TOPEX/POSEIDON Kinetic Energy

The average surface geostrophic kinetic energy from the model over one year is calculated as $1/2 ( var(U')+var(V'))$ . It is computed from 10-day-average surface pressure fields in order to be best compared with the altimeter data (see Fig. 20). The altimeter data allows computation of the surface geostrophic velocity field from the sea surface slope. As expected , the model's surface geostrophic kinetic energy is smaller than the model's total surface kinetic energy. In this case , the ageostrophic energy is almost half the magnitude of the geostrophic energy which may be feasible in such a region of intense atmospheric forcing. Not surprisingly, the spatial pattern of eddy kinetic energy is very similar to the sea surface height variability : highest energies in the North Atlantic current, a local maximum in the rim current, and lower energies in the interior. Topex/Poseidon eddy kinetic energy can be computed in the cross-track direction which yields a full picture if an isotropic eddy field is assumed. This calculation with unfiltered altimeter data gives kinetic energies a magnitude higher than the model (not presently shown). However, it is well known that altimeter data is increasingly noisy at small scales . To compound the problem, calculation of geostrophic velocity involves the subtraction of two noisy quantities and adds to the noise levels of the result. Other studies filter the altimeter data to remove energy at small scales to circumvent this problem. This is not done here; rather I will qualitatively compare the model kinetic energy with the 2 degree gridded data of 1992-1994 previously presented by Stammer et al (1996). Once again, the model does very well in estimating the kinetic energy in the North Atlantic current. The model energy seems to be low by about 20 percent in the North Atlantic Current. The model also accurately predicts a large area of increased eddy energy in the West Greenland Current, although the magnitude is less than half the observed value. The interior energy levels also seem to be low in the model, but this is consistent with the overall spatial pattern and the previous findings with the sea surface height variability.