Final Activity Report Summary - OCEAN3D (Estimation of the upper ocean 3D velocities from remotely sensed Sea Surface Temperature)
A key problem in oceanography is the synoptic estimation of the three-dimensional velocity field of the ocean. For the time being this can only be obtained via direct oceanic measurements which are expensive and labour intensive, and thus are limited in space and time. Satellite oceanography can be used to estimate surface ocean currents as a minimum. Indeed, at present, surface horizontal velocities are regularly estimated from altimetric measurements. Through the combination of several satellite passes it is possible to produce fields of horizontal geostrophic velocities with spatial resolutions of 25 km for mid-latitudes. Although altimetric measurements are widely used to study ocean dynamics, their spatial resolution and sampling geometry limits the study of the smallest mesoscale and sub-mesoscale phenomena. Furthermore, vertical velocities cannot be estimated from altimetry or subsurface velocities. Alternatively, infra-red sensors providing sea surface temperature (SST) have relatively high resolutions, are synoptic over large areas and benefit from numerous dedicated satellites. Such characteristics have led, in the past two decades, in the development of different methodologies to infer motion from their measurements. However, the lack of timely acquisitions and the presence of clouds limit the operational capacity of this method.
In order to overpass these limitations we followed a completely new approach based on the properties of baroclinic flows. Recent theoretical developments in geophysical fluid dynamics suggested that the dynamics of the upper layers of the ocean could be modelled using an effective version of surface quasi-geostrophic equations, namely the eSQG approach. This approach required a single snapshot of SST and the set-up of two parameters, i.e. the mean Brunt-Vaisala frequency and a parameter that determined the energy level at the ocean surface. Firstly, the validity of this approach was tested in different types of numerical simulations. Results showed that the three-dimensional velocity field, including horizontal and vertical velocities, in the upper hundreds of meters of the ocean could be reconstructed under adequate environmental conditions. These conditions occurred after a mixed-layer deepening period. Therefore, the ideal situation for the application of this method would be after strong wind events. The methodology was then applied to real SST data and compared to synthetic aperture radar (SAR) data. Results demonstrated that this approach was able to reconstruct the surface ocean dynamics including the surface divergence field.
In order to overpass these limitations we followed a completely new approach based on the properties of baroclinic flows. Recent theoretical developments in geophysical fluid dynamics suggested that the dynamics of the upper layers of the ocean could be modelled using an effective version of surface quasi-geostrophic equations, namely the eSQG approach. This approach required a single snapshot of SST and the set-up of two parameters, i.e. the mean Brunt-Vaisala frequency and a parameter that determined the energy level at the ocean surface. Firstly, the validity of this approach was tested in different types of numerical simulations. Results showed that the three-dimensional velocity field, including horizontal and vertical velocities, in the upper hundreds of meters of the ocean could be reconstructed under adequate environmental conditions. These conditions occurred after a mixed-layer deepening period. Therefore, the ideal situation for the application of this method would be after strong wind events. The methodology was then applied to real SST data and compared to synthetic aperture radar (SAR) data. Results demonstrated that this approach was able to reconstruct the surface ocean dynamics including the surface divergence field.