Abstract
A linear model based on the quasigeostrophic equations is constructed in order to predict the vertical structure of Rossby waves and, more broadly, of anomalies resolved by altimeter data, roughly with periods longer than 20 days and with wavelengths larger than 100 km. The subsurface field is reconstructed from sea surface height and climatological stratification. The solution is calculated in periodic rectangular regions with a 3D discrete Fourier transform. The effect of the mean flow on Rossby waves is neglected, which the authors believe is a reasonable approximation for low latitudes. The method used has been tested with an idealized double-gyre simulation [performed with the Miami Isopycnal Coordinate Ocean Model (MICOM)]. The linear model is able to give reasonable predictions of subsurface currents at low latitudes (below approximately 30°) and for relatively weak mean flow. However, the predictions degrade with stronger mean flows and higher latitudes. The subsurface velocities calculated with this model using AVISO altimetric data and velocities from current meters have also been compared. Results show that the model gives reasonably accurate results away from the top and bottom boundaries, side boundaries, and far from western boundary currents. This study found, for the regions where the model is valid, an energy partition of the traditional modes of approximately 68% in the barotropic mode and 25% in the first baroclinic mode. Only 20% of the observed kinetic energy can be attributed to free Rossby waves of long periods that propagate energy to the west.
