Satellite-tracked surface drifter data from 1986 to 2008 acquired from NOAA/AOML and the sea-level anomaly (SLA) data of AVISO from 1992-2008 were used in this study to investigate the mesoscale anticyclonic eddies in the northern South China Sea (SCS) and Luzon Strait. A comparison of the concurrent drifter trajectories and SLA for two eddy events (2003/12~2004/02 and 2004/11~2005/01) indicates good agreement between the two datasets. From historical SLA data (1992-2008) it is found that 78 anticyclonic eddies can be identified in the studied region. The number of occurrence is highest in 1994, 1996, 2001 and 2004, and is lowest in 1998. This result is likely due to the ENSO event and the associated wind lessening in the SCS. Most eddies were generated off southwestern Taiwan coast, northern SCS and west of Luzon Strait. During northeastern monsoon the average life time of eddies is 66.88 days, and the average sea level height difference is 10-20 cm, occasionally reached a maximum value of over 30 cm. During southwestern monsoon eddies have an average lifetime of 51.43 days, and the average sea level height difference is mostly less than 15 cm. In particular, eddies off the southwestern Taiwan coast have the characteristics of lower sea level height difference and translational speed. Location of eddy generation has a marked seasonal variation. During northeastern monsoon, most eddies were concentrated in northern Luzon Strait, propagating longer distance toward the west along the continental shelf, even reaching 112¢XE. On the other hand, eddies generated during the southwestern monsoon can only reach 118¢XE. Statistical results indicate a linear relationship exists between the sea level height difference and the life time for eddies, implying that stronger eddies are more long-lived. Finally, from drifter tracks it can be found that as Kuroshio penetrates through the Luzon Strait and forms a loop current off the southwestern Taiwan coast. Subsequently, eddies could often be identified from the SLA data. Therefore, it can be conjectured that in the northern SCS anticyclone are often shed from the Luzon Strait by Kuroshio penetration.

The Wilson’s Plover (Charadrius wilsonia) is a species of concern in most southeastern U.S. coastal states, where it breeds and winters. The U.S. Shorebird Conservation Plan listed this species as a Species of High Concern (Prioritization Category 4), and the U.S. Fish and Wildlife Service has designated it as a Bird of Conservation Concern (BCC). Despite its conservation status, Wilson’s Plover population trends are poorly understood and little research has been conducted examining habitat factors affecting this species’ breeding and foraging ecology. I collected Wilson’s Plover demographic data and explored which habitat characteristics influenced breeding success and foraging site selection among three coastal habitat types (i.e. fiddler crab (Uca spp.) mud flats, beach front, and interdune sand flats) at Onslow Beach, Marine Corps Base Camp Lejeune, North Carolina, 2008-2009. I observed little difference between years in nest success (≥ 1 egg hatched), failure, and overall nest survival. The majority of nest failures were caused by mammalian predators. For those nests that hatched successfully, greater proportions were located in clumped vegetation than on bare ground or sparsely vegetated areas. In-season chick survival for both years was higher for nests that hatched earlier in the season, and for nests farthest from the broods’ final foraging territory. Productivity estimates (chicks fledged per breeding pair) were not significantly different between years (0.88 ± 0.26 fledged/pair in 2008, 1.00 ± 0.25 fledged/pair in 2009) despite a shift in foraging behavior, possibly related to habitat alterations and availability in 2009. My findings indicate that Wilson’s Plover adults and broods were flexible in establishing final foraging territories; in 2008 all final brood foraging territories were on fiddler flats while in 2009, final foraging territories were sometimes split between fiddler flats, beach front, and interdune sand flats. For those Wilson’s Plovers establishing territories on fiddler flats, area of the flat was the most important feature explaining use versus non-use of a particular flat; area ≥ 1250 m² was preferred. Close proximity to water and vegetative cover were also important habitat features in foraging site selection on fiddler crab mud flats, and in all habitat types combined. My findings will directly contribute to population and habitat research goals outlined in the U.S. Shorebird Plan and will supplement limited data about foraging and habitat use related to Wilson’s Plover breeding ecology. / Master of Science

Reconstruire des champs géophysiques à partir d'observations bruitées et partielles est un problème classique bien étudié dans la littérature. L'assimilation de données est une méthode populaire pour aborder ce problème, et se fait par l'utilisation de techniques classiques, comme le filtrage de Kalman d’ensemble ou des filtres particulaires qui procèdent à une évaluation online du modèle physique afin de fournir une prévision de l'état. La performance de l'assimilation de données dépend alors fortement de du modèle physique. En revanche, la quantité de données d'observation et de simulation a augmenté rapidement au cours des dernières années. Cette thèse traite l'assimilation de données d'une manière data-driven et ce, sans avoir accès aux équations explicites du modèle. Nous avons développé et évalué l'assimilation des données par analogues (AnDA), qui combine la méthode des analogues et des méthodes de filtrage stochastiques (filtres Kalman, filtres à particules, chaînes de Markov cachées). Des applications aux modèles chaotiques simplifiés et à des études de cas de télédétection réelle (température de surface de lamer, anomalies du niveau de la mer), nous démontrons la pertinence d'AnDA pour l'interpolation de données manquantes des systèmes dynamiques non linéaires et à haute dimension à partir d'observations irrégulières et bruyantes.Motivé par l'essor du machine learning récemment, la dernière partie de cette thèse est consacrée à l'élaboration de modèles deep learning pour la détection et de tourbillons océaniques à partir de données de sources multiples et/ou multi temporelles (ex: SST-SSH), l'objectif général étant de surpasser les approches dites expertes. / Reconstructing geophysical fields from noisy and partial remote sensing observations is a classical problem well studied in the literature. Data assimilation is one class of popular methods to address this issue, and is done through the use of classical stochastic filtering techniques, such as ensemble Kalman or particle filters and smoothers. They proceed by an online evaluation of the physical modelin order to provide a forecast for the state. Therefore, the performanceof data assimilation heavily relies on the definition of the physical model. In contrast, the amount of observation and simulation data has grown very quickly in the last decades. This thesis focuses on performing data assimilation in a data-driven way and this without having access to explicit model equations. The main contribution of this thesis lies in developing and evaluating the Analog Data Assimilation(AnDA), which combines analog methods (nearest neighbors search) and stochastic filtering methods (Kalman filters, particle filters, Hidden Markov Models). Through applications to both simplified chaotic models and real ocean remote sensing case-studies (sea surface temperature, along-track sea level anomalies), we demonstrate the relevance of AnDA for missing data interpolation of nonlinear and high dimensional dynamical systems from irregularly-sampled and noisy observations. Driven by the rise of machine learning in the recent years, the last part of this thesis is dedicated to the development of deep learning models for the detection and tracking of ocean eddies from multi-source and/or multi-temporal data (e.g., SST-SSH), the general objective being to outperform expert-based approaches.