Temporal Changes in Ocean Eddy Transports

Abstract

New estimates from 11 yr of altimetric data are made of the global time-average variability kinetic energy and its decadal-scale variability. Making the approximation that the variability reflects primarily eddy motions, a time-mean, but spatially varying, eddy mixing coefficient is then estimated along with its changes over the last decade. With a record length more than 2 times that previously available, the time-mean variability kinetic energy KE is statistically more reliable and smoother in its spatial pattern. Minimum values of KE are present in the subpolar North Pacific Ocean and in the eastern South Pacific (both less than 100 cm2 s−2). In contrast to the North Pacific, the subpolar North Atlantic Ocean shows relatively enhanced KE. Eddy kinetic energy and eddy mixing appear to have declined during the last decade over large parts of the western Pacific Ocean, in some regions by as much as 50% of the time-mean value. Increased eddy variability can be found in the Kuroshio and Gulf Stream regions, as well as in the Agulhas region, east of Australia, and at several locations along the Antarctic Circumpolar Current. Somewhat enhanced eddy variability and eddy mixing are also apparent in the eastern tropical Pacific. A numerical simulation of the ocean circulation at 1° spatial resolution over a 10-yr period suggests that variations in eddy mixing of this order of magnitude measurably affect the deep temperature field in the vicinity of permanent frontal structures on a time scale of less than 4 yr. The meridional overturning circulation also reacts on these time scales. If persistent over longer periods in the ocean, these effects would be important for climate simulations.

Observability of the Irminger Sea circulation using variational data assimilation

“…Applications of the ECCO system to global ocean models are described by Stammer et al (2004) at 1 • and Stammer et al (2002) at 2 • resolution. Here we have adapted the ECCO system to a high-resolution regional model of the Irminger Sea (see also Gebbie et al (2006) for a discussion of state estimation in a regional eddy-resolving version of the ECCO model).…”

“…Because the final cost reaches the target value the initial velocity field is less important to the evolution of the model than the initial temperature and salinity field. Studies with the ECCO assimilation system over several years in the global ocean work well without initial velocity controls (for example, see Stammer et al 2002Stammer et al , 2004. The longer assimilation window and lesser inertial flow make the initial velocity field much less influential in these cases, however.…”

Treating strong adjoint sensitivities in tropical eddy‐permitting variational data assimilation

“…It contains twice-daily wind stress vectors and daily net heat flux, net short-wave radiation and water flux at the sea surface. The second forcing dataset was taken from the ECCO flux fields (Stammer et al 2004). These are the NCEP forcings adjusted by a variational 1 • × 1 • global state estimation procedure (Köhl 2005, personal communication).…”

Opportunities for enhanced collaboration within the data assimilation community

“…That is one reason why disciplinary research groups are setting up their own data assimilation programmes. An example is the consortium Estimating the Circulation and Climate of the Ocean (ECCO), formed under the sponsorship of the National Ocean Partnership Program in the USA (Stammer et al 2004). Another example is the Data Assimilation Research Centre (DARC), a consortium of academic institutions set up by the Natural Environment Research Council in the UK.…”