L’objectif général de cette thèse est de contribuer à l’avancement de la connaissance de la variabilité du Courant Nord Méditerranéen (CN) et de ses interactions avec la dynamique côtière, en s’appuyant principalement sur une configuration numérique réaliste à haute résolution de la façade méditerranéenne française, basée sur le modèle de circulation océanique NEMO et nommée GLAZUR64.La validation de cette configuration avec toutes les observations disponibles sur la période d'étude (CTD, gliders, ARGO, radar HF, ADCP, altimétrie et SST satellite) a permis d'évaluer le réalisme des simulations et leur paramétrisation, et de montrer l'apport de la haute-résolution par rapport aux configurations de bassin au 1/12° utilisées aux frontières de GLAZUR64.Enfin, l'utilisation d'un forçage océanique opérationnel a permis d'utiliser une simulation en complément des données d'une campagne en mer, pour l'étude ciblée d'un tourbillon anticyclonique associé à un méandre du CN au large de Toulon, en avril 2011 [Guihou et al., 2013]. / The main objective of this work is to improve our knowledge of the Northern Mediterrranean Current (NC) variability and its interactions with coastal dynamics, using high-resolution modelling of the North-Western Mediterranean Sea, based on the NEMO code and nammed GLAZUR64.The configuration has been validated using all available observations for the period of study (CTD, gliders, ARGO, HF radar, ADCP, satellite altimetry and SST), in order to assess the realism of the simulations and their parameterisations. The contribution of the high-resolution for the simulation of the mesoscale dynamics, compared to 1/12° configurations used at the boundaries is shown.Finally, the use of an operationnal oceanic boundary forcing allowed us to use a combined data-model approach to study the generation and advection of an anticyclonic eddy trapped at the coast and associated to a NC meander, flowing off Toulon in April 2011 [Guihou et al., 2013].

Accurate fishery-independent methods for assessing the abundance of marine fish stocks are important tools for proper pelagic fisheries management. The review and improvement of standardized sampling methods used in fisheries-independent sampling, and the establishment of improved models of abundance and population dynamics utilizing novel statistical techniques for fisheries management will further improve our understanding of the way in which marine fish stocks vary spatially and temporally. One of the most important and longest running surveys of fisheries-independent data in the Southeastern United States is the spring ichthyoplankton survey conducted through the Southeast Area Mapping and Assessment Program (SEAMAP). These surveys of highly migratory pelagic fish larvae and eggs in the exclusive economic zone of the USA in the Gulf of Mexico have been conducted annually during spring spawning periods (March to June) since 1982 by the National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS). A primary focus of this survey is collecting the data used in a larval abundance index of bluefin tuna (Thunnus thynnus) for "tuning" population estimates for the western Atlantic stock of this species using virtual population analysis (VPA). As with many other forms of fishery independent data, there is statistical overdispersion of larval fish and high variability in catch rates related to non-static environmental conditions. This variability in catch rates, along with zero-inflated estimates of abundance, i.e. characterized by high numbers of stations with low or zero catches, present problems in deriving accurate and reliable stock assessments for western Atlantic bluefin tuna. The research presented here was conducted in an effort to improve the decision support system used in managing bluefin tuna in the western Atlantic, by improving catch rates of their larvae during spring ichthyoplankton cruises. A major focus of which was to overcome challenges related to reducing the variance of the larval index of abundance. The research presented here, explored the influence of sampling design and oceanographic conditions on an assemblage of commercially important taxa within the Gulf of Mexico. Such studies are necessary to evaluate differences in habitat utilization within a specific complex or among assemblages of closely related taxa. The spatial and temporal patterns of larval bluefin tuna (Thunnus thynnus) were analyzed in relation to mesoscale patterns of circulation, types of sampling gear used during fishery independent surveys, and in the context of common oceanographic variables associated with the spawning habitat of this iconic pelagic predator. Collaterally the relationship between the above factors on spatial and temporal patterns of the larvae of seven congeneric species of marine fish, whose larvae commonly occur with bluefin tuna, during the spring spawning season in the Gulf of Mexico (Auxis spp., Euthynnus alleteratus Corphaena spp., Katsuwonus pelamis, other Thunnus spp.: Thunnus albacares and Thunnus atlanticus, Istiophoridae, and Xiphias gladius) was investigated. The historical distribution of abundance in these eight taxa of larval epipelagic fish were analyzed in the context of mesoscale and large scale circulation features in the Gulf of Mexico between 1994 and 2008. Characterization of these features was conducted using concurrent, synoptic satellite altimeter, sea surface temperature (SST), and ocean color (OC) observations. Larval abundance among mesocale features was examined for the eight pelagic fish taxa using permutational multivariate analysis of variance (PERMANOVA). All but one taxon, T. thynnus, showed differences in occurrence across features. Additionally only two taxa (X. gladius and Istiophoridae) did not show highest abundances in "Common Waters" of the Gulf of Mexico. The abundance of the eight taxa, when examined in aggregate or as an assemblage, differed significantly between circulation features. However, distributions of individual taxa among circulation features differed for only the other Thunnus spp. taxa, which incorporates Thunnus albacares and atlanticus. The results from this study indicated that for most of the individual taxa studied, and Thunnus thynnus in particular the distribution of spawning habitat in the Gulf of Mexico is not associated with specific patterns in circulation. To analyze the impact of sampling gear on larval epipelagic abundance, the collection methods of the eight taxa during spring surveys, including a new net sampling methodology tested in surveys between 2009 and 2011, were compared. This new sampling methodology consisted of a 1x2 m frame fitted with a 0.505 mm mesh net, towed in a yo-yo fashion between the surface and 10m depth, referred to as the S-10 net. Sampling effectiveness between gears was compared by examining the abundance and length of the eight taxa of larval fish. A PERMANOVA of net type and time of sampling (day/night) on abundance and mean body length (BL) indicated that net type was a significant factor in assessing abundance and BL for all taxa. Highest mean assemblage and individual taxa abundance for seven of the eight taxa were in S-10 samples. Depth discrete sampling of the upper 50 m indicated that highest abundances for all scombrid taxa were found in the upper 30 m and provided justification for the improved sampling efficiency of the S-10 net. These results indicate that distributions of these epipelagic fish are not truly neustonic and that sampling effectiveness for them strongly depends on depth range fished by the net. Finally, the results of two research cruises using the new sampling methodology (S-10 net) during the spring of 2010 and 2011 were investigated to explore larval fish assemblages. Canonical analysis of principal coordinates (CAP) was used to evaluate the horizontal distribution of eight taxa of epipelagic larval fish and to relate these distributions to a total of eight variables observed through in-situ and remotely obtained data. Larval assemblages were established for the CAP via unweighted pair group method with arithmetic mean (UPGMA) hierarchical clustering utilizing similarity profile analysis (SIMPROF) tests for stopping rules. Further refinement to five common assemblage groups was obtained using indicator species values (INDVALS) and percent composition of taxa abundance within groups. Total model accuracy was 54%, with highest classification success for the assemblage group characterized by T. thynnus and the other Thunnus spp. (64%), and lowest classification success for negative stations (22%). Changes in the assemblage of these larval fish were partially explained by the seasonal progression in day-length, and related changes in sea surface temperature. Depth, chlorophyll a concentration, salinity, and optical clarity were also important. Five assemblages of larval fish were identified, characterized by differences in the relative abundance of the five taxa of scombrids, and the Coryphaena spp. taxa, and indicated a high degree of shared spawning habitat. The differences in habitat utilization by these taxa is highly influenced by seasonal changes in sea surface temperature, and large scale differences in depth and water masses in the Gulf of Mexico. This results in a gradual shift from groups with smaller and more abundant coastally influenced taxa with protracted spawning seasons to the larger sub-tropical, more pelagic species across a large expanse of shared spawning habitat. This dissertation research provides a context to the historical assessment of abundance, evidence for the association of specific taxa with particular oceanographic conditions, and improved assessment capabilities for epipelagic larval fish. As this research shows the spawning habitat of these taxa in the Gulf of Mexico are largely shared, and the influence of environmental variables only partly addressed the spatial and temporal variance attributed to larval abundance in the Gulf of Mexico. The improved methods used in this study will be useful to researchers studying the factors impacting larval recruitment and survival of highly migratory species, and comprehensive ecosystem based resource management. The results will improve fisheries-independent sampling and management for highly migratory species, and provides new methods for evaluating larval fish assemblages within the framework of oceanographic conditions for large marine ecosystems.

The temporal and spatial variability in wind speed and direction was investigated in a study of the mesoscale wind fields in the Canberra region. The statistical description of the winds was based on twelve months of three-hourly data at seventeen sites obtained in a joint program carried out by the Division of Land Use Research, CSIRO, the National Capital Development Commission, and the Bureau of Meteorology. The statistical analysis proceeded in stages. The first two stages were concerned with the determination and examination of averages and measures of dipersion. Information on the temporal variability in regional wind, defined as the average of the winds at the seventeen collection sites, provided the first insight into the important determinants of winds in the region. The data were then categorized on the basis of the information thus obtained, and the averages over time for each site were analysed in each category. The variation between sites revealed the extent of the spatial variability in the region. For each category, for each site, there were perturbations around the average state, and in the last stage of the study, the analysis examined how the perturbations were related across sites using correlation coefficients. Generalized Procrustes Analysis was used, followed by the extensive use of cluster analysis. Linear modelling techniques were used at all stages of the study, not only for wind speed, but also for wind direction which is an angular variate and thus required different modelling procedures. The models related the variables of interest to terrain features such as position, elevation and surface roughness. These models allowed an informed judgement to be made on the likelihood of accurately estimating the winds at other locations in the region using interpolation techniques.

<p>The boundary layer is the part of the atmosphere that is in direct contact with the ground via turbulent motion. At mid-latitudes the boundary layer is usually one or a few kilometers deep, while in the Arctic it is much more shallow, typically a few hundred meters or less. The reason is that here the absolute temperature increases in the lowest kilometer, making the boundary layer semi-permanently stably stratified. The exchange of heat, momentum and tracers between the atmosphere, ocean and ground under stable stratification is discussed from an observational, modeling and climate-change point of view. A compilation of six observational datasets, ordered by the Richardson number (rather than the widely used Monin-Obukhov length) reveals new information about turbulence in the very stably stratified regime. An essentially new turbulence closure model, based on the total turbulent energy concept and these observational datasets, is developed and tested against large-eddy simulations with promising results. The role of mesoscale motion in the exchange between the atmosphere and surface is investigated both for observations and in idealized model simulations. Finally, it is found that the stably stratified boundary layer is more sensitive to external surface forcing than its neutral and convective counterparts. It is speculated that this could be part of the explanation for the observed Arctic amplification of climate change.</p>

The influence of mesoscale ocean eddies on near-surface ocean temperature, surface stress and phytoplankton communities is investigated by collocating numerous satellite measurements along with vertical profiles of oceanic temperature and salinity to the interiors of eddies identified and tracked in altimetric sea surface height maps. The surface currents associated with mesoscale ocean eddies impart a curl of the surface stress from the relative motion between surface air and water. This stress curl has a polarity opposite that of the vorticity of the eddy, thus attenuating the eddies by generating Ekman upwelling in the cores of anticyclones and downwelling in the cores of cyclones. Ekman pumping also arises from eddy-induced spatial variability of the sea surface temperature (SST) field that generates a wind stress curl in regions of crosswind SST gradients through a response of surface winds to SST-induced surface heating variations. SST-induced Ekman pumping is shown to be secondary to surface current-induced pumping in most regions of the World Ocean. Eddy-induced Ekman pumping resulting from the combination of surface current effects and air-sea interaction represents an order 1 perturbation of the background, basin-scale Ekman pumping velocities from the large-scale wind fields. In western boundary currents and equatorward-flowing eastern boundary currents, cyclonic eddies preferentially entrain water from the coastal side of the boundary current, which primes the interiors of cyclones to have phytoplankton concentrations that are elevated relative to the background. In contrast, anticyclones formed in these regions contain locally depressed phytoplankton concentrations from the offshore waters. While eddy pumping from vertical displacements of isopycnals during eddy formation can affect the biology in the interiors of cyclones during the transient stage of their development, this ecosystem response cannot be sustained because of the persistent eddy-induced Ekman downwelling throughout the rest of the eddy lifetimes. Likewise, the persistent eddy-induced Ekman upwelling in anticyclones is of little benefit because of their low phytoplankton content at the time of formation. A definitive response to eddy pumping is therefore difficult to detect from satellite observations alone. Eddies formed in regions where anticyclones preferentially entrain water with elevated phytoplankton concentrations, such as the South Indian Ocean, or in some mid-ocean gyre regions where small-amplitude eddies form (e.g., the oligotrophic South Pacific), an ecosystem response to eddy-induced Ekman pumping is observed. Conversely, cyclones in these regions entrain water that is low in chlorophyll, resulting in negative chlorophyll anomalies that are sustained by Ekman downwelling throughout the eddy lifetimes. The phytoplankton response to eddy-induced Ekman upwelling in anticyclones is seasonal, occurring only during the winter. It is proposed that the mechanism for the lack of ecosystem response to eddy-induced Ekman upwelling during the summer is the decoupling of the mixed layer from the nutricline. The observations presented in this dissertation provide a baseline from which coupled ocean circulation and biogeochemical models can be assessed. If coupled models are able to reproduce correctly the observed influence of mesoscale eddies on photoautotrophic communities, further insight into the mechanisms for this variability could be gained from the model output using the methodologies developed in this dissertation together with investigation of subsurface variability in the models below the depth to which chlorophyll can be inferred from the satellite observations. / Graduation date: 2013

The boundary layer is the part of the atmosphere that is in direct contact with the ground via turbulent motion. At mid-latitudes the boundary layer is usually one or a few kilometers deep, while in the Arctic it is much more shallow, typically a few hundred meters or less. The reason is that here the absolute temperature increases in the lowest kilometer, making the boundary layer semi-permanently stably stratified. The exchange of heat, momentum and tracers between the atmosphere, ocean and ground under stable stratification is discussed from an observational, modeling and climate-change point of view. A compilation of six observational datasets, ordered by the Richardson number (rather than the widely used Monin-Obukhov length) reveals new information about turbulence in the very stably stratified regime. An essentially new turbulence closure model, based on the total turbulent energy concept and these observational datasets, is developed and tested against large-eddy simulations with promising results. The role of mesoscale motion in the exchange between the atmosphere and surface is investigated both for observations and in idealized model simulations. Finally, it is found that the stably stratified boundary layer is more sensitive to external surface forcing than its neutral and convective counterparts. It is speculated that this could be part of the explanation for the observed Arctic amplification of climate change.

Polymer nanocomposites refer to a broad range of composite materials with polymer acting as the matrix and any material which has at least one dimension in the order of 1 ~ 100 nanometer acting as the filler. Due to unprecedented improvement observed in properties of the nanocomposites, research interest in this area has grown exponentially in recent years. In designing better nano-composites for advanced technological applications some of the major challenges are: understanding the structure-property relationships, interaction and integrity of the two components at the interface, the role of nanofillers in enhancing the properties of the resulting material. In our work, we have utilized first principle calculations, atomistic simulations, coarse-grained modeling and constitutive equations to develop structureproperty relationships for an amorphous aromatic piezoelectric polyimide substituted with nitrile dipole, carbon nanotubes and resulting nanocomposites. We have studied in detail structure-property relationships for carbon nanotubes and (? ?CN)APB/ODPA polyimide. We have developed chemically sound coarse-grained model based on atomic level simulations of the piezoelectric polyimide to address the larger length and time scale phenomena. The challenge of coarse grain model for these polymers is to reproduce electrical properties in addition to the structure and energetics; our model is the first to successfully achieve this goal. We have compared and analyzed atomistic scale simulation results on the nanocomposite with those predicted from micromechanics analysis. Notably, we have investigated the time dependent response of these highly complex polymers, to our best knowledge this is the first of its kind. In particular we have studied the thermal, mechanical and dielectric properties of the polyimide, nanotube and their nanocomposites through multi-scale modeling technique. We expect the results obtained and understanding gained through modeling and simulations may be used in guiding development of new nanocomposites for various advanced future applications. In conclusion we have developed a computational paradigm to rationally develop next generation nano-materials.

The effects of reactive metal particles on the microstructure and mechanical properties of epoxy-based composites are investigated in this work. To examine these effects castings of epoxy reinforced with 20-40 vol.% Al and 0-10 vol.% Ni were prepared, while varying the aluminum particle size from 5 to 50 microns and holding the nickel particle size constant at 50 microns. In total eight composite materials were produced, possessing unique microstructures. The microstructure is quantitatively characterized and correlated with the composite constitutive response determined from quasi-static and dynamic compressive loading conditions at strain-rates from 1e-4 to 5e3 /s. Microstructures from each composite and at each strain rate were analyzed to determine the amount of particle strain as a function of bulk strain and strain rate. Using computational simulations of representative microstructures of select composites, the epoxy matrix-metallic particle and particle-particle interactions at the mesoscale under dynamic compressive loading conditions were further examined. From computational simulation data, the stress and strain localization effects were characterized at the mesoscale and the bulk mechanical behavior was decomposed into the individual contributions of the constituent phases. The particle strain and computational analysis provided a greater understanding of the mechanisms associated with particle deformation and stress transfer between phases, and their influence on the overall mechanical response of polymer matrix composites reinforced with metallic particles. The highly heterogeneous composite microstructure and the high contrasting properties of the individual constituents were found to drive localized deformations that are often more pronounced than those in the bulk material. The strain rate behavior of epoxy is shown to cause a strain rate dependent deformation response of reinforcement particle phases that are typically strain rate independent. Additionally, the epoxy matrix strength behavior was found to have a higher dependence on strain rate due to the presence of metal particle fillers. Discrepancies between experimental and simulation mechanical behavior results and these findings indicate a need for epoxy constitutive models to incorporate effects of particle reinforcement on the mechanical behavior.

Models of tephra fallout are used to assess volcanic hazards in advance of eruptions and in near-real-time. Current models often approximate the wind field using simplistic assumptions of the atmosphere that cannot account for typical variations in wind velocity that occur in time and three-dimensional space. Here, a widely used mesoscale atmospheric model is used to improve forecasts of the location of the major axis of dispersion for erupting plumes. The Pennsylvania State University-National Center for Atmospheric Research fifth-generation Mesoscale Model (MM5) specializes in atmospheric prediction for regions on the order of ten to hundreds of kilometers on a side. MM5 generates realistic wind fields based on the laws of conservation of mass, energy, and momentum, along with land surface data and atmospheric forecasts and observations.

Tetrad vascular plant data from Staffordshire (VC 39), accumulated 1995 - 2009, were analysed using Two Way Indicator Species Analysis and Principal Components Analysis, explored using GIS, to examine major floristic distribution patterns. Environmental data were also examined using Redundancy Analysis. The strongest floristic trend was a human influence axis, arranged south—north, with the south being most anthropocentric and the north having a high proportion of semi-natural habitats. Indicator species for different parts of the County showed a corresponding strong biogeographic element, with ‘Southern-temperate’ and ‘Boreo-temperate’ characteristics in south and north respectively. The next most important factor in the distribution of plant species in Staffordshire was habitat quality and richness, which separated intensively-managed farmland from areas with seminatural habitats and from human influenced urban areas. Environmental variables associated with the main division in the data were mainly climate and altitude and the extent of developed land. Historic field patterns, identified by Staffordshire County Council, were associated with semi-natural habitats, linking an historical landscape characterisation with vascular plant analysis for the first time in the UK. The investigation set out to characterise and classify spatial units in terms of biodiversity importance, and to produce an objective ecological classification of the Staffordshire region based on botanical and physical features. The resulting classification characterises different areas of the County in terms of plant species composition, and has many potential applications in terms of nature conservation strategy development, targeting of resources and monitoring. A further aim was to facilitate the selection of indicator species for high quality habitats – the study has enabled revision of Staffordshire’s axiophyte plant list. Both the ecological classification and indicator species will be of strategic value because the information provided is authenticated by being based on objective analysis of a wide range of data. In Staffordshire, further work is required to develop environmental data, especially an effective dataset for hedges, while comparison of the data with historical botanical data could provide further information about the County’s previous environment. The techniques developed in the study could be used more widely to: produce of lists of key species, provide information for environmental modeling, climate change monitoring, target nature conservation activity, and to develop landscape strategies.