisation des variables instrumentales alors que le troisième chapitre présente un test statistique capable de valider la construction des instruments. Enfin, dans un quatrième chapitre, les perspectives offerts par ces nouvelles méthodes sont introduites.

[Departement_IRSTEA]Territoires [TR1_IRSTEA]SEDYVIN [ADD1_IRSTEA]Dynamique et fonctionnement des écosystèmes; International audience; The importance of including a historical approach in ecological studies is now acknowledged, as history can help understand structure and functioning of present ecosystems but also assess conservation objectives. In temperate regions, several studies showed ecological differences according to forest temporal continuity, including soil physicochemical properties, vascular plant composition and traits. The definition of ancient and recent forests usually relies on a threshold date, provided by historical data. The best tools for studying the impact of forest continuity on forest plants are historical maps. Those maps should be ancient and precise enough. In France, those conditions are best met by the "Etat-Major" map. Indeed, unlike the Cassini map (18th century), small woodlots were not overlooked, the spatial resolution and precision is reasonable and that map was achieved at the national scale, which allows scientists to compare response patterns among regions in a homogeneous fashion. It has been digitized on about one third of France and is more and more used by scientists. However, is it that simple? Indeed, in lowlands, there are no ambiguities between land uses, and boundaries are very easy to identify. In a semi mountainous region, things are a bit trickier: relief may challenge the identification of land uses and boundaries; some land uses are in fact a combination of two land uses which are distinguished by a criterion that has not yet been clearly defined; in Mediterranean regions, long history of use of forests and particularly sylvo-pastoralism has probably smoothed the difference between ancient forests and recent forests developed on rangelands or pasturelands. We here question the relevance of the "Etat-Major" map as an absolute reference for studying the ecological effect of past land use and particularly in a Mediterranean and semi-mountainous context: the Natural Regional Park of Luberon. A hundred sites were sampled according to past land use (forest, "pâquis" and arable lands) on the "Etat-Major" map. In each plot, we also recorded past land use remnants: terrace, stone removal or none. Understory vegetation and soil physico-chemical properties were recorded and analysed according to past land use, and in situ past land use remnants. We showed that soil properties and understory vegetation significantly responded to past land use remnants, but not to past land use according to the "Etat-Major" map. This suggests that ancient maps may not be sufficient to detect the effect of past land use in the Mediterranean area.

International audience; In this paper, we present an optimized Machine Learning (ML) algorithm for predicting land suitability for crop (sorghum) production, given soil properties information. We set-up experiments using Parallel Random Forest (PRF), Linear Regression (LR), Linear Discriminant Analysis (LDA), KNN, Gaussian Naïve Bayesian (GNB) and Support Vector Machine (SVM). Experiments were evaluated using 10 cross fold validation. We observed that, parallel random forest had a better accuracy of 0.96 and time of execution of 1.7 sec. Agriculture is the main stream of food security. Kenya relies on agriculture to feed its population. Land evaluation gives potential of land use, in this case for crop production. In the Department of Soil Survey in Kenya Agriculture and Livestock Research Organization (KALRO) and other soil research organizations, land evaluation is done manually, is stressful, takes a long time and is prone to human errors. This research outcomes can save time and improve accuracy in land evaluation process. We can also be able to predict land suitability for crop production from soil properties information without intervention of a soil scientist expert. Therefore, agricultural stakeholders will be able to efficiently make informed decisions for optimal crop production and soil management.

National audience

Avec une production mondiale dépassant 750 millions de tonnes en 2017, le blé est considéré comme un aliment de base pour la population mondiale. Sa cartographie et sa surveillance pourraient alors se révéler être un outil très efficace pour atteindre les objectifs de développement durable (ODD2-Faim zéro). Au Liban, en vue d’assurer la sécurité alimentaire nationale, le blé reçoit un soutien financier et technique du gouvernement. Cependant, de nombreuses erreurs dans l'estimation de la superficie en blé à travers le pays sont dues principalement aux déclarations peu fiables des agriculteurs. De plus, le choix du système de culture basé sur le blé (en termes de type de rotation et de pratiques agricoles) peut avoir un impact considérable sur la productivité du système (protéines et rentabilité), l'efficacité d'utilisation des ressources (efficacité d'utilisation de l'eau et utilisation de l’azote), ainsi que sur le risque économique pour les agriculteurs. La présente étude menée dans la plaine de la Bekaa au Liban a pour objectif d'utiliser la télédétection et la modélisation de la croissance des cultures pour fournir aux décideurs politiques et aux utilisateurs finaux les informations dont ils ont besoin sur le blé d’hiver.La première partie évalue le potentiel des images optiques pour la cartographie du blé d'hiver précoce en permettant le transfert de connaissances d'une année à l'autre (2016 et 2017) Les résultats montrent que, lorsque l'approche développée est appliquée à la série chronologique Sentinel-2 de 2017, en utilisant les données de vérité au sol 2016, la précision globale atteint 87,0%, tandis que, lorsqu'elle est mise en œuvre avec les données de vérité au sol 2017, elle est de 82,6% en 2016. Les classifications pour distinguer le blé d'hiver de céréales similaires (orge et triticale) sont réalisées jusqu'à six semaines avant la récolte.La deuxième partie examine la capacité des images radar en bande C du nouveau satellite Sentinel-1 à surveiller la culture de blé d'hiver en identifiant les stades phénologiques économiquement importants qui ne peuvent pas être détectés en utilisant uniquement les indices issus de l’optique (Sentinel-2). Les résultats montrent que la polarisation VV (incidence de 32°-34°) et VH (incidence de 43°-45°) sont respectivement préférables pour estimer les stades de l'épiaison et de pâteux mou. De plus, le rapport VV/VH (incidence de 32°-34°) est préférable pour détecter la germination et la récolte.La troisième partie a pour objectif de vérifier si la rotation blé-féverole, nécessitant des intrants agricoles extensifs (eau et azote), a une performance nettement meilleure que la rotation intensive blé-blé en termes de productivité, d’utilisation rationnelle des ressources, et de minimisation du risque économique à l’échelle parcellaire. Le modèle de croissance des cultures «CropSyst» a été adopté après avoir été calibré et validé sur notre zone d'étude. Les résultats montrent qu’il n’existe pas un scénario de système de culture optimal permettant d’assurer une productivité élevée, de réduire le risque économique et d’obtenir une efficacité élevée de l’utilisation de l’eau et de l’azote. Cependant, le scenario de rotation blé-féverole sans fertilisation du blé semble être un meilleur substitut à la rotation blé-blé en termes de production de protéines. Pour un sol avec une faible capacité de rétention, la production de protéines est de 0,93 t/ha et de 0,8 t/ha respectivement pour les rotations blé-féverole et blé-blé. Pour un sol avec capacité de rétention élevée, la production de protéines est de 1,34 t/ha pour blé-féverole contre 1,17 t/ha pour blé-blé. De plus, ce système de culture blé-féverole pourrait générer un bénéfice net plus élevé que celui du blé-blé. Enfin, la rotation blé-pomme de terre peut générer des bénéfices nets très élevés (8640 US DOLL./ha et 12170 US DOLL./ha) mais avec une faible efficacité des intrants et un risque économique élevé With global production exceeding 750 million tons in 2017, wheat is considered a staple food for the world's population. Wheat mapping and monitoring could then be a very effective tool for achieving the Sustainable Development Goals (SDG2-Zero Hunger). In Lebanon, wheat receives technical and financial support, yet many errors occur in estimating the wheat acreage due to absence of reliable agricultural census and lack of wheat mapping using satellite images. In addition, identifying the best rotation type and agricultural practices leads to identify the most efficient wheat-based cropping system in terms of productivity (protein production and net profit), efficiency (water and nitrogen use), as well as the economic risk on the farmer. Thus, The aim of the current study, which is conducted in the Bekaa plain of Lebanon, is to utilize remote sensing technology and crop modelling for supporting policy makers and end-users in making strategic decisions regarding one of the most food security-driving crop in the Mediterranean (i.e. winter wheat).The first part of the thesis evaluates the potential of optical data for early winter wheat mapping by allowing the transfer of knowledge from one year to another (2016 and 2017 in this study). For its high spatial and temporal resolutions, Sentinel-2 data are employed. Results show that when the developed approach was applied on Sentinel-2 time series of 2017 in using 2016 ground truth data, the overall accuracy reaches 87.0%, whereas, when implemented using 2017 ground truth data, the overall accuracy is 82.6% on 2016 data. The outputs are executed up to six weeks before harvest, as well as distinguishing winter wheat from similar cereals (barley and triticale).The second part of the thesis examines the ability of the SAR (Synthetic Aperture Radar) C-band data of the new radar satellite (Sentinel-1) regarding its ability to monitor winter wheat crop by identifying the economically important phenological phases that cannot be detected relying solely on NDVI derived from optical satellite Sentinel-2. Results show that VV polarization at incidence angle of 32°-34° is best for predicting heading, VH polarization at incidence angle of 43°-45° for predicting soft dough, and the ratio VV/VH at incidence angle of 32°-34° for predicting germination and harvesting.The third part of the thesis is dedicated to test, in contrasted biophysical and management conditions, the hypothesis that promoting wheat-fava bean rotation leads to a significantly better productivity and resources use efficiency, as well as, reducing economic risk than the promoted intensive wheat-wheat and wheat-potato rotations. The cropping simulation model “CropSyst” is used after being calibrated and validated by using experimental data for different wheat-based rotations combining different soil, climate and management options. The results show that there is no particular optimal scenario that can simultaneously ensure high productivity, reduce economic risk, and achieve high wheat- water- and nitrogen-use efficiency. However, the wheat-fava bean rotation cultivated with no wheat fertilization appears to be a better substitute to the wheat-wheat rotation in terms of protein production in both (low and high) Water Holding Capacity (WHC) soils (0.93 t/ha versus 0.8 t/ha in low WHC and 1.34 t/ha versus 1.17 t/ha in high WHC). This cropping system could achieve a higher net profit, showing high resource-use efficiency and good economic sustainability. Moreover, a very high profit could only be attained with the wheat-potato rotation (8640 US DOLL./ha and 12170 US DOLL./ha), yet with low input-efficiency and high economic risk.

Two Radiation Belt Activity Indices, based on electron flux measurement >300 keV and >1.6 MeV, and one Solar Proton Event Alarm, based on proton flux measurement >75 MeV, are developed for post events analysis. Both indices and alarm are plotted over the last 30 days on the CRATERRE project web site and are daily updated.