Today, ocean modeling is fast developing as a versatile tool for the study of earth’s climate, local marine ecosystems and coastal engineering applications. Though the field of ocean modeling began in the early 1950s along with the development of climate models and primitive computers, even today, the state-of-the-art ocean models have their own limitations. Many issues still remain such as the uncertainity in the parameterisation of essential processes that occur on spatial and temporal scales smaller than that can be resolved in model calculations, atmospheric forcing of the ocean and the boundary and initial conditions. The advent of data assimilation into ocean modeling has heralded a new era in the field of ocean modeling and oceanic sciences. “Data assimilation” is a methodology in which observations are used to improve the forecasting skill of operational meteorological models. The study in the present thesis mainly focuses on obtaining a four dimensional realization (the spatial description coupled with the time evolution) of the oceanic flow that is simultaneously consistent with the observational evidence and with the dynamical equations of motion and to provide initial conditions for predictions of oceanic circulation and tracer distribution. A good implementation of data assimilation can be achieved with the availability of large number of good quality observations of the oceanic fields as both synoptic and in-situ data. With the technology in satellite oceanography and insitu measurements advancing by leaps over the past two decades, good synoptic and insitu observations of oceanic fields have been achieved. The current and expected explosion in remotely sensed and insitu measured oceanographic data is ushering a new age of ocean modeling and data assimilation. The thesis presents results of analysis of the impact of data assimilation in an ocean general circulation model of the North Indian Ocean. In this thesis we have studied the impact of assimilation of temperature and salinity profiles from Argo floats and Sea Surface height anomalies from satellite altimeters in a Sigma-coordinate Indian Ocean model. An ocean data assimilation system based on the Regional Ocean Modeling System (ROMS) for the Indian Ocean is used. This model is implemented, validated and applied in a climatological simulation experiment to study the circulation in the Indian Ocean. The validated model is then used for the implementation of the data assimilation system for the Indian Ocean region. This dissertation presents the qualitative and quantitative comparisons of the model simulations with and without subsurface temperature and salinity profiles and sea surface height anamoly data assimilation for the Indian Ocean region. This is the first ever reported data assimilation studies of the Argo subsurface temperature and salinity profile data with ROMS in the Indian Ocean region.

L’archipel du Prince Édouard est situé dans le Courant Circumpolaire Antarctique, entre le Front Subantarctique (SAF) et le front polaire (PF). Les positions de ces fronts se sont révélées très variables aux échelles interannuelles et mensuelles et ont montré une tendance à long terme à migrer vers le sud dans la région. L'activité tourbillonnaire, observée à la dorsale sud-ouest indienne, en amont de l’archipel, a également affiché une variabilité interannuelle et intra-annuelle ainsi qu’une tendance à la baisse sur ces 24 dernières années. A une échelle plus locale, nous avons souligné que les conditions aux îles sont impactées par ces tourbillons mésoéchelle produits en amont. La marée a également été mise en évidence comme jouant un rôle non négligeable dans la variabilité de la circulation entre les îles. Une configuration idéalisée a été mise en place pour la région afin d’étudier les propriétés des tourbillons et les mécanismes physiques à l’origine de leur formation. L'énergie potentielle tourbillonnaire disponible a révélé un maximum d'énergie aux alentours de 800 m de profondeur, confirmant le caractère « deep-reaching » des tourbillons. Cette activité tourbillonnaire a été montrée comme étant le résultat d'une combinaison d’instabilités barotropes et baroclines. Enfin, nous avons enquêté sur les conséquences potentielles d'une migration vers le sud du SAF dans la région des îles. Le modèle étant idéalisé, il nous a permis de simuler un déplacement vers le sud en déplaçant les conditions initiales ainsi que les conditions aux limites. Le principal résultat a été la nette diminution de l'activité mésoéchelle dans la région, ce qui pourrait avoir un impact sur les écosystèmes de l’archipel du Prince Édouard. / The subantarctic Prince Edward Islands lie in the Antarctic Circumpolar Current, between the Subantarctic Front (SAF) and the Polar Front (PF). These fronts positions were found to be highly variable at interannual and monthly time scales and revealed a significant long-term southward trend in the region. The intense mesoscale activity, observed upstream the islands at the South West Indian Ridge, also showed an interannual and intra-annual variability as well a decrease in eddy kinetic energy over 24 years. At a more local scale, we highlighted that the archipelago’s environment was impacted by the mesoscale features produced upstream.Tides appeared to be another important driver of variability of the circulation in between the two islands.An idealised model configuration was designed for the Prince Edward Islands region to study the mesoscale eddy properties and the physical mechanisms of their formation at the ridge. The Eddy Available Potential Energy revealed a maximum of energy around 800 m depth, confirming the deep reaching characteristic of the eddy originated in the region and suggested the presence of a local energy source at this depth. This eddies activity was shown to be the result of a combination of barotropic and baroclinic instabilities occurring at the ridge.Finally, we investigated on the potential consequences of a southward shift of the SAF in the region of the islands.Because the model was idealised, it allowed us to simulate an SAF southward shift by shifting the initial and boundary conditions. The main result was the clear decrease of mesoscale activity in the region which could potentially impact the ecosystems of the Prince Edward Islands.

The most important component of the climate system over the Indian Ocean region is the southwest monsoon, which dictates the life and economy of billions of people in the tropics. Being a phenomena that involves interaction between atmosphere, ocean and land, the southwest monsoon is strongly influenced by upper ocean, primarily through warm sea surface temperature (SST). This is particularly true about the southeastern Arabian Sea (SEAS) and the onset of southwest monsoon over the peninsular India. A localized patch of warm water, known as the Arabian Sea mini warm pool (ASMWP), forms in the SEAS during February–March. It remain as the warmest spot in the northern Indian Ocean till early April. A large region, surrounding the SEAS, attains SST exceeding 30°C during April–May, with often the ASMWP as its core. The ASMWP is believed to have a critical impact on the air-sea interaction during the onset phase of southwest monsoon and on the formation of the onset vortex, during late May or early June. This thesis addresses the formation mechanisms of ASMWP, using a high-resolution Ocean General Circulation Model (OGCM) of the Indian Ocean. In addition to the formation of ASMWP, the SEAS is characterized by several features in its hydrography and circulation, which have been invoked in the past to explain the preferential warming of this oceanic region. During November–January, the prevailing surface currents transport low-salinity water from the Bay of Bengal into the SEAS and leads to strong haline stratification in the upper layer and formation of barrier layer (layer between mixed layer and isothermal layer). The vertical distribution of temperature in the SEAS exhibit inversions (higher subsurface temperature than that at surface) during December–February. A high in sea level and anticyclonic eddies develop in the SEAS during December and they propagate westward. These eddies modify the hydrography through downwelling and play an important role in the redistribution of advected low-salinity water within the SEAS. The seasonally reversing coastal and equatorial currents present in and around SEAS also have a major contribution in setting up the hydrography, through the advection and redistribution of cooler low-salinity water. These features make the SEAS a unique oceanographic region. The first hypothesis on the formation of ASMWP, which has been suggested by diagnostic studies, is based on the barrier layer mechanism. The barrier layer, caused by the influx of low-salinity water at surface, is argued to maintain a shallow mixed layer which can warm more efficiently. In addition, presence of barrier layer can prevent mixed layer cooling, by cutting off the interaction of mixed layer with cooler thermocline water below. However, a coupled model study have shown that there is no significant impact on the ASMWP formation from barrier layer, but only a weak warming effect during it mature phase during April. The second hypothesis, which is based on an OGCM study, has suggested that the temperature inversions present within the barrier layer can heat the mixed layer through turbulent entrainment and in turn lead to the formation of ASMWP during February–March. Both hypotheses rule out the possibility of air-sea heat fluxes being the primary reason in its formation. The strong salinity stratification in the SEAS during December–March is central to the hypotheses about formation of the ASMWP. Observational studies have only limited success in assessing the contribution from barrier layer and temperature inversions, as the ASMWP always form in their presence. OGCMs offer a better alternative. However, modelling processes in the northern Indian Ocean, especially that in the SEAS, is a challenging problem. Previous Indian Ocean models have had serious difficulties in simulating the low-salinity water in the Bay of Bengal and its intrusion into the SEAS. The northward advection of low-salinity water in the SEAS, along the west coast of India, is used to be absent in model simulations. Moreover, the coarse resolution inhibited those models from simulating faster surface currents and vigorous eddies as seen in the observations. In this thesis, we use an OGCM of the Indian Ocean, based on the recent version of Modular Ocean Model (MOM4p0), to study the ASMWP. The model has high resolutions in the horizontal (1/4o x 1/4o) and vertical (40 levels, with 5 m spacing in upper 60 m), and has been forced with daily values momentum, heat and freshwater fluxes. The turbulent (latent and sensible) and long wave heat fluxes have been calculated as a function of model SST. The freshwater forcing consists of precipitation, evaporation and river runoff, and there are no surface restoring or flux adjustments. The river runoff has been distributed over several grid points about the river mouth instead of discharging into a singe grid point, which has resulted in remarkable improvements in salinity simulation. The model simulates the Indian Ocean temperature, salinity and circulation remarkably well. The pattern of model temperature distribution and evolution matches very well with that in the observations. Significant improvements have been made in the salinity simulation, including the Bay of Bengal freshwater plume and intrusion of low-salinity water from the bay into the SEAS. The salinity distribution within the SEAS is also well represented in the model. The use of appropriate horizontal friction parameters has resulted in the simulation of realistic currents. The observed features in the SEAS, including the life cycle of the ASMWP, low-salinity water, barrier layer, temperature inversions, eddies and currents are well represented in the model. Present study has unraveled the processes involved in the life cycle of barrier layer and temperature inversions in the SEAS. Presence of low-salinity water is necessary for their formation. Barrier layer develops in the SEAS during November, after the intrusion of low-salinity water from the Bay of Bengal. The barrier layer is thickest during January–February, and it dissipates during March–April. The variations and peak of barrier layer thickness is controlled by variations in isothermal layer depth, which in turn is dominated by the downwelling effects of anticyclonic eddies. The intense solar heating during March–April leads to the formation of shallow isothermal layer and results in the dissipation of barrier layer. Temperature inversions starts developing in the SEAS during December, reaches its peak during January–February and dissipates in the following months. Advection of cooler low-salinity water over warmer salty water and penetrating shortwave radiation is found to cause temperature inversions within the SEAS, whereas winter cooling is also important to the north and south of the SEAS. There is significant variation in the magnitude, depth of occurrence and formation mechanisms of temperature inversions within the SEAS. Analysis of model mixed layer heat budget has shown that the SEAS SST is mainly controlled by atmospheric forcing, including the life cycle of ASMWP. It has also shown that the heating from temperature inversions do not contribute to the formation of ASMWP. In an experiment in which a constant salinity of 35 psu was maintained over the entire model domain, the ASMWP evolved very similar to that in the standard run, suggesting that the salinity effects are not necessary for the formation of ASMWP. Examination of wind field show that the winds over the SEAS during November–February are low due to the blocking of northeasterly winds by Western Ghats. Several process experiments by modifying the wind and turbulent heat fluxforcing fields have shown that these low winds lead to the formation of ASMWP in the SEAS during February–March. The low winds reduce latent heat loss, resulting in net heat gain by the ocean. This helps the SEAS to keep warmer SST while the surrounding region experience intense cooling under the strong dry northeasterly winds. As the winds are weak over the SEAS, the mixed layer is not able to feel the stratification beneath and the mixed layer depth is determined by solar heating, with or without salinity effects. In addition, the weak winds are not able to entrain the temperature inversions present in the barrier layer. The winds are weak during March–April too, and the air-sea heat fluxes dictate the SST evolution during this period. Therefore, during November–April, the SEAS acts as a low wind heat-dominated regime, where the evolution of sea surface temperature is solely determined by atmospheric forcing. We show that, in such regions, the evolution of surface layer temperature is not dependent on the characteristics of subsurface ocean, including the presence of barrier layer and temperature inversions.

Depuis ces trente dernières années, les systèmes d'observation de la Terre et les modèles numériques se sont perfectionnés et complexifiés pour nous fournir toujours plus de données, réelles et numériques. Ces données, de nature très diverse, forment maintenant un ensemble conséquent d'informations précises mais hétérogènes sur les structures et la dynamique des fluides géophysiques. Dans les années 1980, des méthodes d'optimisation, capables de combiner les informations entre elles, ont permis d'estimer les paramètres des modèles numériques et d'obtenir une meilleure prévision des courants marins et atmosphériques. Ces méthodes puissantes, appelées assimilation variationnelle de données, peinent à tirer profit de la toujours plus grande complexité des informations de par le manque de puissance de calcul disponible. L'approche, que nous développons, s'intéresse à l'utilisation des méthodes multigrilles, jusque là réservées à la résolution de systèmes d'équations différentielles, pour résoudre l'assimilation haute résolution de données. Les méthodes multigrilles sont des méthodes de résolution itératives, améliorées par des corrections calculées sur des grilles de plus basses résolutions. Nous commençons par étudier dans le cas d'un modèle linéaire la robustesse de l'approche multigrille et en particulier l'effet de la correction par grille grossière. Nous dérivons ensuite les algorithmes multigrilles dans le cadre non linéaire. Les deux types d'algorithmes étudiés reposent d'une part sur la méthode de Gauss Newton multigrille et d'autre part sur une méthode sans linéarisation globale : le Full Approximation Scheme (FAS). Ceux-ci sont appliqués au problème de l'assimilation variationnelle de données dans le cadre d'une équation de Burgers 1D puis d'un modèle Shallow-water 2D. Leur comportement est analysé et comparé aux méthodes plus traditionnelles de type incrémentale ou multi-incrémentale. / For these last thirty years, earth observation and numerical models improved greatly and provide now a huge amount of accurate, yet heterogeneous, information on geophysics fluids dynamics and structures. Optimization methods from the eighties called variational data assimilation are capable of merging information from different sources. They have been used to estimate the parameters of numerical models and better forecast oceanic and atmospheric flows. Unfortunately, these powerful methods have trouble making benefit of always more complex information, suffering from the lack of available powerful calculators. The approach developed here, focuses on the use of multigrid methods, that are commonly used in the context of differential equations systems, to solve high resolution data assimilation. Multigrid methods are iterative methods improved by the use of feedback corrections evaluated on coarse resolution. First in the case of linear assimilation, we study the robustness of multigrid approach and the efficiency of the coarse grid correction step. We then apply the multigrid algorithms on a non linear 1-D Burgers equation and on a 2-D Shallow-Water model. We study two types of algorithms, the Gauss Newton Multigrid, which lays on global linearization, and the Full Approximation Scheme. Their behavior is compared to more traditional approaches as incremental and multi-incremental ones.

La Méditerranée Nord-Occidentale, ou bassin Liguro-Provençal, est l'un des rares sièges de la convection océanique profonde. Ce phénomène localisé et intermittent est l'un des principaux modes d'interaction de l'océan profond avec le système climatique. Il est d'une importance primordiale pour la redistribution verticale de chaleur, de dioxyde de carbone et d'éléments biogéochimiques par l'océan, et donc pour le climat et la biologie marine. Le travail de thèse s'inscrit dans le cadre du programme HyMeX, il vise à caractériser le phénomène de convection dans le bassin Liguro-Provençal à partir du cas d'étude de l'année 2012-2013 et à comprendre le rôle de la dynamique de méso-échelle et de la variabilité intrinsèque océanique qui en résulte sur la convection. Le travail de thèse a tout d'abord porté sur la caractérisation du phénomène de convection océanique profonde à partir des observations du cas d'étude 2012-2013. On a estimé le taux de convection hivernale et de restratification printanière et une Expérience de Simulation d'un Système d'Observations (OSSE) a été développée pour estimer l'erreur d'observation associée. On conclut à la validité des observations du réseau MOOSE pour estimer les taux de convection et de restratification sur la période 2012-2013. On caractérise la période comme exceptionnellement convective avec un taux de convection hivernal de 2.3±0.5Sv (1Sv=106m³/s) et on estime pour la première fois un taux de restratification printanière de 0.8±0.4Sv. Deux approches numériques novatrices ont été développées au cours de la thèse pour caractériser le rôle de la méso-échelle et de la variabilité intrinsèque océanique sur le phénomène de convection. On a implémenté l'outil de raffinement de maille AGRIF en Méditerranée Nord-Occidentale dans le modèle régional NEMOMED12 pour documenter l'impact de la méso-échelle sur la convection océanique profonde et sur la circulation thermohaline Méditerranéenne. On a de plus réalisé des simulations d'ensemble à état initial perturbé pour documenter l'impact de la variabilité intrinsèque océanique sur la convection. Après avoir extensivement évalué le réalisme de la convection dans le modèle numérique NEMOMED12 grâce aux données de 2012-2013, on étudie avec ce modèle l'impact de la variabilité intrinsèque océanique sur ce phénomène. Sur le cas d'étude comme sur la période historique 1979-2013, la variabilité intrinsèque océanique module largement la géographie du patch convectif, en particulier dans le domaine hauturier. Aux échelles climatiques, la variabilité intrinsèque module largement la variabilité interannuelle du taux de convection. En moyenne climatologique, elle module aussi la géographie de la convection, mais elle impacte marginalement son intensité et les propriétés climatiques des eaux profondes. Enfin, on étudie avec l'outil AGRIF l'impact de la dynamique de méso-échelle sur la convection profonde et sur la circulation thermohaline. Sur le cas d'étude de 2012-2013, la méso-échelle augmente le réalisme de la convection. On montre qu'elle augmente la variabilité intrinsèque de la convection. Sur cette période comme sur la période historique, elle diminue l'intensité moyenne de la convection et réduit les transformations des eaux profondes. On relie principalement son impact sur la convection à une modification de la circulation stationnaire marquée par un repositionnement et une intensification des courants de bord, et la présence d'un méandre stationnaire du Front Baléare. Par ailleurs, sur la période historique, les échanges avec le bassin Algérien sont intensifiés par la méso-échelle, ce qui modifie les propriétés climatiques des masses d'eau. On montre enfin que la signature de la méso-échelle en surface est susceptible d'impacter les échanges air-mer et donc le climat côtier voire régional Méditerranéen. / The northwestern Mediterranean sea, also named the Liguro-Provençal basin, is one of the few places where ocean deep convection occurs. This localized and intermittent phenomenon is one of the main modes of interaction between the deep ocean and the climate system. It is of paramount importance for the vertical redistribution of heat, carbon dioxyde and biogeochemical elements, and therefore for climate and marine biology. The PhD has been carried out in the framework of HyMeX programme, it aims at characterizing the ocean deep convection phenomenon in the Liguro-Provençal basin from the year 2012-2013 case study and at understanding the role of mesoscale dynamics and of the resulting intrinsic ocean variability on deep convection. The PhD work has first focused on characterizing the ocean deep convection phenomenon from observations collected during the 2012-2013 case study. We estimated the winter deep convection and spring restratification rates and an Observing System Simulation Experiment (OSSE) was developed to estimate the associated observation error. We conclude on the validity of MOOSE network observations to estimate the deep convection and restratification rates in the period 2012-2013. We characterize the period as exceptionally convective with a winter deep water formation rate of 2.3±0.5Sv (1Sv=106m³/s) and we estimate for the first time a spring deep water restratification rate of 0.8±0.4Sv. Two novel numerical approaches were developped during the PhD to characterize the roles of mesoscale dynamics and of intrinsic variability in the deep convection phenomenon. We implemented AGRIF grid refinement tool in the northwestern Mediterranean Sea within NEMOMED12 regional model to document the impact of mesoscale on deep convection and on the Mediterranean thermohaline circulation. In addition, we carried out perturbed initial state ensemble simulations to characterize the impact of ocean intrinsic variability on convection. After extensively evaluating the realism of deep convection in NEMOMED12 numerical model thanks to the 2012-2013 observations, we study with this model the impact of intrinsic variability on deep convection. During the case study as well as in the 1979-2013 historical period, intrinsic ocean variability largely modulates the mixed patch geography, particularly in the open-sea domain. At climatic timescales, intrinsic variability modulates largely the deep convection rate interannual variability. On average over the historical period, it also modulates the mixed patch geography, but it impacts marginally its magnitude and the properties of the deep water formed. Finally, we study with AGRIF tool the impact of mesoscale dynamics on deep convection and on the thermohaline circulation. In the 2012-2013 case study, mesoscale improves the realism of the simulated convection. We show that it increases the deep convection intrinsic variability. In this period as well as during the 1979-2013 historical period, it decreases the mean deep convection rate and it reduces deep water transformations. We mainly relate its impact on convection to the modifincation of the stationary circulation characterized by a relocation and an intensification of boundary currents and the presence of a stationary Balearic Front meander. Also, in the historical period, exchanges with the Algerian basin are increased, which modifies water mass climatological properties. Finally, the surface signature of mesoscale is likely to alter air-sea interactions and the coastal to regional Mediterranean climate.

Le but de cette thèse est d'analyser plusieurs phénomènes de bord ouest de la Mer d'Arabie : (i) le cycle de vie d'un tourbillon de mésoéchelle persistant, le Great Whirl; (ii) la dynamique d'un écoulement d'eau dense (outflow) formée dans une mer adjacente, l'outflow du Golfe Persique; et (iii) une remontée d'eau profonde (upwelling) saisonnière dans la zone côtière d'Oman. Le point commun entre ces phénomènes est leur localisation sur un bord ouest océanique. Ils sont donc influencés par des forçages locaux (notamment les vents de mousson) et les forçages à distance (ondes de Rossby et tourbillons dérivant vers l'ouest). En particulier, ces derniers vont jouer un rôle particulier car la Mer d'Arabie est située à basses latitudes, ce qui implique une propagation rapide des ondes longues et tourbillons. De plus, des ondes sont continuellement excitées par le régime saisonnier des moussons. Nous avons mis au point des expériences numériques de différentes complexités en utilisant un modèle aux équations primitives. Ces expériences permettent soit de simuler de manière réaliste la dynamique complexe de la Mer d'Arabie, soit d'isoler un processus en particulier. Les résultats principaux peuvent se résumer comme suit : (i) le cycle de vie du Great Whirl est significativement impacté par les ondes de Rossby annuelles. Le rotationnel de la tension de vent joue un rôle important dans le maintien, le renforcement et la barotropisation du tourbillon. (ii) La dispersion de l'Eau du Golfe Persique (Persian Gulf Water, PGW) est déterminée par le mélange induit par les tourbillons de mésoéchelle. Précisément, ces tourbillons entrent dans le Golfe d'Oman (où se déverse la PGW), et interagissent avec la topographie. Ces interactions frictionnelles produisent des bandes de vorticité très intenses dans la couche limite de fond. Celles-ci sont arrachées et forment des tourbillons de sous-mésoéchelle. Ces tourbillons capturent de la PGW initialement située sur la pente continentale et la redistribuent dans le golfe d'Oman. Ce mécanisme donne finalement lieu à du mélange, permettant d'expliquer le gradient de salinité climatologique observé en profondeur. (iii) La dynamique de l'upwelling saisonnier au large d'Oman contraste fortement avec la dynamique des upwelling de bord est (Eastern Boundary Upwelling Systems, EBUS). En effet, les ondes de Rossby se propagent vers le large dans les EBUS et vers la côte dans l'upwelling de bord ouest d'Oman. Ces ondes modulent la réponse en température de l'upwelling forcé par le vent.Dans l'ensemble, ces résultats sont relativement spécifiques à la Mer d'Arabie. La faible extension zonale et la basse latitude de la Mer d'Arabie, ainsi que le régime de mousson des vents saisonniers en font une région particulière. La propagation rapide des ondes et tourbillons et leurs interactions avec le bord ouest façonnent les régimes de turbulence de la Mer d'Arabie. / This PhD aims to investigate some western boundary processes in the Arabian Sea : (i) the life cycle of the socalled Great Whirl, a persistent mesoscale eddy; (ii) the dynamics of the Persian Gulf outflow, a marginal sea dense outflow; and (iii) the seasonal Oman upwelling, a coastal upwelling forced by summermonsoonal winds. The cornerstone of all these phenomena is their locationat a western boundary, which makes then being influenced by both localforcing (e.g., monsoonal winds) and remote forcing (Rossby waves and wesward drifting eddies). Specifically, the later are expected to impact the western boundary dynamics since the low latitude of the Arabian Sea implies a fast westward propagation of long Rossby waves and eddies. Moreover, waves are continously excited by the reversing monsoonal winds. Based on a primitive equation model, we designed numerical experiments of different complexity that allowed to either realistically simulate the dynamics in the Arabian Sea or to isolate some processes.Major findings can be summarized as follows : (i) The Great Whirl life cycle is found to be significantly paced by annual Rossby waves, although the strong monsoonal wind stress curl is of major importance to sustain the structure. (ii) The Persian Gulf Water (PGW) spreading in the Gulf of Oman and the northern Arabian Sea can be explained by the stirring done by eddies entering the Gulf. These remotely formed surface intensifed mesoscale eddies propagate into the Gulf and interact with the topography. Frictional interactions produce intense vorticity strips at the boundary that detach and roll up in the interior, forming submesoscale coherent vortices (SCV). These SCV trap PGW initially located on the slope and redistribute it in the interior. This mechanism of transport ultimately produces mixing that explains the large-scale gradient of salinity in the gulf. (iii) We find that the dynamics of the seasonal upwelling of Oman contrasts with the more deeply studied Eastern Boundary Upwelling Systems (EBUS). In particular, Rossby waves, propagating offshore in EBUS vs. onshore in this western boudary upwelling, are found to modulate the wind driven upwelling and its sea surface temperature response.Overall, these results appear to be rather specific to the Arabian Sea. The short zonal extent and the low-latitude of the Arabian Sea, as well as the seasonally reversing wind forcing are the distinguishing features of this region. Fast waves and drifting eddies and their interactions with the western boundary significantly shape the turbulent regimes of the western Arabian Sea.