Mesoscale eddies in satellite observations of SSS and inferred eddy freshwater transports

Mesoscale eddies are ubiquitous throughout the world’s oceans and contribute significantly to the ocean transport of heat and freshwater, key mechanisms that influence ocean properties and global climate. Satellites have provided a unique opportunity to observe mesoscale eddies due to their better spatiotemporal coverage compared to in-situ observing systems and, for nearly a decade, have included observations of sea surface salinity (SSS).

In this study, observations of SSS from NASA’s SMAP, and ESA’s SMOS satellite missions are evaluated to determine the extent to which the existing satellite data can be used to quantify and map eddy contributions to the ocean freshwater transport. With a typical mesoscale eddy signal in SSS of 0.1-0.3 psu and the RMS error of satellite retrievals of 0.3 psu, the signal-to-noise ratio barely approaches 1. On a statistical basis, however, very small standard errors, smaller than 0.01 psu, can be achieved for the average, composite eddies, resulting in robust estimates of the eddy-induced freshwater transports. This approach is applied near-globally to yield estimates in two dimensions.

In the tropics and mid-latitudes, estimates of the eddy freshwater transport from the two satellites demonstrate a remarkably consistent and physically meaningful picture. The transports are divergent in the tropics and convergent in the subtropical gyres, generally consistent with the mean SSS gradients. The estimates start to diverge regionally at higher latitudes resulting, presumably, from a dramatic increase in the level of noise in satellite SSS retrievals (poor sensitivity of the L-band radiometer to SSS in cold water) and/or resolution issues (eddy length scales decrease with latitude). Despite the limitations, our results demonstrate that the role of mesoscale eddies in the ocean freshwater transport can be assessed from existing satellite data. We also discuss spatial patterns of the eddy freshwater transport resulting from our computations.