Este estudo explora a obtenção da altura da camada limite planetária a partir de diferentes métodos e equipamentos, com foco na análise espectral do campo de vento do lidar Doppler. Realizou-se uma campanha de medição por 80 dias com dois lidars Dopplers comerciais com diferentes alcances verticais. Para o lidar com maior alcance vertical e consequetemente uma medida direta da altura da camada limite planetária, foi aplicado o método da variância da razão sinal-ruído e os resultados foram comparados aos valores de altura estimados a partir de dados de radiossondagem através de dois métodos distintos, o método da parcela e o método do número de Richardson bulk, o qual apresentou melhores resultados. Para o lidar com menor alcance vertical, aplicou-se a análise espectral que forneceu valores de comprimento de onda dos picos espectrais, proporcionais à altura da camada limite planetária. A constante de proporcionalidade para obtenção dos valores da altura da camada foi calculada comparando-se os resultados aos valores de altura estimados pela radiossondagem através do método do número de Richardson bulk, obtendo-se um valor de constante de proporcionalidade próximo ao sugerido na literatura. Entretanto, o conjunto de dados que mostrou boas estimativas dos picos espectrais foi bastante restrito, limitando a comparação ao longo do período entre os lidars. / This study explores the obtaining of the planetary boundary layer height through different methods and equipment, focusing on the spectral analysis of the wind field from Doppler lidar. An 80-day measurement campaign was conducted with two commercial Doppler lidars with different vertical ranges. For the lidar with greater vertical range and consequently a direct measurement of the planetary boundary layer heigth, the method of signal-to-noise ratio variance was applied and the results were compared to the height values estimated from radiosonding data using two different methods, the parcel method and the bulk Richardson number method, which presented better results. For the lidar with a lower vertical range, spectral analysis was applied, which provided wavelength values of the spectral peak, proportional to the planetary boundary layer height. The proportionality constant, in order to obtain the boundary height values, were calculated by comparing the height values estimated from the radiosonding data, using the bulk Richardson number method. Proportionality constant value close to the one found in the literature were obtained. However, the set of data that showed good estimates of the spectral peaks was very restricted, limiting the comparison over the period between the lidars.

Este estudo explora a obtenção da altura da camada limite planetária a partir de diferentes métodos e equipamentos, com foco na análise espectral do campo de vento do lidar Doppler. Realizou-se uma campanha de medição por 80 dias com dois lidars Dopplers comerciais com diferentes alcances verticais. Para o lidar com maior alcance vertical e consequetemente uma medida direta da altura da camada limite planetária, foi aplicado o método da variância da razão sinal-ruído e os resultados foram comparados aos valores de altura estimados a partir de dados de radiossondagem através de dois métodos distintos, o método da parcela e o método do número de Richardson bulk, o qual apresentou melhores resultados. Para o lidar com menor alcance vertical, aplicou-se a análise espectral que forneceu valores de comprimento de onda dos picos espectrais, proporcionais à altura da camada limite planetária. A constante de proporcionalidade para obtenção dos valores da altura da camada foi calculada comparando-se os resultados aos valores de altura estimados pela radiossondagem através do método do número de Richardson bulk, obtendo-se um valor de constante de proporcionalidade próximo ao sugerido na literatura. Entretanto, o conjunto de dados que mostrou boas estimativas dos picos espectrais foi bastante restrito, limitando a comparação ao longo do período entre os lidars. / This study explores the obtaining of the planetary boundary layer height through different methods and equipment, focusing on the spectral analysis of the wind field from Doppler lidar. An 80-day measurement campaign was conducted with two commercial Doppler lidars with different vertical ranges. For the lidar with greater vertical range and consequently a direct measurement of the planetary boundary layer heigth, the method of signal-to-noise ratio variance was applied and the results were compared to the height values estimated from radiosonding data using two different methods, the parcel method and the bulk Richardson number method, which presented better results. For the lidar with a lower vertical range, spectral analysis was applied, which provided wavelength values of the spectral peak, proportional to the planetary boundary layer height. The proportionality constant, in order to obtain the boundary height values, were calculated by comparing the height values estimated from the radiosonding data, using the bulk Richardson number method. Proportionality constant value close to the one found in the literature were obtained. However, the set of data that showed good estimates of the spectral peaks was very restricted, limiting the comparison over the period between the lidars.

Sur un site multi-influencé par des émissions urbaines et industrielles, l'analyse de la pollution aux aérosols, au voisinage des sources, requiert une connaissance multi-échelles de la dynamique atmosphérique. une campagne de mesure a été développée afin d'étudier la variabilité météorologique et micro-météorologique et l'évolution des particules, en particulier, submicroniques, sur une durée d'une année. Des oscillations de la concentration en aérosols, autour de la moyenne régionale, ont été identifiées le long du littoral dunkerquois, et attribuées aux phénomènes météorologiques locaux à proximité des industries. Des méthodes de reconnaissance et d'apprentissage supervisé, faisant appel aux mesures par anémomètre ultrasonique et aux profils verticaux du vent par lidar Doppler, ont été mises en œuvre pour établir la variabilité de phénomènes pertinents dans les événements de pollution de l'air : brise de mer, brouillard, front et tempête. L'analyse d'une base de données de six ans a permis de montrer que l’occurrence annuelle des brises de mer est corrélée à celle du nombre de journées anticycloniques. Par ailleurs, la fréquence annuelle des brouillards pourrait être liée à la concentration annuelle régionale en aérosols. L'analyse des covariances du vent a révélé deux situations contrastées, à faible et à fort flux turbulents. Le brouillard et la brise de mer, de faible flux, génèrent une pollution élevée aux PM₁, et sont le siège d'une forte concentration en aérosols organiques oxygénés (aérosols secondaires). Les situations à fort flux, favorisant les échanges verticaux, sont associées à une forte variabilité des sulfates particulaires. L'observation de longue durée a permis de mettre en évidence la construction d'épisodes de pollution particulaire, au cours de séquences de phénomènes météorologiques locaux, du fait des sources locales, mais aussi par incorporation de la pollution à plus grande échelle. / On a site that is multi-influenced by urban and industrial emissions, the analysis of aerosol pollution, in the vicinity of sources, requires a multi-scale knowledge of atmospheric dynamics. A measurement campaign was developed in order to study the meteorological and micro-meteorological variability and the evolution of particles, in particular, submicronic evolution, during a one-year period.Oscillations of the aerosol concentration around the regional average were identified along the Dunkirk coastline, and were attributed to the local meteorological phenomena close to the industries. Recognition and machine learning methods using measurements by an ultrasonic anemometer and vertical wind profiles by a Doppler lidar, were implemented to define the variability of relevant phenomena in air pollution events : sea breeze, fog, front and storm. A six-years database analysis has highlighted a correlation between the annual sea breeze occurrence and the annual number of anticyclonic days. Furthermore, the annual fog frequency could be connected with the annual regional concentration of aerosols. Analysis of wind covariance revealed two contrasting situations, low-level and high-level turbulent fluxes. The fog and the sea breeze, with low-level fluxes, generate a high PM₁ pollution and are in favor of a high organic oxygenated aerosols concentration (secondary aerosols). High-level fluxes situations, favoring vertical exchanges, are associated with a large variability of sulfate aerosols. The long-term observation, made it possible to highlight the development of episodes of particulate pollution during local weather phenomena, owing to the local emissions, but also by taking into account the larger-scale pollution.

This work relates to the remote sensing technologies to determine wind velocity and its related phenomena. Laser-based pulsed systems with heterodyne detection allow a very sensitive measurement of particles’ velocity in the atmosphere. The heterodyne detection theory and the essential principles and properties of this technology are presented. An all-fiber coherent Doppler lidar system in master-oscillator power-amplifier design with a novel multi-wavelength channel configuration was developed and introduced, characterized, and validated in this work. The fiber amplifier is a crucial component of multi-channel all-fiber systems and is therefore discussed with special emphasis on non-linear effects and pulse distortions. A novel feedback controlled pulse-shaping unit was developed to control pulse energy and the shape of the amplified pulses of each channel. Each and every wavelength-channel creates an independent speckle pattern, which is demonstrated by hard targets with diffuse character and atmospheric single-shot measurements, and compared with theoretical results. Utilizing four channels, the precision of the measured wind velocity can be improved and this feature is demonstrated with measured atmospheric return signals. A correlation technique is presented, which enables the enhancement of the SNR at higher backscattered powers by utilizing multiple channels. The multi-wavelength system was designed to work with four wavelength-channels, which also allow measurements along the same line-of-sight with different channel configurations. These channels can be different in pulse shape, pulse length, and pulse repetition frequency. In this work, the resulting benefits of using multiple channels are shown with atmospheric measurements and scientific discussion of these is presented.:1. Introduction to the subject 2. Introduction to pulsed coherent Doppler lidar 3. Methodology: Multi-channel coherent Doppler lidar system 4. Erbium-doped fiber amplifier in a multi-channel lidar 5. Correlation properties of a multi-channel lidar system 6. Benefits of a multi-channel coherent Doppler lidar 7. Conclusions A. Wind measurement examples B. Range gate weighting function for different pulse shapes / Diese Arbeit befasst sich mit der Fernerkundungstechnologie zur Erfassung der Windgeschwindigkeit und den damit verbundenen Phänomenen. Die heterodyne Detektion der laserbasierten gepulsten Systeme ermöglicht eine sehr sensitive Messung von Partikelgeschwindigkeiten in der Atmosphäre. Sowohl die Theorie der heterodynen Detektion als auch die wesentlichen Eigenschaften und Grundsätze werden in dieser Arbeit diskutiert. Ein neuartiges faserbasiertes kohärentes Doppler Mehrkanal-Lidarsystem wurde entwickelt, vorgestellt, charakterisiert und validiert. Dabei ist der faserbasierte Verstärker eine wesentliche Komponente in diesem System und wird somit in Bezug auf Nichtlinearitäten und Impulsverzerrungen besonders analysiert. Eine neuartige geregelte Impulsformung wurde entwickelt, um die Impulsenergie und die Impulsform unabhängig für jeden Kanal zu stabilisieren. Jeder einzelne Kanal erzeugt ein unabhängiges Specklebild, welches durch Messungen am festen Ziel und durch atmosphärische Messungen präsentiert und mit theoretischen Berechnungen verglichen wird. Somit wird unter der Verwendung aller vier Kanäle des Systems die Geschwindigkeitsgenauigkeit erhöht, welches durch atmosphärische Messungen dargelegt wird. Zudem wird eine Korrelationstechnik vorgestellt, die das SNR bei höherer Rückstreuintensität weiter erhöht. Das Mehr-Wellenlängensystem wurde für vier Kanäle ausgelegt, welches eine Beobachtung der Atmosphäre entlang der Sichtlinie unter sonst gleichen Bedingungen ermöglicht. Diese Kanäle können sich sowohl in der Impulsform, in der Impulslänge als auch in der Pulsfolgefrequenz unterscheiden. Diese Arbeit zeigt die wesentlichen Vorteile eines solchen Mehrkanal-Lidarsystems mit atmosphärischen Messungen und wissenschaftlichen Ausarbeitungen.:1. Introduction to the subject 2. Introduction to pulsed coherent Doppler lidar 3. Methodology: Multi-channel coherent Doppler lidar system 4. Erbium-doped fiber amplifier in a multi-channel lidar 5. Correlation properties of a multi-channel lidar system 6. Benefits of a multi-channel coherent Doppler lidar 7. Conclusions A. Wind measurement examples B. Range gate weighting function for different pulse shapes