O oceano desempenha papel fundamental na configuração e manutenção do clima da Terra, sendo considerado um dos componentes principais do sistema climático.Diversos estudo foram conduzidos para avaliar as mudanças nos processos climáticos e como o clima, em contrapartida, é afetado por tais mudanças. O presente trabalho visa investigar o impacto das mudanças climáticas na formação de massas de água do oceano austral. Foram analisados resultados de simulação numérica para os séculos XX e XXI pelo modelo CCSM3 para os cenários 20c3m e SRESA1B do IPCC. Através da técnica de separação de mássas de água Análise Otimizada de Parâmetros Múltiplos (OMP) foram identificadas 3 massas de água no Mar de Ross: Água Profunda Circumpolar (CDW); Água da Plataforma de Gelo (ISW); Água de Plataforma de Baixa Salinidade (LSSW). A ISW, precursora da Água de Fundo Antártica (AABW), apresenta maior variação espacial tornando-se mais rasa no século XX e assumindo camadas mais profundas no século XXI. A variação da ISW está relacionada à variação do Modo Anular Sul (SAM) e à variação do gelo marinho. / It has been known for a long time that the ocean plays the most important role on Earth\'s heat budget, what turns it into a major component of the global climate system. Therefore, many studies have been made to assess whether features of climate processes are changing and how may climate itself be affected by these changes. This work aims to look at the impact of climate changes on water masses formation in the Southern Ocean. Results from the 20th century and SRESA1b CCSM3/NCAR simulation (1870 to 2100) were analyzed using the Optimum Multiparameter Analysis (OMP) to separate water masses. Three water masses were identified in the Ross Sea: Circumpolar Deep Water (CDW); Ice Shelf Water (ISW); Low Salinity Shelf Water (LSSW). Simulation results have shown that the ISW gets shallower during the 20th century and then, during the 21stcentury, it gets deeper and occupies the deepest layer by 2100 while it flows towards higher latitudes as AABW. Much closely to what has been shown by observational studies, water masses formation in the Southern Ocean is intrinsically linked to atmospheric vaiability modes, such as the southern annular mode--SAM, and to sea ice variation.

Mapping ice and open water in ocean bodies is important for numerous purposes including environmental analysis and ship navigation. The Canadian Ice Service (CIS) currently has several expert ice analysts manually generate ice maps on a daily basis. The CIS would like to augment their current process with an automated ice-water discrimination algorithm capable of operating on dual-pol synthetic aperture radar (SAR) images produced by RADARSAT-2. Automated methods can provide mappings in larger volumes, with more consistency, and in finer resolutions that are otherwise impractical to generate. We have developed such an automated ice-water discrimination system called MAGIC. The algorithm first classifies the HV scene using the glocal method, a hierarchical region-based classification method. The glocal method incorporates spatial context information into the classification model using a modified watershed segmentation and a previously developed MRF classification algorithm called IRGS. Second, a pixel-based support vector machine (SVM) using a nonlinear RBF kernel classification is performed exploiting SAR grey-level co-occurrence matrix (GLCM) texture and backscatter features. Finally, the IRGS and SVM classification results are combined using the IRGS approach but with a modified energy function to accommodate the SVM pixel-based information. The combined classifier was tested on 61 ground truthed dual-pol RADARSAT-2 scenes of the Beaufort Sea containing a variety of ice types and water patterns across melt, summer, and freeze-up periods. The average leave-one-out classification accuracy with respect to these ground truths is 95.8% and MAGIC attains an accuracy of 90% or above on 88% of the scenes. The MAGIC system is now under consideration by CIS for operational use.

Scene classification is performed on countless remote sensing images in support of operational activities. Automating this process is preferable since manual pixel-level classification is not feasible for large scenes. However, developing such an algorithmic solution is a challenging task due to both scene complexities and sensor limitations. The objective is to develop efficient and accurate unsupervised methods for classification (i.e., assigning each pixel to an appropriate generic class) and for labeling (i.e., properly assigning true labels to each class). Unique from traditional approaches, the proposed bootstrap approach achieves classification and labeling without training data. Here, the full image is partitioned into subimages and the true classes found in each subimage are provided by the user. After these steps, the rest of the process is automatic. Each subimage is individually classified into regions and then using the joint information from all subimages and regions the optimal configuration of labels is found based on an objective function based on a Markov random field (MRF) model. The bootstrap approach has been successfully demonstrated with SAR sea-ice and lake ice images which represent challenging scenes used operationally for ship navigation, climate study, and ice fraction estimation. Accuracy assessment is based on evaluation conducted by third party experts. The bootstrap method is also demonstrated using synthetic and natural images. The impact of this technique is a repeatable and accurate methodology that generates classified maps faster than the standard methodology.

Die in dieser Arbeit präsentierten Genomanalysen erweitern das Wissen um das genomische Potential der Roseobacter-Gruppe und zeigen mögliche Adaptionen an ökologische Nischen innerhalb mariner Lebensräume auf. In den polaren Meereisorganismen Octadecabacter arcticus 238 und O. antarcticus 307 konnten neue Eigenschaften identifiziert werden, welche bislang nicht in Vertretern der Roseobacter-Gruppe beschrieben wurden und wahrscheinlich Anpassungen an polare bzw. Meereis-assoziierte Lebensräume darstellen. Ein besonderes Highlight dieser Analysen ist die Charakterisierung einer neuen Untergruppe von Xanthorhodopsinen in den Octadecabacter-Vertretern. Diese neue Xantho¬rhodopsin-Unter¬gruppe unterscheidet sich von den bisher beschriebenen Xantho¬rhodopsinen nicht nur durch phylogenetische Verwandtschaftsbeziehungen, sondern auch in ihrer mangelnden Befähigung zur Keto-Carotenoid-Bindung und ihrer vorwiegenden Verbreitung in Organismen Eis-assoziierter Habitate. Für beide polare Octadecabacter-Vertreter wurde eine ungewöhnlich hohe Genom-plastizität festgestellt. Hierbei scheint es sich um eine Anpassung an das einzigartige Meereis¬habitat dieser Organismen zu handeln, welches als hot spot für horizontalen Gentransfer (HGT) gilt. Zudem bietet diese Genomplastizität eine Erklärung für die zahlreichen genomischen Unterschiede zwischen den Octadecabacter-Stämmen, welche in direktem Widerspruch zu der nahen Verwandtschaft dieser Organismen auf 16S rRNA-Gen¬sequenz¬ebene stehen. Trotz dieser Unterschiede weist die genetische Ausstattung von O. arcticus und O. antarcticus auffällige Übereinstimmungen auf, welche auf einen gemeinsamen exklusiven Genpool von Octadecabacter-Vertretern beider Polargebiete hindeuten. Dies wird durch 16S rRNA-basierte phylogenetische Analysen von Octadecabacter-Vertretern verschiedener Habitate unterstützt. Somit scheint zwischen Bakteriengemeinschaften beider Polarregionen eine direkte Verbindung zu existieren. Von den Polargebieten ausgehende Tiefenströmungen, welche sich über beide Hemisphären erstrecken, könnten diese Verbindung darstellen. Anhand der bislang verfügbaren Genomsequenzen wurden Verwandtschaftsbeziehungen sowie allgemeine Unterschiede zwischen Vertretern der Roseobacter-Gruppe auf vielfältigen Ebenen untersucht. Die Ergebnisse dieser Analysen geben wertvolle Einblicke in unterschiedliche Nischenadaptionen zwischen nah verwandten Roseobacter-Vertretern und in die Bedeutung von horizontalem Gentransfer für diese Gruppe. Zudem bieten sie eine Grundlage für die vereinfachte Einteilung und Analyse zukünftiger Roseobacter-assoziierter Genom– und Metagenomsequenzen.

Mapping ice and open water in ocean bodies is important for numerous purposes including environmental analysis and ship navigation. The Canadian Ice Service (CIS) currently has several expert ice analysts manually generate ice maps on a daily basis. The CIS would like to augment their current process with an automated ice-water discrimination algorithm capable of operating on dual-pol synthetic aperture radar (SAR) images produced by RADARSAT-2. Automated methods can provide mappings in larger volumes, with more consistency, and in finer resolutions that are otherwise impractical to generate. We have developed such an automated ice-water discrimination system called MAGIC. The algorithm first classifies the HV scene using the glocal method, a hierarchical region-based classification method. The glocal method incorporates spatial context information into the classification model using a modified watershed segmentation and a previously developed MRF classification algorithm called IRGS. Second, a pixel-based support vector machine (SVM) using a nonlinear RBF kernel classification is performed exploiting SAR grey-level co-occurrence matrix (GLCM) texture and backscatter features. Finally, the IRGS and SVM classification results are combined using the IRGS approach but with a modified energy function to accommodate the SVM pixel-based information. The combined classifier was tested on 61 ground truthed dual-pol RADARSAT-2 scenes of the Beaufort Sea containing a variety of ice types and water patterns across melt, summer, and freeze-up periods. The average leave-one-out classification accuracy with respect to these ground truths is 95.8% and MAGIC attains an accuracy of 90% or above on 88% of the scenes. The MAGIC system is now under consideration by CIS for operational use.

Reactive halogen species (RHS) are responsible for ozone depletion and oxidation of gaseous elemental mercury and dimethyl sulphide in the polar boundary layer, but the sources and mechanisms controlling their catalytic reaction cycles are still not completely understood. To further investigate these processes, ground– based Multi–Axis Differential Optical Absorption Spectroscopy (MAX-DOAS) observations of boundary layer BrO and IO were made from a portable instrument platform in McMurdo Sound during the Antarctic spring of 2006 and 2007. Measurements of surface ozone, temperature, pressure, humidity, and wind speed and direction were also made, along with fourteen tethersonde soundings and the collection of snow samples for mercury analysis. A spherical multiple scattering Monte Carlo radiative transfer model (RTM) was developed for the simulation of box-air-mass-factors (box-AMFs), which are used to determine the weighting functions and forward model differential slant column densities (DSCDs) required for optimal estimation. The RTM employed the backward adjoint simulation technique for the fast calculation of box-AMFs for specific solar zenith angles (SZA) and MAX-DOAS measurement geometries. Rayleigh and Henyey-Greenstein scattering, ground topography and reflection, refraction, and molecular absorption by multiple species were included. Radiance and box-AMF simulations for MAX-DOAS measurements were compared with nine other RTMs and showed good agreement. A maximum a posteriori (MAP) optimal estimation algorithm was developed to retrieve trace gas concentration profiles from the DSCDs derived from the DOAS analysis of the measured absorption spectra. The retrieval algorithm was validated by performing an inversion of artificial DSCDs, simulated from known NO2 profiles. Profiles with a maximum concentration near the ground were generally well reproduced, but the retrieval of elevated layers was less accurate. Retrieved partial vertical column densities (VCDs) were similar to the known values, and investigation of the averaging kernels indicated that these were the most reliable retrieval product. NO₂ profiles were also retrieved from measurements made at an NO₂ measurement and profiling intercomparison campaign in Cabauw, Netherlands in July 2009. Boundary layer BrO was observed on several days throughout both measurement periods in McMurdo Sound, with a maximum retrieved surface mixing ratio of 14.4±0.3 ppt. The median partial VCDs up to 3km were 9.7±0.07 x 10¹² molec cm ⁻ in 2007, with a maximum of 2.3±0.07 x 10¹³ molec cm⁻², and 7.4±0.06 x 10¹² molec cm⁻² in 2006, with a maximum of 1.05 ± 0.07 x 1013 molec cm⁻². The median mixing ratio of 7.5±0.5 ppt for 2007 was significantly higher than the median of 5.2±0.5 ppt observed in 2006, which may be related to the more extensive first year sea ice in 2007. These values are consistent with, though lower than estimated boundary layer BrO concentrations at other polar coastal sites. Four out of five observed partial ozone depletion events (ODEs) occurred during strong winds and blowing snow, while BrO was present in the boundary layer in both stormy and calm conditions, consistent with the activation of RHS in these two weather extremes. Air mass back trajectories, modelled using the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model, indicated that the events were locally produced rather than transported from other sea ice zones. Boundary layer IO mixing ratios of 0.5–2.5±0.2 ppt were observed on several days. These values are low compared to measurements at Halley and Neumayer Stations, as well as mid-latitudes. Significantly higher total mercury concentrations observed in 2007 may be related to the higher boundary layer BrO concentrations, but further measurements are required to verify this.

In dieser Arbeit werden Konzepte für die Diagnostik der großskaligen Zirkulation in der Troposphäre und Stratosphäre entwickelt. Der Fokus liegt dabei auf dem Energiehaushalt, auf der Wellenausbreitung und auf der Interaktion der atmosphärischen Wellen mit dem Grundstrom. Die Konzepte werden hergeleitet, wobei eine neue Form des lokalen Eliassen-Palm-Flusses unter Einbeziehung der Feuchte eingeführt wird. Angewendet wird die Diagnostik dann auf den Reanalysedatensatz ERA-Interim und einen durch beobachtete Meerestemperatur- und Eisdaten angetriebenen Lauf des ECHAM6 Atmosphärenmodells. Die diagnostischen Werkzeuge zur Analyse der großskaligen Zirkulation sind einerseits nützlich, um das Verständnis der Dynamik des Klimasystems weiter zu fördern. Andererseits kann das gewonnene Verständnis des Zusammenhangs von Energiequellen und -senken sowie deren Verknüpfung mit synoptischen und planetaren Wellensystemen und dem resultierenden Antrieb des Grundstroms auch verwendet werden, um Klimamodelle auf die korrekte Wiedergabe dieser Beobachtungen zu prüfen. Hier zeigt sich, dass die Abweichungen im untersuchten ECHAM6-Modelllauf bezüglich des Energiehaushalts klein sind, jedoch teils starke Abweichungen bezüglich der Ausbreitung von atmosphärischen Wellen existieren. Planetare Wellen zeigen allgemein zu große Intensitäten in den Eliassen-Palm-Flüssen, während innerhalb der Strahlströme der oberen Troposphäre der Antrieb des Grundstroms durch synoptische Wellen verfälscht ist, da deren vertikale Ausbreitung gegenüber den Beobachtungen verschoben ist. Untersucht wird auch der Einfluss von arktischen Meereisänderungen ausgehend vom Bedeckungsminimum im August/September bis in den Winter. Es werden starke positive Temperaturanomalien festgestellt, welche an der Oberfläche am größten sind. Diese führen vor allem im Herbst zur Intensivierung von synoptischen Systemen in den arktischen Breiten, da die Stabilität der troposphärischen Schichtung verringert ist. Im darauffolgenden Winter stellen sich barotrope bis in die Stratosphäre reichende Änderungen der großskaligen Zirkulation ein, welche auf Meereisänderungen zurückzuführen sind. Der meridionale Druckgradient sinkt und führt so zu einem Muster ähnlich einer negativen Phase der arktischen Oszillation in der Troposphäre und einem geschwächten Polarwirbel in der Stratosphäre. Diese Zusammenhänge werden ebenfalls in einem ECHAM6-Modelllauf untersucht, wobei vor allem der Erwärmungstrend in der Arktis zu gering ist. Die großskaligen Veränderungen im Winter können zum Teil auch im Modelllauf festgestellt werden, jedoch zeigen sich insbesondere in der Stratosphäre Abweichungen für die Periode mit der geringsten Eisausdehnung. Die vertikale Ausbreitung planetarer Wellen von der Troposphäre in die Stratosphäre ist in ECHAM6 mit sehr großen Abweichungen wiedergegeben. Somit stellt die Wellenausbreitung insgesamt den größten in dieser Arbeit festgestellten Mangel in ECHAM6 dar. / In this study concepts for the diagnostics of the large-scale circulation in the troposphere and the stratosphere are developed. Therefore the energy budget, wave propagation and the interaction between waves and the mean flow are analyzed. A corresponding set of diagnostic methods is derived. Furthermore a new type of localized Eliassen Palm Fluxes including moisture fluxes is introduced. These diagnostic methods are then applied to the ERA-Interim reanalysis and to a run of the ECHAM6 atmospheric model forced with observed sea surface temperatures and sea ice data. The diagnostics of the large scale circulation are useful to enhance the understanding of the climate system dynamics. Furthermore the knowledge of the relation between energy sources and sinks, atmospheric waves on planetary and synoptic scales and their forcing of the mean flow is applicable to validate global climate models. The results presented here show small deviations in terms of the energy balance in ECHAM6 but large discrepancies in terms of wave propagation. On the one hand Eliassen Palm fluxes connected to planetary waves are generally too strong. On the other hand the mean flow forcing within upper tropospheric jet streams by synoptic scale waves does not agree with observations, since the vertical propagation is shifted. A second part of this study analyses the influence of Arctic sea ice anomalies at the sea ice minimum in August/September on atmospheric conditions. Strong positive temperature anomalies with their maximum at the surface are observed. In autumn they lead to intensified synoptic scale systems, because of a reduced atmospheric vertical stability. A large-scale barotropic circulation pattern up to the stratosphere appears in winter related to previous late summer sea ice changes. A reduced meridional pressure gradient leads to a pattern related to a negative phase of the Arctic Oscillation in the troposp here as well as related to a weaker stratospheric polar vortex. The same analysis performed with an ECHAM6 model run shows a too small warming of Arctic latitudes. While tropospheric changes in the Arctic are covered by the model to some degree, the stratosphere shows large discrepancies in reproducing the observed changes in the low ice period. The vertical propagation of planetary waves from the troposphere into the stratosphere is reproduced with large differences. Accordingly this study shows the largest errors in ECHAM6 related to atmospheric wave propagation.

Scene classification is performed on countless remote sensing images in support of operational activities. Automating this process is preferable since manual pixel-level classification is not feasible for large scenes. However, developing such an algorithmic solution is a challenging task due to both scene complexities and sensor limitations. The objective is to develop efficient and accurate unsupervised methods for classification (i.e., assigning each pixel to an appropriate generic class) and for labeling (i.e., properly assigning true labels to each class). Unique from traditional approaches, the proposed bootstrap approach achieves classification and labeling without training data. Here, the full image is partitioned into subimages and the true classes found in each subimage are provided by the user. After these steps, the rest of the process is automatic. Each subimage is individually classified into regions and then using the joint information from all subimages and regions the optimal configuration of labels is found based on an objective function based on a Markov random field (MRF) model. The bootstrap approach has been successfully demonstrated with SAR sea-ice and lake ice images which represent challenging scenes used operationally for ship navigation, climate study, and ice fraction estimation. Accuracy assessment is based on evaluation conducted by third party experts. The bootstrap method is also demonstrated using synthetic and natural images. The impact of this technique is a repeatable and accurate methodology that generates classified maps faster than the standard methodology.

A quase inacessibilidade de grandes partes do Oceano Austral torna o conhecimento da espessura do gelo marinho limitado. Esta informação é essencial para a determinação do balanço de massa deste componente da criosfera. Na transição do inverno para a primavera de 2006, uma equipe de pesquisadores, coletou uma série de perfis de espessura de gelo marinho no norte e noroeste do mar de Weddell. Eles estavam a bordo do navio de pesquisa alemão Polarstern do Alfred-Wegener-Institut, Helmholtz-Zentrum für Polar- und Meeresforschung (AWI) e utilizaram um sistema composto por um sensor eletromagnético, altímetro laser e um Sistema de Posicionamento Global Diferencial (DGPS) aerotransportado por helicóptero. Simultaneamente, a Agência Espacial Europeia (ESA), adquiriu imagens ENVISAT ASAR WSM da área de pesquisa. O objetivo principal desta tese é determinar a relação existente entre o retroespalhamento medido pelo Radar de Abertura Sintética e a espessura do gelo marinho obtida pelo HEM (Helicopter-borne ElectroMagnetic system) usando os dados citados acima. Utilizamos os programas de computador NEST®, MATLAB®, EXCEL®, ArcGIS®/ArcMAP®, para processar, analisar e selecionar as imagens, para determinar a relação entre o retroespalhamento e as medidas, quase-tempo-coincidentes, de espessura do gelo. Projetamos as trajetórias dos voos sobre as imagens obtidas nas mesmas datas e extraímos os pixels referentes aos locais onde foram medidas as espessuras de gelo marinho. Apropriamos os valores de espessura do gelo para a área de cada pixel sobre o qual se referiam. Uma análise estatística determinou que o parâmetro que melhor representa a espessura do gelo dentro da área do pixel é a média. A regressão linear é a melhor forma de ajuste das relações entre o valor de retroespalhamento do pixel e a espessura do gelo marinho contido na área deste pixel. O coeficiente de correlação linear de Pearson, resultante de análise paramétrica, indica uma forte correlação (0,75) entre retroespalhamento e espessura do gelo marinho. Porém, a análise não paramétrica de Spearman resultou em um coeficiente de correlação baixo (0,06) o que pode indicar que os dados analisados são compostos por duas populações distintas (e.g., gelo de primeiro ano e plurianual). Contudo, a análise não paramétrica de Kolmogorov-Smirnov aventa a possibilidade de que não tenhamos amostrado toda população. Esta seria a razão de não haver valores de espessuras relativas ao intervalo entre -9.21dB e -1.35dB, o que poderia induzir ao baixo valor do coeficiente de correlação na análise de Spearman. Aplicamos a equação linear: y=0,6345x+12,015 às imagens e pudemos separá-las em doze classes: uma para água e onze para gelo marinho, estas com intervalos de um metro. Isto possibilitou a observação de importantes feições oceanográficas como: canais de águas abertas; cristas de compressão, decaimento do gelo marinho, deslocamento de icebergs (como indicadores de correntes, marés e ventos), liberação de gelo por geleiras de maré, desprendimento de iceberg da plataforma de gelo Larsen C, esteiras de ondas e vórtices oceânicos. Com base no exposto, podemos afirmar que, estatisticamente, nossos resultados são robustos e significantes, com nível de confiança entre 95% e 99%. A equação que propomos é um primeiro passo para inferir-se a espessura do gelo marinho a partir de coeficientes de retroespalhamento SAR. / The almost inaccessibility of large parts of the Southern Ocean makes the knowledge on the sea-ice thickness limited. This information is essential for determining the mass balance of this cryosphere component. During the transition from winter to spring 2006, a researcher team performed several sea-ice thickness profiles in the north and northwest of the Weddell Sea. They were aboard the German research vessel Polarstern from the Alfred-Wegener-Institute, Helmholtz-Zentrum für Polar- und Meeresforschung (AWI) and utilizing for those measurements a Helicopter-borne Electromagnetic system (HEM). The HEM is composed of an electromagnetic sensor, laser altimeter, and a Differential Global Positioning System (DGPS). Simultaneously, the European Space Agency (ESA) ENVISAT ASAR WSM acquired images from the research area. The main objective of this thesis is to determine the relationship between the Synthetic Aperture Radar backscattering and the HEM’s sea-ice thickness measurements using data listed above. We used the computer programs: NEST®, MATLAB®, EXCEL®, ArcGIS® / ArcMap®, to process, analyze and select images, in order to determine the relationship between the backscattering and the quasi-time-coincident ice thickness measurements. We projected the flights trajectories on the images obtained on the same dates, extracting the pixels pertaining to the places where the sea-ice thicknesses were determined. We ascribed ice thickness values for each area covered by the pixel. A statistical analysis determined that the best ice thickness parameter within a pixel area is its mean. Linear regression is the best way to adjust the relationship between the pixel backscatter value and the sea-ice thickness contained within the pixel area. The Pearson linear correlation coefficient, resulting from parametric analysis, indicates a strong correlation (0.75) between backscatter and sea-ice thickness. However, the nonparametric Spearman analysis resulted in a low correlation coefficient (0.06), which may indicate that the analyzed data consist of two distinct populations (e.g., first-year and multi-year ice). However, the Kolmogorov-Smirnov nonparametric analysis brought up the possibility that we just have not sampled the entire population. This could explain the no existence of sea-ice thicknesses values on the interval from -9.21dB to -1.35dB, which could lead to the low correlation coefficient in the Spearman analysis. We applied the linear equation: =,+, to the images, separating them into twelve classes: one for water and eleven to sea-ice, the latter in one-meter thickness intervals. The results enabled the observation of important oceanographic features such as open water channels, pressure ridges, sea-ice decay, icebergs motion (as indicators of currents, tides and winds), glaciers discharge, iceberg calving from Larsen C ice shelf, wakes and oceanic eddies. Based on the foregoing, we can say that, statistically, our results are robust and significantly, with a confidence level from 95% to 99%. The proposed equation is a first step to inferring sea-ice thickness from SAR backscatter coefficients.

La fonte rapide des glaces de l'Arctique dans le contexte actuel du réchauffement climatique est un sujet scientifique majeur de ces 30 dernières années. L'Arctique joue un rôle fondamental dans l'équilibre du climat et requiert une attention particulière. Les régions arctiques sont alors surveillées par des observations satellitaires et des mesures in-situ. L'impact climatique de la fonte totale de la glace arctique est encore spéculatif. Des recherches sont donc nécessaires pour le suivi à long terme de l'Océan Arctique, en particulier la dynamique spatio-temporelle de la couverture de glace et ses conséquences sur les écosystèmes. Notre travail s'inscrit dans ce contexte, et est porté sur le paysage sonore des régions polaires avant leur possible industrialisation qui accompagnera la fonte de la glace. Ainsi, nous avons d'abord examiné les conséquences de la disparition du couvert de glace sur les paysages sonores de ces régions. Nous avons alors étudié les variations saisonnières du bruit de fond et ses pilotes environnementaux. De ce fait, nous avons développé un algorithme d'estimation du bruit ambiant afin de pouvoir constituer des séries temporelles à partir des données acoustiques longue durée. Deuxièmement, nous avons étudié les transitoires générés par le comportement mécanique de la banquise en Arctique. Cette étude vise d'une part à comprendre le mécanisme de production de ces transitoires sous la glace, et d'autre part à évaluer leur potentiel comme moyen d'observation de la dynamique de la glace de mer. / The Arctic sea ice melting, in the global warming context, has become a major scientific topic during the last 30 years. The Arctic Ocean plays a fundamental role in the global climate balance and requires a particular attention. The Arctic Regions are then monitored by satellite observations and in-situ measurements. The climatic impact of the total melting of the Arctic sea ice is not yet understood and researches are still needed for long term monitoring of Arctic Ocean, particularly the dynamics of the ice cover and its consequences on the ecosystems. Our work focused on the natural soundscapes of these Polar Regions prior to their possible industrialization. So, we first examined the impact of climate warming alone on polar soundscapes by studying the seasonal variability of ambient noise and its environmental drivers. We then developed an ambient noise estimation algorithm for automatic extraction of this noise component from long term measurements. In second, we examined the acoustic transients generated by the mechanical behavior of the ice cover at its maximum extent. This aims to better understanding of the physical processes involved in under-ice noise production and their potential use for sea ice monitoring.