Exploration and extraction activities in the Alberta Oil Sands region have degraded thousands of hectares of land, about 40 % of which were once covered by peatlands. To date, there is very little information regarding the reclamation of peatlands, and even less related to their reclamation after oil sands disturbances. The results of this study will be used to develop protocols for the reclamation of peatlands after disturbance due to oil sands exploration and extraction operations in the Alberta Oil Sands region. Carex aquatilis is considered to be an ideal candidate species for initiating reclamation because it is a common pioneer species of disturbed landscapes, and has been shown to have a wide range of tolerances for several environmental factors in other areas of its overall distribution. The results of this study showed even wider ranges of tolerance than previous studies have reported, and did so using an ecological field study, a greenhouse propagation study, and a greenhouse Na+ tolerance study. A field study used vegetation surveys and environmental data to determine that Carex aquatilis is abundant in several wetland types and is not strongly associated with any of the species with which it was commonly found. In addition, C. aquatilis also was present across wide ranges of environmental variables including pH, conductivity, shade, water depth, Na+ concentration of surface water and percent organic carbon content of the substrate. These wide ranges of tolerance should allow C. aquatilis to establish after many types of disturbance. A greenhouse propagation study tested for population differences in percent germination after four stratification treatments, three stratification durations and four germination treatments. The overall best stratification treatment was to store seeds moist at 4°C for 30 days. The most effective germination treatment across all treatments was to plant seeds just below the surface of the substrate with the water level maintained at 1cm below the substrate surface. Only 9% of all sown seeds germinated. There were differences in germination rates between populations, although it is unknown whether these differences were due to differences in viability, timing of seed maturation, stratification requirements or germination requirements. A greenhouse Na+ tolerance experiment tested for differences in Na+ tolerance between four populations and two age classes to assess the efficacy of C. aquatilis as a pioneer species after oil sands surface mining disturbance. Five Na+ concentrations ranging from 5 to 2000 mg L-1 were tested. Plants showed no differences in Na+ tolerance between populations, but younger plants were more susceptible to damage from Na+ stress than older plants so it is recommended that older plants are used to initiate reclamation. It is also recommended that Na+ be managed on surface mine reclamations so that it is below 500 mg L-1.

In recent years, studying soft errors has become an issue of greater importance. There have been many methods developed that estimate the Soft Error Rate. Those methods are either deterministic or statistical. The proposed deterministic model aims to improve Soft Error Rate estimation by accurately approximating the generated pulse and all subsequent pulses. The generated pulse is approximated by a piecewise function consisting of two Weibull cumulative distribution functions. This method is an improvement over existing methods as it offers high accuracy while requiring less pre-characterization. The proposed algorithm reduces pre-characterization by allowing the beta Weibull parameter to be calculated during runtime using gate parameters such as the gate delay.

L’étude de l’interaction entre une impulsion électromagnétique et un matériau dispersif est un vieux sujet qui remonte au moins au tournant du vingtième siècle avec Sommerfeld et Brillouin. Depuis lors, de nombreux scientifiques ont apporté leur contribution mais il reste un certain nombre de questions, y compris dans des domaines supposés complètement défrichés. Le but de cette thèse est d’aborder, de la manière la plus systématique qui soit, les problèmes les plus emblématiques du domaine de ce qu’il est convenu d’appeler “l’électrodynamique classique dans la matière”. Dans un premier temps, nous avons cherché à répondre à la question simple et fondamentale suivante: un certain nombre de points expérimentaux étant donnés trouver une procédure qui permette d'obtenir un modèle analytique pour la permittivité. Maintenant équipé d’un modèle phénoménologique pour la permittivité, on s’attache à déterminer les caractéristiques essentielles liées à la propagation d’une impulsion électromagnétique dans un milieu arbitrairement dispersif. Après quoi, nous changeons de dimension en nous attaquant aux problèmes tridimensionnels. Nous commençons par l’étude du champ électromagnétique diffusé par une particle chargée ponctuelle oscillant à une fréquence fixe donnée, et placée à côté d’un objet. La dernière partie est une partie plus pratique où l'on aborde des problèmes qui touchent à la technologie. Chercher les résonances de structures ouvertes n’est pas une mince affaire et met en scène des résonances dans le plan complexe. Ici on détermine à la main les QNM (Quasi-Normal Modes) d’une cavité Fabry-Pérot. / The study of the interaction between electromagnetic pulses with dispersive and possible passive materials has a long tradition that can be traced, at least, to the works of Sommerfeld and Brillouin. As time has passed many scientist have contribute to a better understanding of this kind of phenomena. However some well established concepts need to be revisited and some questions remain open. The aim of this thesis is then, to tackle in a very systematic way, some of the most representative problems in this area that has been called “classical electrodynamics in matter”: I) We have developed a mathematical procedure that allows to find analytical models that in the frequency domain fulfils two fundamental physical properties: reality in time domain and causality. II) Our second task is to determine the velocity of an electromagnetic pulse that propagates in a highly dispersive medium. III) The problem of describing the electromagnetic field generated by an oscillating charge and itsinteraction with some dispersive 3D-object is also studied. The first part deals with the pure description of an EM field generated by arbitrary charge density $\rho$ and corresponding current density $j$. Second part is related to the interaction of this polyharmonic EM field and a dispersive object (a sphere). IV) Finding the resonant frequencies and consequently the resonant modes is a well known problem in physics, when the fields are not strictly confined and can leak to the whole universe we can say that we are dealing with Quasi Normal Modes (QNMs). In this thesis we give a brief and straightforward way of deriving the QNMs of a Fabry-Perot cavity.

Depuis plus de deux décennies, I'enregistrement des signaux hydroacoustiques par des hydrophones dans le canal SOFAR (SOund Fixing And Ranging) a permis la détection et la localisation de nombreux séismes de faible magnitude dans I'océan. Cependant, I'interprétation de ces signaux est actuellement incomplète. La complexité de la conversion des ondes sismiques en ondes acoustiques, appelées « ondes T », au niveau du plancher océanique, et de leur propagation dans la tranche d'eau n'est pas intégralement comprise et ne permet pas d'extraire beaucoup d'informations sur les séismes à leur origine. Une simulation numérique de ces signaux est proposée pour identifier et comprendre les paramètres environnementaux et les phénomènes majeurs entrant en jeu dans la génération des ondes T. L'approche proposée est la méthode des éléments spectraux, mise en œuvre dans le code SPECFEM2D. Cette approche s'avère adaptée à cet usage car elle permet d'aborder le phénomène de la génération et la propagation des ondes T dans son ensemble en tenant compte ses principales caractéristiques : le couplage fluide/solide (propagation sismique, conversion, et propagation acoustique), les profils de vitesse d'onde dans I'eau et dans la croute, et le tenseur des moments sismiques donnant le diagramme de radiation de la source. Les signaux simulés présentent des formes, des durées, des temps d'arrivée, et des amplitudes relatives tout à fait comparables aux enregistrements réels. Les différences observées proviennent sans doute du traitement 2D du problème et d'une représentation trop simpliste des environnements et de la source sismique. / For more than 2 decades, recording of hydroacoustic signals in the sopen channel (Sound Fixing And Ranging) has allowed the detection and localization of many low-magnitude earthquakes in the ocean. However the interpretation of these signals is still incomplete.The generation of acoustic waves, known as T-waves, resulting from the conversion of seismic waves into acoustic waves at the sea-bottom, and their propagation in the water column are not yet fully understood, which prevents to extract more information about the earthquakes they originate from. Here, we use numerical modelling to identify and understand the main environmental parameters and phenomena that control the generation and propagation of acoustic T-waves in the ocean. The proposed approach is a spectral element method, implemented in the code SPECFEM2D, which allows to address the problem in a comprehensive way taking into account the solid/fluid coupling (seismic propagation, conversion, acoustic propagation), velocity profiles of the waves in the crust and the ocean, and the moment tensor of the earthquake that defines the radiation pattern of the seismic source. Simulated acoustic signals present many similarities in the shape, duration, arrival times and amplitudes of the predicted T-waves with observed T-waves. Differences are likely due to the 2D representation of the problem and to the simplistic representation of the environment and of the seismic source.

L'objet de cette thèse est l'étude de l'accélération de la propagation dans les équations de réaction-diffusion par un nouveau mécanisme d'échange avec une ligne de diffusion rapide. On répondra à la question de l'influence de ce couplage avec forte diffusivité sur la propagation en généralisant un résultat de Berestycki, Roquejoffre et Rossi de 2013. Le système d'équations étudié a été proposé pour donner une explication mathématique de l'influence des réseaux de transports sur les invasions biologiques. Dans un premier chapitre, on étudiera l'existence et l'unicité de solutions de type ondes progressives via une méthode de continuation. La transition se fait par l'intermédiaire d'une perturbation singulière qui paraît nouvelle dans ce contexte, connectant le système initial à un problème au bord de type Wentzell. Le second chapitre s'intéresse à la vitesse des ondes sus-mentionnées. On y démontre qu'elle croît comme la racine carrée de la diffusivité de l'espèce sur la route, ce qui généralise et démontre la robustesse du résultat de Berestycki, Roquejoffre et Rossi. De plus, on caractérise précisément le ratio de croissance comme unique vitesse admissible pour les ondes d'un système hypoelliptique a priori dégénéré. Enfin dans une dernière partie on s'intéresse à la dynamique. On y montre que ces ondes attirent une large classe de données initiales. En particulier on met en lumière un nouveau mécanisme d'attraction qui permet aux ondes d'attirer des données dont la taille est indépendante de la diffusivité sur la route ; c'est un résultat nouveau au sens où usuellement, l'accélération de fronts de réaction-diffusion se paie en renforçant les hypothèses nécessaires sur la taille des données initiales attirées. / The aim of the thesis is the study of enhancement of propagation in reaction-diffusion equations, through a new mechanism involving a line with fast diffusion. We answer the question of the influence of such a coupling with strong diffusion on propagation by generalizing a result of Berestycki, Roquejoffre and Rossi (2013). The model under study was proposed to give a mathematical understanding of the influence of transportation networks on biological invasions. The first chapter shows existence and uniqueness of travelling waves solutions with a continuation method. The transition occurs through a singular perturbation - new in this context - connecting the system with a Wentzell boundary value problem. The second chapter is concerned with the speed of the waves : we show that it grows as the square root of the diffusivity on the line, generalizing and showing the robustness of the result by Berestycki, Roquejoffre and Rossi. Moreover, the growth ratio is characterized as the unique admissible velocity for the waves of an hypoelliptic a priori degenerate system. The last part is about the dynamics : we show that the waves attract a large class of initial data. In particular, we shed light on a new mechanism of attraction which enables the waves to attract initial data with size independent of the diffusivity on the line : this is a new result, in the sense than usually, enhancement of propagation has to be paid by strengthening the assumptions on the initial data for invasion to happen.

Master of Science / Department of Mathematics / Nathan Albin / This thesis is focused on the use of networks to identify potentially malicious Android applications. There are many techniques that determine if an application is malicious, and they are ever-changing. Techniques to identify malicious applications must be robust as the schemes of creating malicious applications are changing as well. We propose the use of a network-based approach that is potentially effective at separating malicious from benign apps, given a small and noisy training set. The applications in our data set come from the Google Play Store and have been scanned for malicious behavior using Virus Total to produce a ground truth dataset. The apps in the resulting dataset have been represented as binary feature vectors (where the features represent permissions, intent actions, discriminative APIs, obfuscation signatures, and native code signatures). We use the feature vectors corresponding to apps to build a weighted network that captures the \closeness" between applications. We propagate labels, benign or malicious, from the labeled applications that form the training set to unlabeled applications (which we aim to label), and evaluate the effectiveness of the proposed approach in terms of precision, recall and F1-measure. We outline the algorithms for propagating labels that were used in our research and discuss the fine tuning of hyper-parameters. We compare our results to known supervised learning algorithms, such as k-nearest-neighbors and Naive Bayes, that can be used to learn classifiers from the training labeled data and subsequently use the classifiers to label the unlabeled test data. We discuss potential improvements on our methods and ways to further this research.

Thesis (MTech (Horticulture))--Cape Peninsula University of Technology, 2007. / Gethyllis multifo/ia and Gethyl/is villose (Family: AMARYLLIDACEAE) are indigenous geophytes, growing naturally in the Worcester area, Western Cape. G. muliiiolie falls in the Vulnerable category of the Red Data List of Southern African Plants while G. vil/osa is not threatened at all. Both these species are winter growers and start their growing phase between March and April. These bulbs start their dormant phase between September and October when their leaves start to die down. Flowers of both species are short-lived and borne towards the end of November and early December when no leaves are present. The leaves and berries are simultaneously pushed above ground at the onset of the new growing phase. The fruit of some Gethyl/is species is sweet, juicy, pleasantly aromatic and good to eat and has medicinal properties for the cure of various ailments. The genus is difficult to propagate asexually and very little is known about its propagation and cultivation. The fragrance and medicinal value of the fruit of G. multifolia necessitates future research in the commercial production of this species. A habitat observation study of the two species was conducted to assess the vulnerable status of G. multitolie. Asexual propagation experiments were conducted to find ways of reproducing these two species successfully. A hydro culture study was also conducted to ascertain whether this method of cultivation could be incorporated in the general cultivation of the two species. Finally an in vitro propagation study was conducted to look at faster methods of reproducing these two species. This is of extreme importance in the conservation of the vulnerable G. multitolie. Leaf, root and basal plate cuttings were unsuccessful with no rooting in both species. G. mulfifo/ia bulbs were propagated successfully using twin scaling, bulb cuttings, scooping and scoring propagation techniques with between 80% and 100% rooting success. G. vil/osa was unsuccessful using the above propagation techniques with a 0 - 40% rooting rate. Although Gefhyl/is species in general are sensitive to over-watering, the hydro culture experiment with the sub-irrigation system and leca pellet medium proved to be an effective method of cultivating both species throughout the growing phase. G. mulfifo/ia proved to be unsuccessful during the initial in vitro propagation experiments with no surviving explants during the initiation phase. Results improved with an increased number of trials. It is possible to grow both species by means of in vitro propagation, but more emphasis in future research, should be placed on the multiplication aspect of G. mu/tifo/ia, since not many new buibiets were produced. It was observed through this study that grazing domestic livestock, urban expansion (this includes agricultural extension) and in some cases the lack of interest shown in our indigenous plant species, are some of the main factors influencing the decline in numbers of this species. It is also recommended that more emphasis be placed on the conservation of South Africa's indigenous flora and that the vulnerable status of G. multifolia according to The Red Data List of Southern African Plants, be changed to the "Endangered category" as the factors causing its decline continue to increase.

[pt] O objetivo deste trabalho é apresentar o método das linhas aplicado ao estudo das características de propagação em estruturas planares. Na sua aplicação, discretiza-se variáveis especiais em um conjunto de linhas perpendiculares as fitas. Conseqüentemente, equações diferenciais parciais são reduzidas a equações diferenciais ordinárias com soluções bem conhecidas. Apesar da aplicação de diferenças finitas, o Método das Linhas é mais eficiente em termos de velocidade computacional e uso de memória, quando comparado com método de diferenças finitas usual. Algumas estruturas planares em guias retangulares com três camadas dielétricas foram analisadas e os resultados comparados com outros métodos. Nos apêndices mostra-se como é possível implementar programas com discretização não uniforme, o que aumenta a eficiência do método, bem como a discretização em duas direções, o que permite análise de ressoadores de microondas. Este trabalho pretende servir como auxilio básico, podendo ser útil em outros trabalhos empregando o Método das Linhas. / [en] In this work, the Method of Lines applied to the study of the propagation caracteristics of planar structures is presented. In this method, the spacial variables are discretized in several lines orientated in a direction normal to the conductor strips. Then, the partial differential equations are reduced to ordinary differential equations having well known solutions. Although this method uses finite differences, it is much more efficient with respect to computer speed and memory necessary, in comparison to the conventional finite difference method. A few planar structures with three dieletric layers, inside rectangular waveguides were analysed, and the results are compared with those using other reported methods. In the appendix, it is shown how to develop computer programs using non-uniform discretization for the improvement of the efficiency of this method. Also, the discretization in two directions is presented, which is adequate for analysis of microwave resonators. This work is intended as a basic approach to hel in further analysis using the Method of Lines.

D.Phil. (Botany) / A taxonomic revision of the genus Centella is presented that is based on vegetative and reproductive characters. Genetic data was included in the extensive research ofpossible hybridisation. Hybridization as a possible mechanism of evolution may account for the lack of conguence between taxonomic characters in Centella. Phylogenetic and phenetic methods of analysis were used to explore the relationships among species. The results show that Centella is a particularly difficult genus, with many ofthe problems experienced at species level being a reflection ofthe problematic nature of the Apiaceae. The taxonomic treatment includes nomenclature, synonymy, typification, diagnostic characters, detailed illustrations and geographical distribution of all 45 species that are recognized. Keys to the subgenera, sections and species are provided. The revision includes several taxonomic and nomenclatural changes. The genus is subdivided as follows: Subgenus Trisanthus Subgenus Solandra Subgenus Centella 1 species 4 species Section Erianthae Section Tri/obae Section Centella Section Virgatae 9 species 3 species 16 species 12 species The following 13 new species were described: C. annua, C. calcaria, C. rotundifolia, C. umbellata, C. ternata, C. crenata, C. cryptocarpa, C. gymnocarpa, C. pi/osa, C. brachycarpa, C. dolichocarpa, C. thesioides and C. glauca. One ofthese species C. cryptocarpa, has been collected for the first time. Centalla rotundifolia, C. cordata and C. longifolia were elevated to species rank and C. rigescens was reinstated as a species of Centella. The number of varieties was significantly reduced and variation within the species is mostly described informally. Varieties were retained for only two species. New varieties are: C. tridentata var. tridentata, C. tridentata var. littoralis, C. tridentata var. hermanniifolia, C. tridentata var. dregeana, C. graminifolia var. graminifolia, C.graminifolia var. natalensis and C. graminifolia var. stipitata.

Cette thèse propose une étude de la propagation non linéaire des infrasons dans l’atmosphère par résolution temporelle des équations de la mécanique des fluides compressibles en s’appuyant sur des algorithmes développés pour l’aéroacoustique. L’étude se restreint aux ondes émises par les explosions de forte amplitude. Une étude bibliographique est d’abord réalisée pour identifier les phénomènes physiques influents et pour détailler un modèle d’atmosphère réaliste qui soit compatible avec une résolution numérique directe. Les équations de Navier-Stokes sont ensuite formulées pour intégrer les effets hors-équilibre de la relaxation de la vibration moléculaire. Cette étape est accompagnée de développements analytiques servant de base `a une discussion des effets dissipatifs et des effets de l’inhomogénéité atmosphérique de grande échelle. Les méthodes numériques optimisées retenues pou l’implémentation du solveur sont ensuite présentées. La stabilité et la précision sont étudiées analytiquement, puis le code de calcul est validé par des séries de simulations dans des configurations simples pour lesquelles des solutions analytiques sont disponibles. Une discrétisation adaptée pour la simulation des infrasons non linéaires se propageant dans l’atmosphère est alors proposée. Le solveur est ensuite validé dans la configuration complexe d’un modèle réaliste d’atmosphère, d’abord par une convergence de maillage, puis par confrontation avec des méthodes numériques préexistantes. L’influence des effets dissipatifs et des effets non linéaires est ensuite discutée sur la base des résultats numériques. L’influence générale des vents atmosphériques est par ailleurs présentée et des phénomènes spécifiques tels que les réflexions partielles ou les ondes rampantes sont évoqués. Enfin, une confrontation des simulations avec une expérience de référence est réalisée et les similitudes entre les enregistrements numériques et expérimentaux sont soulignées. / This thesis proposes a study of non linear infrasound propagation through the atmosphere by a time integration of compressible fluid mechanics equations using aeroacoustics algorithms. The survey is restricted to the case of infrasound emitted by high amplitude explosions. Existing literature is reviewed to identify influential physical phenomena and to detail a realistic atmosphere model compatible with direct numerical methods. Afterwards, the Navier-Stokes equations are formulated to take non-equilibrated vibrational relaxation effects into account. Acoustic absorption and effects of high-scale atmospheric in homogeneities are then discussed on the basis of analytical developments. Subsequently, the optimized numerical methods implemented in the solver are introduced. Numerical stability and accuracy are analyzed, then the method is validated by series of simple configuration simulations and an adapted discretization is proposed to simulate non linear infrasound propagating in the atmosphere. The solver is validated in the configuration of a complex realistic atmosphere, first by a grid convergence method and second by confronting the results to those of other numerical approaches. The influence of acoustic absorption and non linear effects are then discussed by analyzing numerical results. General influence of winds is also introduced and specific phenomena, such as partially reflected waves or creeping waves, are evoked. Finally, simulations are compared with a full-scale experience and similarities between both kind results are underlined.