L’environnement maritime du littoral implique de nombreux processus complexes, mais le manque de données en haute résolution couvrant une large zone sur une longue période est souvent l’obstacle principal à des recherches plus approfondies. Le radar haute-fréquence (HF) est un moyen de faire de la télédétection afin d’obtenir pratiquement en temps réel de l’information sur la surface de la mer et sur une large zone. Ainsi l’étude de l’inversion des paramètres marins à partir de données issues de radars HF est réellement porteuse de sens. Cette thèse fait l’usage d’un jeu de données collectées durant 13 mois par deux réseaux de radar HF à commande de phase pour étudier les caractéristiques de signaux d’échos de la mer, étudier les données à traiter et les méthodes d’inversion, calculer les paramètres de la surface de la mer et évaluer la précision de l’inversion radar des paramètres de la houle.Cette thèse se réfère à l’onde de sol radar HF, dont les ondes radio interagissent avec l’océan du fait de la diffraction de résonance de Bragg. Nous passons en revue l’historique et les applications du radar HF. Nous rappelons les bases de la théorie des ondes électromagnétiques. Nous décrivons les principes d’inversion des courants de surface de la mer, direction du vent et paramètres de houles. La faisabilité de l’inversion de paramètres de houle est examinée. A partir de l’analyse théorique et des études statistiques de nombreux échantillons de données, cette thèse propose une série de méthodes sur le traitement du signal brut et le contrôle de qualité, ce qui inclut la détermination du niveau de bruit, le moyennage des données dans l’espace et le temps, l’identification correcte des pics spectraux, le seuil de largeur de pic, etc. Respectant les caractéristiques de différents processus physiques, les inversions de courant et de vent utilisent des spectres collectés toutes les 20 minutes ; l’inversion des paramètres de houle utilise des spectres moyennés sur 1 heure. Les statistiques des spectres utilisés pour le calcul des paramètres de houle sont présentées pour chacune des stations. Un ensemble de programme efficaces de calculs automatiques avec une complexité algorithme réduite sont développés pour réaliser le traitement et en tirer les paramètres marins.Les vitesses de courants radiales sont obtenues à partir d’une unique station radar. Les champs de vecteurs de courants sont obtenus en combinant chaque station. On montre une année de débit moyen dans la mer d’Iroise, ainsi que le calcul de la vorticité et de la divergence. On étudie un ensemble de données d’un mois du radar SeaSonde de Qingdao. Les schémas de débit moyen, ainsi que la vorticité et la divergence sur un mois sont présentés.La direction relative du vent par rapport à la direction de visée du radar est mesurée à travers le ratio des amplitudes des pics de Bragg. Différents modèles empiriques sont employés pour obtenir la vitesse relative du vent par inversion radar. Les résultats présentés sont en accord avec les estimations prédites par le modèle. Différents modèles de distribution directionnelle sont utilisés pour mesurer le facteur de diffusion pour la mer d’Iroise.Cette thèse se concentre sur l’étude des paramètres de houle. Les résultats sont validés à l’aide de bouées et de données du modèle de vagues (Wavewatch III). L’estimation montre que la précision de la fréquence de houle est très bonne, la précision sur la hauteur significative de houle est très raisonnable et la précision sur la direction absolue de la houle est faible. La cohérence des mesures par chacune des stations radars est vérifiée par comparaison entre les deux. L’utilisation conjointe des échantillons est également prise en charge pour réaliser l’inversion. L’utilisation de deux radars n’améliore pas seulement la précision, mais résout aussi l’ambiguïté de direction relative de houle à partir d’une unique station et donne la direction absolue de vague avec une certaine précision. / Nearshore marine environment contains many complex processes, but the lack of high-resolution data over a large area during a long time is often the primary obstacle to further research. High-frequency (HF) radar is a mean of remote sensing which obtains continuous near-real time sea surface information over a large area. Thus the study of inversion of marine parameters from HF radar data is very meaningful. Thisthesis makes use of a 13-month-long dataset collected by two phased array HF radar to investigate the characteristics of the sea echo signals, study the data processing and inversion methods, compute sea surface parameters and evaluate the accuracy of radarinversion of swell parameters.The thesis refers to the ground wave HF radar, whose radio waves interact with ocean by Bragg resonance scattering. The development history and applications of HF radar is introduced. The basic theory of electromagnetic wave is reviewed. The principles of inversion of sea surface current, wind direction and swell parameters are described. The feasibility of the swell parameter inversion is investigated. Based on theoretical analysis and statistical studies of a large number of samples, the thesis proposes a series of methods on raw signal processing and quality control, including the determination of the noise level, data averaging in space and time, the proper identification of spectral peaks, the peak width threshold, etc. Respecting the characteristics of different physical processes, inversions of current and wind use spectra collected every 20 min; inversion of swell parameters uses one-hour averaged spectra. The statistics of qualified spectra for swell parameter calculations are presented for both stations. A set of efficient, with a reduced computational cost, automatic computing programs are developed to do the processing and derive marine parameters. Radial current velocities are derived from single radar station. Current vector fields are obtained by combination of both stations. One-year mean flow field in the Iroise Sea is shown, together with the computation of vorticity and divergence. A one-month SeaSonde radar dataset off Qingdao is studied. One-month mean flow pattern together with vorticity and divergence are presented.Relative wind direction with respect to radar look direction is measured through ratio of Bragg peaks amplitudes. Different empirical models are employed to derive radar-inverted relative wind direction. Results show reasonable agreement with model estimations. Different directional distribution models are used to measure the spreading factor for the Iroise Sea. The thesis focuses on the study of swell parameters. Results are validated by buoy and wave model (WAVEWATCH III) data. The assessments show that the accuracy of swell frequency is very good, the accuracy of swell significant waveheight is reasonable, and the accuracy of relative swell direction is low.Consistency of measurements by both radar stations is verified by comparison between the two. This also supports the use of double samples to do the inversion. Use of two radars not only further improves the accuracy but also solves the ambiguity of relative swell direction from single station and gives the absolute wave direction to a certain precision. The thesis proposes a constant relative directionmethod to derive swell significant waveheight, based on the studies of radar integral equation and the inverted results of relative swell direction. This proposal is demonstrated to improve the agreement of radar inversion and buoy/model provided significant waveheight and increases significantly the number of samples. The thesis investigates the accuracy of swell parameters obtained by HF radar. Contributions of random errors in radar observations are quantified. Comparing the differences between radar and buoy/model estimations gives assessments of the contribution of radar intrinsic uncertainty and contribution of other factors.

A pesquisa tem por finalidade avaliar dois modelos econométricos de mudanças de preços, que podem ser utilizados em estratégias de arbitragem estatística, o probit ordenado e o de decomposição, estimando seus parâmetros em quatro pregões de mini contratos de dólar futuro negociados na bolsa de valores brasileira. O estudo da negociação em alta frequência com a utilização de dados de transação a transação revela informações relativas à microestrutura de mercado que o ferramental mais tradicional não é capaz de desvendar. Uma das razões é que modelos tradicionais trabalham com variações de preço como variáveis contínuas, enquanto que ao considerar as variações de preço uma variável contínua e não uma variável discreta, como nos modelos aqui avaliados. Este trabalho acrescenta à literatura sobre microestrutura de mercado ao aplicar os modelos estudados em um ativo distinto daqueles avaliados nos papers originais, voltados ao exame do mercado de ações. Esta pesquisa concluiu que os modelos probit ordenado e de decomposição podem ser utilizados para previsão de mini contratos de dólar futuro e que o modelo de decomposição apresenta parâmetros mais significantes. Também concluiu-se que, no modelo probit ordenado, as variáveis de volume e time duration não se apresentaram relevantes na determinação do preço desse contrato e que a quantidade de defasagens utilizadas nos parâmetros estimados pode variar dentre os pregões. / The research aims to evaluate two econometric models of price change, which can be used in strategies of statistical arbitrage, the ordered probit model and the decomposition model, estimating its parameters in four trading sessions of mini US dollar futures contracts traded on the Brazilian Stock Exchange. The study of high frequency trading with the use of trade-by-trade price movements reveals information related to the market microstructure that the more traditional econometric tools are not able to solve when considering the price changes as a continuous variable and not a discrete one, like in the models evaluated here. This work adds to the literature on market microstructure by applying the models studied in an asset different from those evaluated in the original papers, aimed at examining the stock market. This research concluded that the ordered probit and decomposition models can be used to predict mini US dollar futures price changes and that the decomposition model presents more significant parameters. It was also concluded that, in the ordered probit model, the volume and time duration variables were not relevant in determining the price of this contract and that the number of lags used to estimate parameters can vary among the trading sessions.

No description available.

No description available.

No description available.

Notre connaissance des mécanismes d’acquisition de la chimie des eaux continentales est limitée par la fréquence des prélèvements qui peuvent être réalisés sur le terrain. Échantillonner une rivière tous les 15 jours, c’est négliger une variabilité de sa composition qu’on commence à suspecter et qui renseigne sur des processus ou des variations de sources de courte période intégrés sur l’ensemble du bassin versant. L’automatisation des prélèvements et la mesure en temps réel est un défi important non seulement du point de vue de la compréhension des mécanismes contrôlant la chimie des eaux, de leur dynamique temporelle, des relations entre concentrations et débits, mais aussi du point de vue appliqué pour la détection des pollutions ponctuelles. Ce développement intéresse les professionnels de l’eau qui ne sont pas à ce jour contraint par la loi, mais qui pourraient le devenir. Dans le cadre du projet équipex CRITEX, un prototype de maison chimique appelé ”RiverLab”, véritable laboratoire d’analyses en temps réel sur le terrain a été développé et installé en novembre 2014 sur la bassin versant expérimental de l’Orgeval, un site étudié depuis 50 ans par les équipes de l’IRSTEA. Ce prototype est destiné à être déployé sur deux autres sites du réseau national des bassins versants (dispositif SOERE). Nous nous focalisons sur les deux principaux forçages naturels que subi un bassin versant à savoir les apports de précipitations et l’énergie solaire revisité sous l’angle de la haute fréquence / Our knowledge of the mechanisms of acquisition of inland water chemistry is limited by the frequency of sampling that can be carried out in the field. Sampling a river every 15 days means neglecting a variability in its composition that is beginning to be suspected and which informs processes or variations of short-term sources integrated over the entire catchment area. The automation of sampling and real-time measurement is an important challenge not only from the point of view of understanding the mechanisms controlling water chemistry, their temporal dynamics, the relations between concentrations and flows, but also from the point of view of applied for the detection of point pollution. This development is of interest to water professionals who are not yet legally constrained, but who could become so.As part of the CRITEX team project, a prototype of a chemical house called "RiverLab", a real laboratory for real-time field analysis, was developed and installed in November 2014 on the experimental watershed of Orgeval, a site studied for 50 years by the IRSTEA teams. This prototype is intended to be deployed on two other sites of the national watershed network (SOERE device). We focus on the two main natural forces that have undergone a watershed, namely precipitation inputs and solar energy revisited in terms of high frequency

Pour réduire significativement les émissions de CO2 dans les moteurs à combustion interne, un levier majeur est la réduction des pertes thermiques pariétales lors de la combustion. Ces pertes présentent un pic de plusieurs MW/m2 près du point mort haut, et sont liées à des phénomènes complexes d'interaction flamme-paroi qui dépendent du mode de combustion. Afin de mieux appréhender les phénomènes associés, il est nécessaire de caractériser le flux thermique à des échelles temporelles inférieures à la milliseconde.Dans ces travaux, une machine à compression rapide et une cellule à précombustion à volume constant sont utilisées pour simuler les phénomènes de combustion rencontrés en moteurs. Des thermocouples à jonction fines permettent une mesure de flux thermique instantanée avec une résolution temporelle de 0.1 ms. Ces moyens d'essais permettent de reproduire trois modes de combustion : flamme de propagation, flamme de diffusion et auto-inflammation. Ces travaux permettent également d'évaluer les différentes technologies envisageables de mesure de transfert thermique en combustion (thermocouples, thermorésistances et thermométrie phosphore rapide) au regard des caractéristiques métrologiques requises par la rapidité des phénomènes mis en jeu.Le flux lors du transfert thermique atteint des valeurs de plusieurs MW/m2 avec une forme qui dépend du mode de combustion. Le flux lors de la propagation d'une flamme prémélangée est dominé par un pic lors de l'interaction flamme paroi,d'environ 5 MW/m2 et de durée 0.5 ms. Le flux lors de la combustion d'un jet Diesel est approximativement un plateau pendant la durée de l'injection ; il est dominé par l'effet d'entraînement d'air par le jet qui cause une augmentation du coefficient de transfert convectif jusqu'à des valeurs de 10 kW/m2/K, l'augmentation de température liée à la combustion étant secondaire. Dans le cas d'ondes de pression générées par une auto-inflammation rapides de gaz(cliquetis lors d'un allumage commandé ou HCCI à fort contenu énergétique), une corrélation est observée entre l'intensité du cliquetis et le flux thermique associé, quel que soit le mode de combustion qui génère les oscillations de pression. Le flux lors du cliquetis est 3 à 5 fois plus élevé que lors d'une combustion par flamme de propagation comparable. / CO2 emissions in internal combustion engines are linked with inefficiencies due to wall heat losses during combustion.Those losses exhibit a sharp peak of a few MW/m2 close to top dead center and are linked to complex flame/wall interaction phenomena that vary with the combustion mode. A fine understanding of the associated phenomena requires experimental characterization of wall heat flux with a time resolution better than the millisecond. In this PhD work, a rapid compression machine and a precombustion cell are used to reproduce engine combustion phenomena. Thin-junction thermocouples allow an instantaneous measurement of the wall heat flux with a time resolution of 0.1 ms. Three combustion modes are generated: propagation flame, diffusion flame and auto-ignition.Different possible measurement technologies and procedures (thermocouples, thermoresistances and rapid phosphor thermometry) are compared and benchmarked against the features of combustion phenomena. Flux during wall heat transfer reaches values of a few MW/m2 and its shape varies with the combustion mode. During premixed flame propagation, flux is dominated by a peak during flame-wall interaction of about 5 MW/m2 in amplitude and 0.5 ms in duration. During Diesel combustion, heat flux is approximately constant during the injection duration; itsevolution is driven by an increase of the convection coefficient up to 10 kW/m2/K, which is attributed to air entrainment by the spray; the temperature increase from combustion is considered a second-order effect. During combustion presenting a pressure wave propagation (e.g. knock for some spark-ignition cases or HCCI with high energy content), the intensity of pressure oscillations and wall heat flux are shown to be correlated. That correlation is independent of the phenomenon creating the pressure wave; heat flux during knock is 3-5 times higher than for a comparable premixed propagation flame.

The runoff of suspended solids and nutrients from land into the nation's lakes and rivers can have severe impacts on the health of these systems and their uses. High-frequency environmental data from sensors can provide insight into fundamental biogeochemical processes that dictate water quality and provide regulators with valuable knowledge on how to manage critical resources. The goal of this research was to utilize sensor technology, telemetry hardware, cyberinfrastructure, and water quality models to create a sensing system that will allow the investigation of the fate and transport of dissolved oxygen, suspended solids, nutrients, and other water quality parameters throughout a watershed dominated by agricultural activity. Deploying these sensors at multiple locations along the stream enabled the investigation of these processes from the fine scale to the larger watershed scale. Results from this research addressed both fundamental science and resource management issues regarding water quality. Using high-frequency data, a dramatic diel cycle in dissolved oxygen was observed with nonlinear dynamics which was successfully modeled mathematically, and excursions in water quality criteria were observed. In addition, a diel pattern in turbidity was discovered with higher levels at night likely caused by bioturbation (i.e. nocturnal activity of bottom feeding fishes) which resulted in higher suspended solids loadings during nighttime. Furthermore, the QUAL2K model was successfully calibrated for water quality using sensor measurements and grab samples from volunteer, IOWATER data. Nutrient loading rates (nitrate-N, orthophosphate, and total dissolved solids) were estimated along the entire creek and were similar to other Iowa streams. Volunteer environmental data were found to be helpful in model calibration for some parameters (e.g. TSS and nitrate). The construction and operation of a sensing system in Clear Creek contributed to water quality science and engineering. Findings from the configuration and field testing of sensing station components such as water quality sensors, power systems and communication hardware will aid the design of future sensing systems and environmental observatories. Integrating the methodology of this research with future observing systems will further our understanding of water quality processes and help maintain the health and value of our nation's water environment.

High frequency water quality monitoring presents unique and unlimited opportunities of exploring spatio-temporal variation in water quality. Knowledge gained from analyzing high frequency water quality data can provide more clarity regarding transportation and processing of water constituents over time and space and scale. This study analyzes high frequency discharge, nitrate load and concentration data for three watersheds of different sizes - Cedar River Watershed, North Raccoon and Middle Raccoon. Each of these sites were monitored for 2-3 calendar years. Sudden spikes in discharge, nitrate concentration and load data, also defined as "events" were analyzed in great detail to understand the patterns in event occurrence and event intensity. Smaller watersheds seemed to have sharper and "flashier" events compared to bigger watersheds. Nitrate concentration events were flatter in shape compared to discharge and nitrogen load events. The relationship between nitrogen concentration and discharge was found to be varying over time, unlike the relationship between nitrate load and discharge, which were almost perfectly correlated for most site-year combinations. Based on more than 40,000 simulations, it was determined that high frequency water quality sampling is not only efficient in capturing minute spatio-temporal variations but can also capture nitrate exceedances to a greater degree. High frequency sampling was also associated with higher yield ratio in nitrate load estimates, not only during high flow periods, but also during the non-high-flow period.

Purpose: To study the impact of visual cues, speech materials and age on the frequency bandwidth necessary for optimizing speech recognition performance in listeners with normal hearing. Method: Speech recognition abilities of adults and children with normal hearing were assessed using three speech perception tests that were low-pass (LP) filtered and presented in quiet and noise. The speech materials included the Multimodal Lexical Sentence Test (MLST) that was presented in auditory-only and auditory-visual modalities for the purpose of determining the listener's visual benefit. In addition, The University of Western Ontario Plurals Test (UWO) assessed listeners' ability to detect high frequency acoustic information (e.g., /s/ and /z/) in isolated words and The Maryland CNC test that assessed speech recognition performance using isolated single words. Speech recognition performance was calculated as percent correct and was compared across groups (children and adults), tests (MLST, UWO, and CNC) and conditions (quiet and noise). Results: Statistical analyses revealed a number of significant findings. The effect of visual cues was significant in adults and children. The type of speech material had significant impact on the frequency bandwidth required for adults and children to optimize speech recognition performance. The children required significantly more bandwidth to optimize performance than adults across speech perception tests and conditions of quiet and noise. Adults and children required significantly more bandwidth in noise than in quiet across speech perception tests. Conclusion: The results suggest that children and adults require significantly less bandwidth for optimizing speech recognition performance when assessed using sentence materials which provide visual cues. Children, however, showed less benefit from visual cues in the noise condition than adults. The amount of bandwidth required by both groups decreased as a function of the speech material. In other words, the more ecologically valid the speech material (e.g., sentences with visual cues versus single isolated words), the less bandwidth was required for optimizing performance. In all, the optimal bandwidth (except for the noise condition of the UWO test) is achievable with current amplification schemes.