Les gyres subtropicaux sont au coeur des changements observés au cours des dernières décennies. On y observe entre la surface et la pycnocline permanente une augmentation du contenu thermique de l’océan. La pycnocline permanente délimite un important réservoir de chaleur et joue un rôle majeur en empêchant la chaleur accumulée en surface d’atteindre les profondeurs de l’océan. La pycnocline permanente est donc d’un intérêt important dans un contexte de changement climatique. Pour la première fois et grâce au réseau de données Argo, nous avons été capables de déterminer les propriétés de la pycnocline permanente. L’objectif de cette thèse est de déterminer la structure de la pycnocline permanente et d’étudier sa variabilité au cours des dernières décennies. Une méthode de détermination objective de la pycnocline permanente a été développée. Cette méthode a d’abord été appliquée à l’océan Atlantique nord avec les données Argo puis à l’océan global. Une structure complexe de la pycnocline permanente a été mise en évidence avec de fortes différences d’un gyre à l’autre. La pycnocline permanente est la plus profonde et la plus épaisse dans le gyre subtropical nord Atlantique. Cela explique que le gyre subtropical nord Atlantique soit le plus grand réservoir de chaleur au monde. Ensuite, les relations entre la variabilité du contenu de chaleur et les propriétés de la pycnocline permanente ont été étudiées en s’appuyant sur des réanalyses océaniques. Au cours des dernières décennies, un réchauffement important de l’océan a été observé et particulièrement dans l’océan Atlantique nord. Ce réchauffement est principalement dominé par un approfondissement des isopycnes. Les déplacements verticaux des isopycnes induisent des changements dans la stratification et affectent les propriétés de la pycnocline permanente (profondeur et densité potentielle). / Subtropical gyres are central to the observed climate changes throughout the last decades. It is observed between the surface and the permanent pycnocline an intense increase in the ocean heat content. The permanent pycnocline delineates thus an important heat reservoir. The permanent pycnocline has a major role in preventing heat to reach the deep ocean and it thus of a relative importance in the context of climate change. For the first time and thanks to the development of the Argo array, we have been able to characterize the observed structure of the permanent pycnocline. The objective of this PhD thesis is to investigate the structure of the permanent pycnocline and its variability over the last decades. We developed an objective method to characterize the properties of the permanent pycnocline. This method has been first applied to the North Atlantic Ocean with Argo data and then to the global ocean. A complex structure of the permanent pycnocline emerges with strong differences from one gyre to another. The permanent pycnocline is found to be the deepest and the thickest in the North Atlantic subtropical gyre. It implies that the North Atlantic subtropical gyre is the largest heat reservoir on Earth. Then, ocean reanalyses have been used to investigate the changes in the permanent pycnocline properties in the North Atlantic subtropical gyre. Over the last decades, there is a strong warming of the upper ocean, especially in the North Atlantic subtropical gyre. The warming in the ocean is dominated by the heaving of isopycnal surfaces. This heaving strongly affects the depths of isopycnals and the stratification. This in turn affects the properties of the permanent pycnocline, especially its depth and potential density.

Le but de cette thèse est d’analyser les impacts de la variabilité atmosphérique grande échelle sur la circulation océanique. Ceci a déjà fait l’objet de nombreuses publications, dans lesquelles la variabilité atmosphérique est analysée en termes de modes de variabilité, déterminés par analyse en composantes principales (EOF en anglais) des anomalies de pression de surface. Ces modes sont l’Oscillation Nord-Atlantic (NAO), le Pattern Est-Atlantique (EAP) et le Pattern Scandinave (SCAN). La décomposition en EOF implique que les modes sont orthogonaux et symétriques. Cette dernière hypothèse a été montrée comme étant invalide pour la NAO. Par conséquent, un nouveau concept est proposé dans cette étude pour estimer la variabilité atmosphérique, celui des régimes de temps. Ces derniers sont des structures spatiales de grande échelle, récurrents et quasi-Stationnaires qui permettent de capturer la variabilité des forçages atmosphériques. De plus, ils permettent de séparer les patterns spatiaux des deux phases de la NAO. Ces régimes de temps sont donc une alternative prometteuse pour l’analyse de la variabilité océanique forcée par l’atmosphère. A partir d’observation et de modèles numériques (réalistes ou idéalisés), nous avons montré que les régimes Atlantic Ridge (AR), NAO− et NAO+ induisent une réponse rapide (échelles mensuelles à interannuelles) des gyres subtropical et subpolaire (via un mécanisme de Sverdrup topographique) et de la cellule de retournement (MOC, ajustement aux anomalies de transport d’Ekman). Aux échelles décennales, le gyre subpolaire s’intensifie lors de conditions NAO+ et BLK persistantes via un ajustement barocline aux flux de flottabilité et s’affaiblit pour AR via un ajustement barocline aux anomalies de rotationnel de vent. Ce dernier mécanisme explique aussi l’augmentation du gyre subtropical pour une NAO+ persistante et son affaiblissement pour un AR persistant. La réponse des gyres pour des conditions de NAO− persistantes est un déplacement vers le sud des gyres (l’intergyre gyre). L’intensité de la MOC est augmentée pour des conditions de NAO+ et BLK persistantes, dû à l’augmentation de la formation d’eau dense en mer du Labrador, et inversement pour NAO− et AR. Finalement, des bilans de contenu de chaleur dans la gyre subpolaire et les mers nordiques ont été effectués dans quatre modèles océaniques globaux. Les moyennes d’hiver de convergence océanique de chaleur dans la partie ouest de la gyre subpolaire sont positivement corrélées aux occurrences d’hiver de NAO−, ce qui est dû à la présence de l’intergyre, tandis que cette convergence est négativement corrélée aux occurrences d’AR, ce qui est dû à la réduction des deux gyres qui lui est associée. Les flux de chaleur vers l’océan dans la gyre subpolaire sont négativement corrélés aux occurrences d’hiver de la NAO+ et inversement pour la NAO−. Dans les mers Nordiques, ils sont positivement corrélés aux occurrences de BLK et, dans une moindre mesure, aux occurrences de AR. De plus, nous suggérons que la variabilité du contenu de chaleur dans la partie ouest du gyre subpolaire est la réponse décalée (lag de 6 ans) à l’intégration temporelle du forçage lié au régime NAO+, due à la combinaison de la réponse en phase (0-Lag) des flux de chaleur et à la réponse décalée (lag de 3 ans) de la convergence de chaleur. / The aim of the PhD is to investigate the impacts of the large-Scale atmospheric variability on the North- Atlantic ocean circulation. This question has already been addressed in a large number of studies, in which the atmospheric variability is decomposed into modes of variability, determined by decomposing sea-Level pressure anomalies into Empirical Orthogonal Function (EOFs). These modes of variability are the North-Atlantic Oscillation (NAO), the East-Atlantic Pattern (EAP) and the Scandinavian Pattern (SCAN). EOF decomposition assumes that the modes are orthogonal and symmetric. The latter assumption, however, has been shown to be inadequate for the NAO. Hence, a different framework is used in this study to assess the atmospheric variability, the so-Called weather regimes. These are large-Scale, recurrent and quasi-Stationary atmospheric patterns that have been shown to capture well the interannual and decadal variability of atmospheric forcing to the ocean. Furthermore, they allow to separate the spatial patterns of the positive and negative NAO phases. Hence, these weather regimes are a promising alternative to modes of variability in the study of the ocean response to atmospheric variability. Using observations and numerical models (realistic or in idealised settings), we have shown that the Atlantic Ridge (AR), NAO− and NAO+ regimes drive a fast (monthly to interannual) wind-Driven response of the subtropical and subpolar gyres (topographic Sverdrup balance) and of the meridional overturning circulation (MOC, driven by Ekman transport anomalies). At decadal timescales, the subpolar gyre strengthens for persistent NAO+ and Scandinavian Blocking (BLK) conditions via baroclinic adjustment to buoyancy fluxes and slackens for persistent AR conditions via baroclinic adjustment to wind-Stress curl anomalies. The latter mechanism also accounts for the strengthening of the subtropical gyre for persistent NAO+ conditions and its weakening for persistent AR conditions. The gyres response to persistent NAO− conditions reflects the southward shift of the gyre system (the intergyre gyre). The MOC spins-Up for persistent NAO+ and BLK conditions via increased deep water formation in the Labrador Sea, and conversely for the NAO− and AR regimes. Last, heat budget calculations in the subpolar gyre and the Nordic Seas have been performed using four global ocean hindcasts. The winter averaged heat convergence in the western subpolar gyre is positively correlated with the NAO− winter occurrences, which is due to the intergyregyre circulation, while it is negatively correlated with AR winter occurrences, because of the wind-Driven reduction of both gyres. Downward surface heat flux anomalies are negatively correlated with NAO+ occurrences, and conversely for the NAO−. In the Nordic Seas, they are positively correlated with BLK and to a lesser extent AR occurrences. Furthermore, we suggest that the heat content variability in the western subpolar gyre is the signature of the delayed response (6-Year lag) to the time-Integrated NAO+ forcing, due to the combination of the immediate (0-Lag) response of surface heat flux and the lagged (3 year lag) response of ocean heat convergence.

Submitted by Fabio Sobreira Campos da Costa (fabio.sobreira@ufpe.br) on 2016-03-18T13:04:49Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Tese_Gbekpo_Aubains_2015.pdf: 10869276 bytes, checksum: 0b22521ef971c75f3112dcbc74fcbb7a (MD5) / Made available in DSpace on 2016-03-18T13:04:50Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Tese_Gbekpo_Aubains_2015.pdf: 10869276 bytes, checksum: 0b22521ef971c75f3112dcbc74fcbb7a (MD5) Previous issue date: 2015-04-08 / As interações do sistema oceano-atmosfera no Atlântico tropical e suas contribuições à grande variabilidade da precipitação ao longo da costa do nordeste do Brasil (NEB) foram investigadas para os anos de 1974-2008. Os núcleos das estações chuvosas de Março-Abril e de Junho-Julho foram identificados para a parte norte do Nordeste do Brasil (NNEB) e a parte leste do Nordeste do Brasil (ENEB), respectivamente. As regressões lineares defasadas entre as anomalias da Temperatura da Superfície do Mar (TSM), da Pseudo tensão de cisalhamento de vento (PWS), do Fluxo de calor latente (LHF), da Umidade especifica do ar, e as anomalias (positivas e negativas) de precipitação forte no NNEB e no ENEB mostram que a variabilidade da precipitação dessas regiões é diferentemente influenciada pela dinâmica do Atlântico tropical. Quando a zona de convergência intertropical (ZCIT) é anormalmente deslocada para o sul alguns meses antes da estação chuvosa do NNEB, a fase negativa do Modo Meridional do Atlântico (AMM) (fortalecimento dos ventos alísios do nordeste, relaxamento dos ventos alísios do sudeste, maior evaporação no hemisfério norte, menor evaporação no hemisfério sul, TSM mais fria no hemisfério norte, e TSM mais quente no hemisfério Sul), aumenta a precipitação durante a estação chuvosa. O efeito oposto ocorre durante a fase positiva do AMM. Além disso, o estudo mostra a grande influência e um efeito preditivo da região Noroeste do Atlântico Equatorial noroeste (NWEA) sobre a precipitação do NNEB. Com relação ao estado subsuperficial do oceano, os resultados indicam que uma camada de barreira mais fina na NWEA de Novembro-Dezembro até Março-Abril é associada ao resfriamento progressivo da TSM, ao reforço do componente meridional do vento nordeste e precipitações intensas sobre o NNEB. Já a influência da dinâmica do Atlântico tropical sobre a variabilidade da precipitação no ENEB em Junho-Julho indica uma propagação para noroeste de uma área de forte correlação positiva de TSM e de Umidade específica do ar, deslocando-se da parte sudeste do Atlântico tropical (de Fevereiro-Março) para a região da Piscina Quente do Atlântico Sudoeste (SAWP), situada ao largo do Brasil (Junho-Julho). A área de propagação das anomalias, observada segue globalmente o caminho do ramo sul da Corrente Sul Equatorial (sSEC), que é responsável pelo transporte de calor oceânico de leste para oeste no Atlântico tropical sul. O deslocamento da fase mais quente da advecção horizontal de calor oceânico, na camada de mistura, de leste da bacia (entre 5º-15ºS) para a costa do Brasil em Junho-Agosto corrobora a influência da sSEC sobre o núcleo da chuva do ENEB. Uma aceleração dos ventos alísios de sudeste, associada a uma convergência da anomalia do vento sobre a SAWP, produz excesso de umidade do ar sobre a região e provoca mais precipitação sobre ENEB. O efeito oposto ocorre para os episódios menos chuvosos. De acordo com o estudo, a SAWP se mostra como uma área de potencial para o estabelecimento de um índice de previsão de chuvas no ENEB. / Tropical Atlantic Ocean-atmosphere interactions and their contributions to strong variability of rainfall along the Northeast Brazilian coast (NEB) were investigated for the years 1974-2008. The core rainy seasons of March-April and June-July were identified for northern Northeast Brazil (NNEB) and eastern Northeast Brazil (ENEB), respectively. Lagged linear regressions between sea surface temperature (SST), pseudo wind stress (PWS), latent heat flux (LHF) and air specific humidity anomalies over the entire tropical Atlantic and strong rainfall anomalies in NNEB and ENEB show that the rainfall variability of these regions is differentially influenced by the dynamics of the tropical Atlantic. When the intertropical convergence zone (ITCZ) is abnormally displaced southward a few months prior to the NNEB rainy season, the associated meridional mode (strengthening of the northeast trade winds, relaxation of the southeast trade winds, strong evaporation in the north, weak evaporation in the south, colder SST in the North, and warmer SST in the South) increases precipitation during the rainy season. The opposite effect occurs during the positive phase of the dipole. Additionally, this study shows strong influence and predictive effect of the Northwestern Equatorial Atlantic (NWEA) on the NNEB rainfall. Thinner barrier layer in the NWEA from November-December to March-April is associated with progressive cooling of SST, strengthening of meridional component of the northeasterly wind and intense precipitations over the NNEB. The dynamical influence of the tropical Atlantic on the June-July ENEB rainfall variability shows a northwestward-propagating area of strong, positively correlated SST and air specific humidity from the southeastern tropical Atlantic (February-March) to the Southwestern Atlantic Warm Pool (SAWP) offshore of Brazil (June-July). The northwestward-propagating area, observed from February-March to June-July, follows the same pathway of the southern branch of south equatorial current (sSEC), which is responsible of the oceanic heat transport from east to west in the southern tropical Atlantic. The displacement of the warmest phase of horizontal advection of the oceanic mixed layer heat from the eastern part (between 5-15ºS) to the Brazilian coast in June-August confirms this influence of the sSEC on core rainy season in the ENEB. Furthermore, according to our study, the SAWP could be used as index of rainfall prediction in ENEB. An early acceleration of the southeasterly trade winds, associated with a strong convergence of the wind anomaly over the SAWP, produces excessive humidity over the region and causes more precipitation over ENEB. The opposite effect occurs for less rainy episodes.

No description available.