Grid computing is a virtualized distributed computing environment aimed at enabling the sharing of geographically distributed resources. Grid resources have traditionally consisted of dedicated supercomputers, clusters, or storage units. With the present ubiquitous network connections and the growing computational and storage capabilities of modem everyday-use computers, more resources such as PCs, devices (e.g., PDAs and sensors), applications, and services are on grid networks. Grid is expected to evolve from a computing and data management facility to a pervasive, world-wide resource-sharing infrastructure. To fully utilize the wide range of grid resources, effective resource discovery mechanisms are required. However, resource discovery in a global-scale grid is challenging due to the considerable diversity, large number, dynamic behavior, and geographical distribution of the resources. The resource discovery technology required to achieve the ambitious global grid vision is still in its infancy, and existing applications have difficulties in achieving both rich searchability and good scalability. In this thesis, we investigate the resource discovery problem for open-networked global-scale grids. In particular, we propose a distributed semantics-based discovery framework. We show how this framework can be used to address the discovery problem in such grids and improve three aspects of performance: expressiveness, scalability, and efficiency. Expressiveness is the first characteristic that a grid resource-searching mechanism should have. Most existing search systems use simple keyword-based lookups, which limit the searchability of the system. Our framework improves search expressiveness from two directions: First, it uses a semantic metadata scheme to provide users with a rich and flexible representation mechanism, to enable effective descriptions of desired resource properties and query requirements. Second, we employ ontological domain knowledge to assist in the search process. The system is thus able to understand the semantics of query requests according to their meanings in a specific domain; this procedure helps the system to locate only semantically related results. The more expressive the resource description and query request, however, the more difficult it is to design a scalable and efficient search mechanism. We ensure scalability by reconfiguring the network with respect to shared ontologies. This reconfiguration partitions the large unorganized search space into multiple well-organized semantically related sub-spaces that we call semantic virtual organizations. Semantic virtual organizations help to discriminatively distribute resource information and queries to related nodes, thus reducing the search space and improving scalability. To further improve the efficiency of searching the virtual organizations, we propose two semantics-based resource-integrating and searching systems: GONID and OntoSum. These two systems address searching problems for applications based on different network topologies: structured and unstructured peer-to-peer overlay networks. Queries in the search systems are processed in a transparent way, so that users accessing the data can be insulated from the fact that the information is distributed across different sources and represented with different formats. In both systems, ontological knowledge is decomposed into different coarse-grained elements, and then these elements are indexed with different schemes to fit the requirements of different applications. Resource metadata reasoning, integrating, and searching are based on the index. A complex query can be evaluated by performing relational operations such as select, project, and join on combinations of the indexing elements. We evaluate the performance of our system with extensive simulation experiments, the results of which confirm the effectiveness of the design. In addition, we implement a prototype that incorporates our ontology-based virtual organization formation and semantics-based query mechanisms. Our deployment of the prototype verifies the system's feasibility and its applicability to real-world applications.

Grid computing is a virtualized distributed computing environment aimed at enabling the sharing of geographically distributed resources. Grid resources have traditionally consisted of dedicated supercomputers, clusters, or storage units. With the present ubiquitous network connections and the growing computational and storage capabilities of modem everyday-use computers, more resources such as PCs, devices (e.g., PDAs and sensors), applications, and services are on grid networks. Grid is expected to evolve from a computing and data management facility to a pervasive, world-wide resource-sharing infrastructure. To fully utilize the wide range of grid resources, effective resource discovery mechanisms are required. However, resource discovery in a global-scale grid is challenging due to the considerable diversity, large number, dynamic behavior, and geographical distribution of the resources. The resource discovery technology required to achieve the ambitious global grid vision is still in its infancy, and existing applications have difficulties in achieving both rich searchability and good scalability. In this thesis, we investigate the resource discovery problem for open-networked global-scale grids. In particular, we propose a distributed semantics-based discovery framework. We show how this framework can be used to address the discovery problem in such grids and improve three aspects of performance: expressiveness, scalability, and efficiency. Expressiveness is the first characteristic that a grid resource-searching mechanism should have. Most existing search systems use simple keyword-based lookups, which limit the searchability of the system. Our framework improves search expressiveness from two directions: First, it uses a semantic metadata scheme to provide users with a rich and flexible representation mechanism, to enable effective descriptions of desired resource properties and query requirements. Second, we employ ontological domain knowledge to assist in the search process. The system is thus able to understand the semantics of query requests according to their meanings in a specific domain; this procedure helps the system to locate only semantically related results. The more expressive the resource description and query request, however, the more difficult it is to design a scalable and efficient search mechanism. We ensure scalability by reconfiguring the network with respect to shared ontologies. This reconfiguration partitions the large unorganized search space into multiple well-organized semantically related sub-spaces that we call semantic virtual organizations. Semantic virtual organizations help to discriminatively distribute resource information and queries to related nodes, thus reducing the search space and improving scalability. To further improve the efficiency of searching the virtual organizations, we propose two semantics-based resource-integrating and searching systems: GONID and OntoSum. These two systems address searching problems for applications based on different network topologies: structured and unstructured peer-to-peer overlay networks. Queries in the search systems are processed in a transparent way, so that users accessing the data can be insulated from the fact that the information is distributed across different sources and represented with different formats. In both systems, ontological knowledge is decomposed into different coarse-grained elements, and then these elements are indexed with different schemes to fit the requirements of different applications. Resource metadata reasoning, integrating, and searching are based on the index. A complex query can be evaluated by performing relational operations such as select, project, and join on combinations of the indexing elements. We evaluate the performance of our system with extensive simulation experiments, the results of which confirm the effectiveness of the design. In addition, we implement a prototype that incorporates our ontology-based virtual organization formation and semantics-based query mechanisms. Our deployment of the prototype verifies the system's feasibility and its applicability to real-world applications.

Grid computing is a virtualized distributed computing environment aimed at enabling the sharing of geographically distributed resources. Grid resources have traditionally consisted of dedicated supercomputers, clusters, or storage units. With the present ubiquitous network connections and the growing computational and storage capabilities of modem everyday-use computers, more resources such as PCs, devices (e.g., PDAs and sensors), applications, and services are on grid networks. Grid is expected to evolve from a computing and data management facility to a pervasive, world-wide resource-sharing infrastructure. To fully utilize the wide range of grid resources, effective resource discovery mechanisms are required. However, resource discovery in a global-scale grid is challenging due to the considerable diversity, large number, dynamic behavior, and geographical distribution of the resources. The resource discovery technology required to achieve the ambitious global grid vision is still in its infancy, and existing applications have difficulties in achieving both rich searchability and good scalability. In this thesis, we investigate the resource discovery problem for open-networked global-scale grids. In particular, we propose a distributed semantics-based discovery framework. We show how this framework can be used to address the discovery problem in such grids and improve three aspects of performance: expressiveness, scalability, and efficiency. Expressiveness is the first characteristic that a grid resource-searching mechanism should have. Most existing search systems use simple keyword-based lookups, which limit the searchability of the system. Our framework improves search expressiveness from two directions: First, it uses a semantic metadata scheme to provide users with a rich and flexible representation mechanism, to enable effective descriptions of desired resource properties and query requirements. Second, we employ ontological domain knowledge to assist in the search process. The system is thus able to understand the semantics of query requests according to their meanings in a specific domain; this procedure helps the system to locate only semantically related results. The more expressive the resource description and query request, however, the more difficult it is to design a scalable and efficient search mechanism. We ensure scalability by reconfiguring the network with respect to shared ontologies. This reconfiguration partitions the large unorganized search space into multiple well-organized semantically related sub-spaces that we call semantic virtual organizations. Semantic virtual organizations help to discriminatively distribute resource information and queries to related nodes, thus reducing the search space and improving scalability. To further improve the efficiency of searching the virtual organizations, we propose two semantics-based resource-integrating and searching systems: GONID and OntoSum. These two systems address searching problems for applications based on different network topologies: structured and unstructured peer-to-peer overlay networks. Queries in the search systems are processed in a transparent way, so that users accessing the data can be insulated from the fact that the information is distributed across different sources and represented with different formats. In both systems, ontological knowledge is decomposed into different coarse-grained elements, and then these elements are indexed with different schemes to fit the requirements of different applications. Resource metadata reasoning, integrating, and searching are based on the index. A complex query can be evaluated by performing relational operations such as select, project, and join on combinations of the indexing elements. We evaluate the performance of our system with extensive simulation experiments, the results of which confirm the effectiveness of the design. In addition, we implement a prototype that incorporates our ontology-based virtual organization formation and semantics-based query mechanisms. Our deployment of the prototype verifies the system's feasibility and its applicability to real-world applications. / Science, Faculty of / Computer Science, Department of / Graduate

Perfluoroalkylated substances (PFAS) are persistent contaminants that are widely distributed in the global environment. Despite the fact that these chemicals have been manufactured and used for over 50 years, there has been little scientific and regulatory interest until very recently. An important research priority over the past decade has been to gain a better understanding of the mechanisms and pathways explaining the presence of these compounds in remote regions. One explanation is related to the use and release of volatile precursor compounds which undergo atmospheric transport and are also susceptible to degradation to PFAS through gas phase reactions with radical species. The main purpose of this doctoral thesis was to investigate an alternative explanation, namely the long-range transport (LRT) of PFAS themselves, which have been released into the environment in substantial quantities during manufacturing and product use. Papers I – III explore the LRT potential of perfluorocarboxylic acids and perfluorocarboxylates and demonstrate that both oceanic and atmospheric transport are efficient pathways of dispersion from source to remote regions of the Northern Hemisphere. Oceanic transport of perfluorooctane sulfonate (PFOS) was shown to be an important process in Paper IV as well. The role of precursor transport and degradation to PFOS was also examined in this paper. The most interesting aspect of the fate and transport of PFOS precursors is the rapid response in ambient concentrations exhibited by these compounds in the model simulations following production phase-out. Since precursor compounds are known to degrade to PFOS in vivo, the modeling results demonstrate that this exposure pathway is a plausible explanation for the declining trends in PFOS concentrations reported for marine mammals in some remote environments.

One of the recognized weaknesses of land surface models as used in weather and climate models is the assumption of constant soil thickness because of the lack of global estimates of bedrock depth. Using a 30-arc-s global dataset for the thickness of relatively porous, unconsolidated sediments over bedrock, spatial variation in soil thickness is included here in version 4.5 of the Community Land Model (CLM4.5). The number of soil layers for each grid cell is determined from the average soil depth for each 0.9 degrees latitude x 1.25 degrees longitude grid cell. The greatest changes in the simulation with variable soil thickness are to baseflow, with the annual minimum generally occurring earlier. Smaller changes are seen in latent heat flux and surface runoff primarily as a result of an increase in the annual cycle amplitude. These changes are related to soil moisture changes that are most substantial in locations with shallow bedrock. Total water storage (TWS) anomalies are not strongly affected over most river basins since most basins contain mostly deep soils, but TWS anomalies are substantially different for a river basin with more mountainous terrain. Additionally, the annual cycle in soil temperature is partially affected by including realistic soil thicknesses resulting from changes in the vertical profile of heat capacity and thermal conductivity. However, the largest changes to soil temperature are introduced by the soil moisture changes in the variable soil thickness simulation. This implementation of variable soil thickness represents a step forward in land surface model development.

La constante de temps du débit de base (τ) représente le temps moyen pour que l'eau souterraine arrive à la rivière depuis la zone de recharge dans un bassin donné. C’est un élément clé pour simuler le débit de base dans les modèles simples des eaux souterraines, tels qu’ORCHIDEE. τ a été estimée à l’échelle globale à partir d’une solution de l’équation de Boussinesq pour les aquifères libres en pente. τ dépend de la porosité efficace, de la transmissivité, de la pente de l'aquifère et de la densité de drainage (δ). Calculées à partir de bases de données globales, les valeurs de τ sont surestimées par rapport à celles obtenues par analyse des courbes de récession. Une analyse de sensibilité a montré que la transmissivité et δ sont les principales sources d’incertitude de τ. L’extraction d’un nouveau réseau de drainage, qui dépend de la lithologie, du climat, de la pente et des δ observées, a permis d’obtenir des δ conformes aux valeurs observées aux échelles régionales et à la variabilité spatiale. L’utilisation de ces nouvelles δ et la combinaison de deux jeux de données de conductivité hydraulique pour le sol et l’aquifère a réduit τ de deux ordres de grandeur, mais les valeurs calculées restent surestimées. L’utilisation de τ dans le modèle de surface ORCHIDEE a montré une forte sensibilité du débit simulé à l’augmentation de τ, qui dégrade les débits simulés par rapport aux observations. Cette méthodologie nécessite des valeurs plus adaptées de transmissivité et porosité efficace par rapport aux jeux de données globaux actuellement disponibles pour obtenir des valeurs de τ plus proches de celles attendues et qui permettent de reproduire les débits observés. / The base flow time constant (τ) represents the mean amount of time the groundwater takes to reach the stream from the recharge zone in a given watershed. τ is a key element to simulate base flow in simple groundwater models as ORCHIDEE. τ was estimated at global scale based on a solution of the Boussinesq equation for unconfined sloping aquifers. τ depends on the effective porosity, transmissivity, aquifer slope, and drainage density (δ). When estimated from global available datasets, τ results are overestimated when compared to recession analysis results. A sensitivity analysis showed that transmissivity and δ are the main uncertainty sources of τ. A river network extraction based on lithology, climate, slope, and observed δ allowed to obtain δ values close to reference data and spatially variable at regional scale. The use of a new δ and the combination of two hydraulic conductivity datasets of soil and aquifer reduced τ of two orders of magnitude, however the values remained overestimated. The use of τ in ORCHIDEE land surface model showed a strong sensitivity of the river discharge buffer effect to τ, which worsen simulated river discharge when compared to observations. This methodology needs more adequate porosity and transmissivity values when compared to global available datasets that will result in close results to observed river discharge.

Le bassin versant du fleuve Niger est directement influencé par les fluctuations de la mousson africaine, qui impactent les ressources en eau et entraînent des évènements extrêmes tels que des inondations ou des sécheresses. En retour, les forts taux d'évaporation observés dans le Delta intérieur du Niger, large région annuellement inondée, impactent le climat, au moins à l'échelle régionale. Une meilleure compréhension des processus hydrodynamiques de ce bassin ne peut cependant être obtenue sans un réseau d'observations ayant une couverture spatiale et temporelle suffisante. La mission SWOT fournira des cartes 2D de hauteurs et pente des eaux de surface avec une résolution encore jamais atteinte en altimétrie (50 à 100 mètres). Cette thèse s'inscrit dans le cadre de la phase de préparation à la mission SWOT et se propose d'offrir des perspectives d'utilisation de ces données satellites pour l'amélioration des modèles d'hydrologie globale. Dans un premier temps, le modèle hydrologique du CNRM, ISBA-TRIP, incluant un schéma d'inondations et un réservoir simple d'aquifères ajouté durant cette thèse est évalué sur le bassin du Niger à l'aide de multiples observations in-situ et satellites. L'étude montre que le modèle simule de façon cohérente l'évolution des eaux de surface, des zones inondées, et les anomalies de stock d'eau sur le bassin. Ensuite, un schéma d'assimilation de données est mis en place afin d'optimiser un des paramètres clés en hydrologie, le coefficient de Manning. Ce coefficient, décrivant la propriété du sol à 'retenir' les flux d'eau, influence fortement la dynamique des eaux de surface, et notamment les hauteurs d'eau et le débit. L'assimilation des données SWOT est appliquée dans le cadre d'une expérience jumelle, qui consiste à considérer une simulation de référence, appelée 'vérité', de laquelle sont issues les observations virtuelles de hauteur d'eau SWOT. L'étude montre que l'assimilation des hauteurs d'eau SWOT permet l'optimisation du coefficient de Manning, distribué spatialement, malgré l'hypothèse d'équifinalité. Les hauteurs d'eau et les débits sont considérablement améliorés, et on obtient une meilleure simulation des anomalies de stocks d'eau sur le bassin ainsi que des zones inondées sur le Delta intérieur du Niger (occurrence, intensité). Enfin, le potentiel des données SWOT pour améliorer les prévisions hydrologiques sur des périodes plus longues que celle de la phase d'assimilation est mis en évidence. / The hydrologic and hydrodynamic processes of the Niger basin are largely influenced by the West African monsoon variabilty. In the last 3 decades these variations have resulted in an increase of extreme events such as floods and droughts. Retrospectively, the climate might be impacted by the evaporation fluxes from the inner Delta flooded region, at least regionally. A better understanding of the Niger basin water cycle is a crucial issue for water resources management but requires observation datasets with a large spatial and temporal coverage. The SWOT satellite mission will provide 2D global maps of water level and slope at an unprecedented resolution (50 to 100 meters). Within the framework of the preparation of the SWOT mission, this thesis aims at proposing a SWOT data assimilation strategy for the improvement of global scale hydrological models. First, the ISBA-TRIP hydrological model from CNRM is evaluated over the Niger basin. This model includes an inundation scheme and simple aquifer reservoir. The model diagnostics are compared to an extensive set of in-situ and satellite observations. According to its relative simple physics, the model is able to simulate in a realistic manner, the continental water dynamics : discharge, water levels, floods, total water storage variations. Sensitivity tests are also performed to determine the most sensitve ISBA-TRIP parameters. Among them, the Manning coefficient has a key role in the flow dynamics but its estimation is difficult and usually based on geomorphologic relationships. The second part of this work consists in setting up a SWOT data assimilation strategy for the optimization of the ISBA-TRIP parameters. Since the SWOT observations are not available yet and also to assess the skills of the assimilation method, the study is carried out in the framework of an Observing System Simulation Experiment (OSSE). The corrected parameter is the Manning coefficient, spatially distributed over the river. The assimilation allows a good improvement of the relative bias of discharge and water level over the river. The Manning coefficient is also globally improved and tends to an optimal value. Moreover, the water storage anomalies and flooded fraction are also better simulated. Finally, the study shows that the method is useful for hydrological forecasting over longer time periods than those of the calibration.

The Milazzo peninsula, North-Eastern Sicily, Italy, is the implantation site of potentially polluting anthropogenic activities. Among them, a refinery and a thermoelectric plant. An epidemiological study has shown a decrease of human health quality in the area, but there was no further research as to the reasons of this observation. The aim of this work is to better understand how the Milazzo peninsula's anthropogenic emissions impact the local human health and ecosystems. The Life Cycle Assessment calculating approach was chosen in order to study the consequences of the emissions resulting from the following local anthropogenic activities: the thermoelectric plant, the refinery, the road traffic, the illegal combustion of domestic waste. The Impact 2002+ and the ReCiPe 2008 global-scale models were chosen, as well as a site-specific approach. In order to conduct this site-specific calculation, Milazzo soils and seawater samples were gathered and the specific local parameters were deduced from their analyses. The global-scale results pointed out the waste combustion emissions and the thermoelectric plant as the main causes of the human health quality decrease, while the thermoelectric plant was pointed out as the anthropogenic activity causing the most environmental damages. The site-specific study applied to the soil and seawater ecotoxicity confirmed the important role of the thermoelectric plant's emissions in the damages upon the environment, especially the seawater. Moreover, the site-specific study also corrected the soil ecotoxic impact scores, showing the importance of both thermoelectric plant and illegal waste combustion contributions to the damaging of the environment. As a conclusion, LCA study as a whole enlightens and explains the previous observations of human health and ecosystem damages in the site of study while the site-specific study proved relevant in the peculiar area of Milazzo.

Le plancton constitue l’essentiel de la biomasse pélagique et est un acteur majeur des cycles biogéochimiques globaux qui régulent le système Terre. Il comprend l'ensemble des organismes portés par les courants, des bactéries aux méduses géantes. La communauté n'est que très rarement étudiée dans son ensemble mais plutôt par fraction. L’expédition Tara Oceans constitue le premier effort de collecte simultané de toutes les classes de taille de plancton à l’échelle de l’océan global. Pour démontrer la faisabilité de cette approche à grande échelle, des échantillons hebdomadaires de plancton, depuis les bactéries jusqu’au macroplancton gélatineux, ont d’abord été analysés en combinant plusieurs instruments d’imagerie sur une période de 10 mois, en un site de référence (point B) dans la rade de Villefranche sur mer. L’imagerie nous a permis de comparer 1) l’information fonctionnelle définie comme l’agrégation de taxons en 18 Groupes Ecologiques de Plancton (GEP), et 2) la structure en taille des communautés échantillonnées sur un intervalle de taille de 6 ordres de grandeur (0.1 µm à 10000 µm). La communauté planctonique au point B évolue en une succession écologique complexe impliquant tous les groupes planctoniques, depuis les bactéries jusqu’aux prédateurs gélatineux du macroplancton. Des évènements impulsifs, tels que des coups de vent, déclenchent des réorganisations de la communauté par un jeu d’interactions entre des contrôles « bottom-up » et « top-down ». Toutefois, le biovolume planctonique total ne varie que d’un seul ordre de grandeur au cours de la période échantillonnée. De même, la structure en taille des communautés planctoniques totales ne varie pas significativement au cours du temps. La stabilité du biovolume total et de la structure en taille suggère que des mécanismes structurant et de compensation forts maintiennent les communautés planctoniques dans un intervalle de biomasse restreint. Le couplage entre données de taille et de taxonomie révèle une réorganisation du réseau trophique entre l’été et l’hiver. En hiver, Le réseau trophique microplancton-zooplancton est dominé par la fonction de broutage. En été, le réseau trophique microplancton-zooplancton est dominé par la fonction de prédation (chaetognathes et gélatineux carnivores). En été, ce réseau trophique s’organise en deux chaines trophiques parallèles et distinctes discriminées par des relations de taille entre proies et prédateurs. Cette réorganisation souligne le rôle clef du zooplancton et de la prédation dans la structuration des communautés planctoniques. Parallèlement à cette analyse temporelle en un point fixe, nous avons montré l’existence de types caractéristiques de communautés zooplanctoniques, associés à des conditions environnementales distinctes, à partir des échantillons de l’expédition Tara Oceans, à l’échelle globale. En utilisant la même méthodologie que pour l’analyse de la dynamique temporelle, nous avons identifié trois types de communautés mésozooplanctoniques à l’échelle globale selon le type d’environement: 1) des communautés associées aux environnements productifs (upwellings côtiers et équatoriaux), 2) des communautés associées aux zones de minimum d’oxygène (OMZs, « Oxygen Minimum Zones »), et 3) des communautés associées aux gyres océaniques oligotrophes. Ce travail constitue une première typologie des communautés zooplanctoniques, structurées en taille et GEP, à l’échelle globale. Il sera complété dans le futur par l’intégration de données issus des autres compartiments planctoniques, et de données d’export vertical de matière organique particulaire pour affiner les estimations des relations qui existent entre phytoplancton, zooplancton et flux biogéochimiques. / Plankton constitutes the bulk of pelagic biomass and plays a major role in the global biogeochemical cycles that regulate the earth system. It encompasses all the organisms that drift with the water masses movements, from bacteria to giant medusae. Studies of the entire community are scarce, and plankton has been traditionally studied by fractions. The Tara Oceans expedition is the first attempt to simultaneously collect plankton in every size classes at the global scale. To demonstrate the feasibility of this approach, samples of plankton from bacteria to gelatinous macroplankton were collected weekly over ten months at a reference site (point B), in Villefranche Bay, northwestern Mediterranean, and analyzed using imaging techniques. Imaging enabled us to compare 1) the functional taxonomic information as derived from the analysis of 18 Plankton Ecological Groups (PEGs), and 2) the size structure of the same planktonic community over 6 orders of magnitude in size. The plankton dynamics at point B are driven by a complex succession process involving all plankton groups, from bacteria to macroplanktonic gelatinous predators. Environmental impulsive events such as wind events trigger sharp community level reorganizations via interplay of bottom-up controls followed by top-down controls. However, the total biovolume of the planktonic community varies within only one order of magnitude over the period studied. In addition, the size structure of the entire community does not vary significantly over time. The total biovolume and size structure stability suggest that strong and compensative mechanisms drive community dynamics within a narrow range of biomass variation. The use of both taxonomic and size structured data reveals a reorganization of the food web between winter and summer. In winter and spring the microplanktoniczooplanktonic food web is shaped by the grazing function. In summer, it is shaped by the predation function (chaetognaths and gelatinous predators). In summer, the food web self organizes in two distinct food chains discriminated by size relations between predators and preys. This reorganization underlines the key role of zooplankton and predation in structuring planktonic communities. In parallel to this temporal dynamics study, we used the Tara Oceans expedition samples to study the global scale distribution of mesozooplankton. We showed that characteristic mesozooplanktonic communities were associated with distinct environmental conditions, at the global scale. Using a similar methodology as for the temporal study we found that three different mesozooplanktonic communities were associated with 1) productive environments (e.g. upwellings), 2) Oxygen Minimum Zones, and 3) Oligotrophic oceanic gyres. This work is the first typology of mesozooplanktonic communities at the global scale. It will be further developed in the future by the integration of other planktonic compartments and particulate organic matter fluxes data, to improve our knowledge on the relations between phytoplankton, zooplankton and particulate organic matter fluxes.

La péninsule de Milazzo, au Nord-Est de la Sicile, en Italie, est le site d'implantation d'activités anthropiques potentiellement polluantes. Parmi elles, une raffinerie et une centrale thermoélectrique. Une étude épidémiologique a montré une diminution de la qualité de la santé humaine dans la zone, mais aucune recherche n'a approfondi les raisons de cette observation. Le but de ce travail est de mieux comprendre comment les émissions des activités anthropiques de la péninsule de Milazzo impactent la santé humaine à l'échelle locale ainsi que les écosystèmes. L'approche par l'Analyse du Cycle de Vie a été choisie afin d'étudier les conséquences des émissions provenant des activités anthropiques suivantes: la centrale thermoélectrique, la raffinerie, la circulation automobile, la combustion sauvage de déchets domestiques. Les modèles à l'échelle globale Impact2002+ et ReCiPe2008 furent choisis, ainsi qu'une approche site-spécifique. Afin de réaliser ces calculs à l'échelle site-spécifique, des échantillons de sols et d'eau marine provenant de Milazzo ont été rassemblés et les paramètres locaux spécifiques ont été déduits de leurs analyses. Les résultats à l'échelle globale on montré que la combustion des déchets ménagers et que la centrale thermoélectrique sont les activités anthropiques causant le plus de dommages environnementaux. L'étude site-spécifique appliquée à l'écotoxicité du sol et de l'eau de mer a confirmé l'importance du rôle des émissions de la centrale thermoélectrique dans les dommages sur l'environnement, en particulier sur l'eau de mer. De plus, l'étude site-spécifique a permis de corriger les scores d'impact du sol, montrant l'importance à la fois de la centrale thermoélectrique et de la combustion sauvage de déchets ménagers dans les dommages sur l'environnement. En conclusion, l'étude ACV a globalement montré et expliqué les observations précédentes concernant les dommages à la santé humaine et à l'environnement sur le site d'étude alors que l'étude site-spécifique s'est montrée pertinente pour la zone particulière de Milazzo. / The Milazzo peninsula, North-Eastern Sicily, Italy, is the implantation site of potentially polluting anthropogenic activities. Among them, a refinery and a thermoelectric plant. An epidemiological study has shown a decrease of human health quality in the area, but there was no further research as to the reasons of this observation. The aim of this work is to better understand how the Milazzo peninsula's anthropogenic emissions impact the local human health and ecosystems. The Life Cycle Assessment calculating approach was chosen in order to study the consequences of the emissions resulting from the following local anthropogenic activities: the thermoelectric plant, the refinery, the road traffic, the illegal combustion of domestic waste. The Impact 2002+ and the ReCiPe 2008 global-scale models were chosen, as well as a site-specific approach. In order to conduct this site-specific calculation, Milazzo soils and seawater samples were gathered and the specific local parameters were deduced from their analyses. The global-scale results pointed out the waste combustion emissions and the thermoelectric plant as the main causes of the human health quality decrease, while the thermoelectric plant was pointed out as the anthropogenic activity causing the most environmental damages. The site-specific study applied to the soil and seawater ecotoxicity confirmed the important role of the thermoelectric plant's emissions in the damages upon the environment, especially the seawater. Moreover, the site-specific study also corrected the soil ecotoxic impact scores, showing the importance of both thermoelectric plant and illegal waste combustion contributions to the damaging of the environment. As a conclusion, LCA study as a whole enlightens and explains the previous observations of human health and ecosystem damages in the site of study while the site-specific study proved relevant in the peculiar area of Milazzo.