The Upper Equatorial Indian Ocean. The Climatological Seasonal Cycle

Abstract

The climatological seasonal cycle of the upper equatorial Indian Ocean is discussed. A summary of the observations is given. Near the surface and below the equatorial thermocline, the observations indicate an intense variability of the equatorial currants, primarily at the semiannual frequency. In the thermocline, an eastward flowing equatorial undercurrent present in March and April is the dominant feature. An analysis of the temperature profiles in the upper 500 meters also indicates at all depths, seasonal vertical displacements well above the estimated uncertainty. At the semiannual period, the different isotherms are oscillating in phase in the upper western Indian Ocean and an upward phase propagation is present in the east. Displacements at the annual period have shorter spatial scales and are usually smaller, the near surface and the deeper oceans am separated at this period by a level of no displacements in the upper thermocline. The equatorial thickening of the thermocline in March–April in the central and western Indian Ocean also suggests the presence of an eastward flowing undercurrent at that time. To investigate the dynamics, a simplified wind forced linear model in the equatorial beta-plane is presented for a stratified Indian Ocean. The wind stress is applied as a body force over a mixed layer of specified thickness and below a dynamically induced mixing is included. A decomposition in vertical modes is done and the lower modes are solved with a numerical scheme from Cane and Patton. The dynamics of the higher modes are characterized by Yoshida jets (negligible zonal gradients) and an approximate solution is sought for them. The model response to wind forcing is sensitive to thew hypotheses with the surface equatorial currants being strongly dependent on the mixed layer depth and with the currents in the thermocline being controlled by mixing. The model reproduces many of the most salient characteristics of the observed seasonal cycle, both for the currents and the isotherm vertical displacement. As in earlier studies, this suggests that the wind stress near the equator is an important forcing of the equatorial Indian Ocean, at least to depths of 500 m. However, typical differences between the observed seasonal cycle and the modeled one are of the order of 10 days in phase and a factor 1.5 in amplitude. Sensitivity studies suggest that these differences could not be overcome by tuning the model adjustable parameters, but that they are likely to arise from the crude representation of the forcing as a body force and the neglect of the nonlinearities.