L’upwelling du Sud Vietnam (SVU) joue un rôle clef dans la dynamique océanique et la productivité biologique en Mer de Chine du Sud. Cette thèse vise à quantifier la variabilité interannuelle du SVU et identifier les facteurs et mécanismes en jeux. Pour cela un jeu de simulations numériques pluri-annuelles à haute résolution a été utilisé. Le réalisme du modèle a été évalué et optimisé par comparaison aux observations in-situ et satellites. Les résultats montrent que la forte variabilité du SVU est fortement pilotée par le rotationnel du vent estival, et liée à l’oscillation ENSO via son impact sur le vent. Cependant cette influence du vent est significativement modulée par la variabilité intrinsèque océanique liée aux interactions entre la vorticité associée aux tourbillons océaniques et le vent, et dans une moindre mesure par la circulation océanique de grande échelle et les fleuves. Ces conclusions sont robustes aux choix effectués pour corriger la dérive de surface du modèle. |
The summer South Vietnam Upwelling (SVU) is a major component of the South China Sea circulation that also influences the ecosystems. The objectives of this thesis are first to quantitatively assess the interannual variability of the SVU in terms of intensity and spatial extent, second to quantify the respective contributions from different factors (atmospheric, river and oceanic forcings; ocean intrinsic variability OIV; El-Niño Southern Oscillation ENSO) to the SVU interannual variability, and third to identify and examine the underlying physical mechanisms. For that we use a set of sensitivity eddy-resolving simulations of the SCS circulation performed with the ROMS_AGRIF ocean regional model at 1/12° resolution for the period 1991-2004. The ability of the model to represent realistically the water masses and dynamics of the circulation in the SCS and SVU regions was first evaluated by comparison with available satellite and in-situ observations. We then defined a group of sea-surface-temperature upwelling indexes to quantify in detail the interannual variability of the SVU in terms of intensity, spatial distribution and duration. Our results reveal that strong SVU years are offshore-dominant with upwelling centers located within 11-12oN and 110-112oE, whereas weak SVU years are coastal-dominant with upwelling centers located near the coast and over a larger latitude range (10-14oN). The first factor that triggers the strength and extent of the SVU is the summer wind curl associated with the summer monsoon. However its effect is modulated by several factors, including first the OIV, whose contribution reaches 50% of the total SVU variability, but also the river discharge and the remote ocean circulation. The coastal upwelling variability is strongly related to the variability of the eastward jet that develops from the coast. The offshore upwelling variability is impacted by the spatio-temporal interactions of the ocean cyclonic eddies with the wind stress curl, which are responsible for the impact of the OIV. The ocean and river forcing also modulate the SVU variability due to their contribution to the eddy field variability. ENSO has a strong influence on the SVU, mainly due to its direct influence on the summer wind. Those results regarding the interannual variability of the SVU are robust to the choice of the surface bias correction method used in the model. We finally present in Appendix preliminary results about the impacts of tides |