A radiation model was developed for retrieving cloud visible optical depth, droplet effective radius, and cloud top emission temperature using AVHRR satellite observations at 0.63, 3.7, and 11 ��m. The model was used to determine the sensitivity of the retrieved properties to various approximations often employed in such retrievals. Droplet effective radius appears to be the most sensitive to the commonly used approximations. Cloud properties retrieved using a 16-stream scheme were within ��5% of those retrieved using a 148-stream scheme. Cloud properties retrieved using double Henyey-Greenstein phase functions were within ��10% of those retrieved using Mie scattering. The retrieved cloud properties were used to investigate biases that arise when partly cloudy pixels were assumed to be overcast and biases that arise due to oblique satellite view angles. On average, cloud visible optical depths retrieved for partly cloudy pixels were 40-60% of those retrieved for overcast pixels. Likewise, cloud liquid water paths were 30-50%, droplet effective radii were 1-3 ��m smaller, and cloud top emission temperatures were 2-4K larger. Cloud visible optical depths retrieved at 60�� satellite zenith angles were 60-70% of those retrieved at nadir. The retrieved droplet effective radii and cloud top emission temperatures varied little with changing satellite zenith angle. For March 1989, cloud optical depths and cloud emission temperatures retrieved for pixels overcast by single-layer, low-level clouds were negatively correlated. Cloud optical depth, liquid water path, and droplet effective radius were positively correlated with the sea surface-cloud top temperature difference. The retrieved cloud properties were also compared for the spatial coherence, CLAVR (Clouds from AVHRR), and a threshold method based on International Satellite Cloud Climatology Project procedures. For regions containing single-layered cloud systems, fractional cloud cover and cloud brightness temperatures derived by the ISCCP-like threshold method were systematically larger than those derived by the spatial coherence method, whereas cloud reflectivities were systematically smaller. Cloud reflectivities and brightness temperatures derived by CLAVR and the spatial coherence method were in better agreement. / Graduation date: 1998

Physical-biological interactions in the Southern Ocean were investigated using remote sensing data from several different satellite sensors. Satellite sea surface temperature data were used to study the dynamics of the Antarctic Polar Front (PF). Satellite ocean color data were used to estimate surface chlorophyll concentrations and their relation to various physical forcings within the Southern Ocean. A detailed study of phytoplankton blooms at the Antarctic Polar Front revealed that elevated chlorophyll concentrations (phytoplankton blooms) occur most often in areas where the PF interacts with large topographic features within the Southern Ocean. The physical dynamics of the PF are strongly influenced by the topography, and in turn strongly influence phytoplankton bloom dynamics. The analysis of satellite data from the modern Southern Ocean indicates that phytoplankton are limited by the availability of the micronutrient iron in most areas. This iron-limitation implies that the elevated iron inputs during glacial periods would have led to increased phytoplankton primary and export production and a stronger sink for atmospheric CO��� in the Southern Ocean. / Graduation date: 2000

Large marine vertebrate species can exhibit vast movements, both horizontally and vertically, which challenges our ability to observe their behaviours at extended time-scales. There is a growing need to understand the intra- and inter-annual movements of mobile marine species of conservation concern in order to develop effective management strategies. The basking shark (Cetorhinus maximus) is the world's second largest fish species, however, a comprehensive understanding of this species’ ecology, biology and spatial behaviour in the north-east Atlantic is currently lacking. This thesis seeks to investigate the movement ecology of basking sharks using a suite of technologies to integrate biologging, biotelemetry, remotely sensed data, and ecological modelling techniques. I use satellite telemetry data from basking sharks tracked in 2012, 2013 and 2014 to quantify movements in coastal waters off the west coast of Scotland within the Sea of the Hebrides proposed MPA. Sharks exhibited seasonal residency to the proposed MPA, with three long-term tracked basking sharks demonstrating inter-annual site fidelity, returning to the same coastal waters in the year following tag deployment (Chapter 2). I reveal that sharks tracked into winter months exhibit one of three migration strategies spanning nine geo-political zones and the High Seas, demonstrating the need for multi-national cooperation in the management of this species across its range (Chapter 3). I examine the vertical space-use of basking sharks to improve an understanding of the processes that influence movements in all dimensions. Basking sharks exhibit seasonality in depth-use, conduct deep dives to over 1000 m, and alter their depth-use behaviour in order to remain within thermal niche of between 8 and 16 oC (Chapter 4). Finally, I combine contemporaneous data recorded by deployed satellite tags with remotely sensed environmental data to employ novel ecological modelling techniques to predict suitable habitat for basking sharks throughout the Atlantic Ocean (Chapter 5).

The quantification and description of sea surface temperature (SST) is critically important because it can influence the distribution, migration, and invasion of marine species; furthermore, SSTs are expected to be affected by climate change. Recent research indicates that there has been a warming trend in ocean temperatures over the last 50 years. Hence, we sought to identify and demonstrate how a particularly germane SST dataset can be used within the scope of global climate change research. For this project we assembled a 29-year nearshore time series of mean monthly SSTs along the North Pacific coastline, as well as mean monthly SSTs for ice-free regions of the Arctic, using remotely-sensed satellite data collected with the Advanced Very High Resolution Radiometer (AVHRR) instrument. By providing detailed information concerning both dataset generation and data limitations, we aimed to make these data comprehensible to an expanded audience concentrating on life sciences rather than the traditionally physical science-based community. Furthermore, by making these data freely and publically available in multiple formats, including GIS (geographic information systems) layers, we expand their visibility and the extent of their use. We then used the dataset to describe SST patterns of nearshore (< 20 km offshore) regions of 16 North Pacific ecoregions, and of ice-free regions of 20 Arctic ecoregions, as delineated by the Marine Ecoregions of the World (MEOW) hierarchical schema. Our work creates a better understanding of present temperature regimes in these critically sensitive areas, from which we can draw several basic conclusions. 1) AVHRR SST measurements alone are sufficient to identify temperature patterns pertinent to determining health of ecosystems; 2) Within the nearshore North Pacific, ecoregions along the California Current System are most vulnerable to habitat-altering SST changes; 3) sea ice distribution is a major factor affecting SSTs in Arctic ecoregions, causing concern for the welfare of Arctic species. / Graduation date: 2012

La caractérisation des dynamiques de sous-mésoéchelle (<10km) à la surface de l'océan et leurs impacts sur les processus océaniques globaux sont des enjeux scientifiques majeurs. L'imagerie satellitaire est un outil essentiel dans ce contexte, qui présente toutefois des limitations liées aux instruments de télédétection. Dans le cas des images de température de surface des océans (SST), les mesures satellitaires des structures océaniques sont limitées par la résolution grossière des capteurs micro-ondes (~50km) d'une part, et par la sensibilité aux conditions climatiques (e.g., couverture nuageuse) des instruments de mesure infrarouge haute-résolution. Dans cette thèse, nous nous intéressons à l'analyse, la modélisation et la reconstruction des structures turbulentes haute-résolution capturées par imagerie satellitaire de SST, et proposons quatre contributions principales. Dans un premier temps, nous développons une méthode de filtrage conjointe Fourier-ondelettes pour le prétraitement d'artefacts géométriques dans les observations satellitaires infrarouges. Dans un deuxième temps, nous nous focalisons sur la caractérisation de la variabilité géométrique de champs de température de surface (SST) en utilisant des modèles de marches aléatoires appliqués aux lignes de niveaux. En particulier, nous considérons des processus aléatoires de type schramm Loewner (SLE). Nous nous intéressons ensuite à la modélisation stochastique des variabilités inter-échelles de champs de SST. Des modèles stochastiques de textures multivariées sont introduits. Ces modèles permettent de reproduire des propriétés statistiques et spectrales similaires à celles des données ayant servi à les calibrer. Nous développons ensuite des méthodes de super-résolution de champs de SST conditionnellement à une observation basse-résolution. Nous utilisons des modèles multivariés de textures formulés dans le domaine des ondelettes, en exploitant l'apprentissage d'à priori statistiques (i.e., covariances et covariances croisées) des différentes sous-bandes à partir d'images haute-résolution. Des contraintes supplémentaires imposées sur la phase de Fourier des différentes sous-bandes simulées permettent la reconstruction de structures géométriques marquées tels que les fronts. Nous démontrons la pertinence de la méthode proposée sur des images satellitaires de SST obtenues à partir du capteur Modis/Aqua. / The characterization of sub-mesoscale dynamics (<10 km) in the ocean surface and their impact on global ocean processes are major scientific issues. Satellite imagery is an essential tool within this framework. However, the use of remote sensing techniques still raise challenging. For instance, regarding Sea Surface Temperature (SST) images, satellite measurements of oceanic structures are limited by the coarse resolution of microwave sensors (~50km) on one hand, and by sensitivity to climatic conditions (eg., Cloud cover) of high-resolution infrared instruments on the other hand. In this thesis, we are interested in analysis, modeling and reconstruction of high-resolution turbulent structures captured by satellite SST imagery. In this context, we propose four main contributions. First, we develop a joint Fourier-Wavelet filtering method for the pre-processing of geometrical noises in satellite-based infrared observations, namely the striping noises. Secondly, we focus on the characterization of the geometric variability of sea surface temperature (SST) fields using random walk models applied to SST isolines. In particular, we consider the class of Schramm Loewner evolution curves (SLE). We then focus on the stochastic modeling of the cross-scale variabilities of SST fields. Stochastic multivariate texture-based models are introduced. These models are designed to reproduce several statistics and spectral properties that are observed on the data that are used to calibrate the model. We then develop our framework for stochastic super-resolution of SST fields conditionally to low-resolution observations. We use multivariate texture-based models formulated in the wavelet domain. These models exploit the formulation of statistical and spectral priors (i.e., covariances and cross-covariances) on wavelet subbands. These priors are directly learned from exemplar high-resolution images. Additional constraints imposed on the Fourier-phase of the different simulated subbands allow the reconstruction of coherent geometric structures such as the edge information. Our method is tested and validated using infrared high-resolution satellite SST images provided by Aqua Modis sensor.

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.

Les instruments, satellites ou systèmes aéroportés, actuellement utilisés pour la détection et la caractérisation d'hydrocarbure sur la mer sont basés sur des moyens optiques ou radars. Ces moyens présentent une performance dégradée due à une fréquence encore trop importante de fausses alarmes ou à un temps de traitement des données trop conséquent. Les méthodes de détection, d'identification et de quantification des fuites d'hydrocarbures offshores peuvent donc être améliorées en associant robustesse et réactivité. Cette amélioration suppose une compréhension approfondie des phénomènes océanographiques et électromagnétiques à l'œuvre dans cette scène particulière. La thèse s'appuie sur des données regroupant des images optiques et SAR aéroportées ou satellites ainsi que des mesures réalisées en laboratoire. Ce jeu de données permet de vérifier la cohérence des résultats obtenus par modélisation. L'objectif de la thèse est de distinguer une surface de mer polluée d'une surface de mer propre à l'aide de la signature électromagnétique de la surface totale puis de détailler le type et la quantité d'hydrocarbure présent. La thèse se divise en deux domaines, à savoir modélisation océanographique et modélisation électromagnétique. La modélisation océanographique intègre la simulation de la surface rugueuse imitant une surface de mer propre, et polluée. Cette surface de mer doit être générée sur une superficie importante et doit conserver une résolution restituant les petites vagues avec un temps de génération minimal. La partie électromagnétique est centrée sur les modèles asymptotiques de diffusion des ondes électromagnétiques par une interface rugueuse. Ces modèles sont adaptés au contexte de la thèse, complexité de la scène et rapidité du traitement, mais nécessitent plusieurs hypothèses pour être appliqués. / Satellites or airborne systems currently used for the detection and characterization of oil slicks on sea surface are based on optical or radar means. These means have a lack of performance due to a too high frequency of false alarms or to an excessively long data processing time. The methods for detecting, identifying and quantifying offshore pollutant can therefore be improved by combining robustness and reactivity. This improvement implies an in-depth understanding of the oceanographic and electromagnetic phenomena at work in this particular scene. The thesis is based on data gathering aerial and satellite images and SAR as well as measurements carried out in laboratory. This dataset makes it possible to check the consistency of the results obtained by modeling. The objective of the thesis is to distinguish a polluted sea surface from a clean sea surface using the electromagnetic signature of the total surface and then to detail the type and quantity of pollutant. The thesis is divided into two domains, namely oceanographic modeling and electromagnetic modeling. Oceanographic modeling integrates the simulation of the rough surface imitating a clean or polluted sea surface. This sea surface must be generated over a large area with a thin resolution. The electromagnetic part is centered on the asymptotic models for the electromagnetic waves diffraction by a rough interface. These models are adapted to the context of the thesis, the complexity of the scene and the speed of processing, but require several hypotheses to be applied.

No description available.