The total, annual, and monthly mean values of the number of NBC rings and other warm eddies occurring from 1993 through 1998
in the 1/16° data assimilating global NLOM are comparable to the numbers obtained by Goni and Johns (2001; GJ, hereafter)
from their analysis of TOPEX/POSEIDON altimeter data for the same
time period. Interannual variability of NBC ring and warm eddy formation is also similar
in the model.
Most of the individual rings reported by GJ were also found in the model with data
assimilation. A few NBC rings occurred in NLOM that were not noted by GJ and vice versa.
In agreement with the analysis by GJ, the data assimilating model indicates that warm rings occur
in the NBC region even when the retroflection does not appear to be present.
Some are clearly not of South Atlantic origin. Others appear to carry South Atlantic
water because they form in the same localized area where the NBC retroflection, when
it is present, sheds NBC rings. Indeed, they are fed by the NBC and contribute
to the western boundary layer portion of upper MOC transport identified by Fratantoni et al. (2000).
Finally, others may carry modified South Atlantic water into the region from the
ocean interior. In this case, the intergyre exchange mechanism is interior
surface Ekman transport (e.g. Fratantoni et al., 2000) instead of the
more direct NBC or NBC ring pathway along the western boundary.
The number
of clearly identifiable NBC retroflection eddies per year in the data assimilating model
is in agreement with the results of Fratantoni et al. (1995) based on observations
and an earlier, non-data-assimilating version of NLOM.
Not all NBC rings and warm eddies
found in the western tropical Atlantic off
the northeast coast of South America translate far enough westward to enter the
Caribbean Sea via the passages of the Lesser Antilles. Others decay before reaching
the island arc at the eastern edge of the Caribbean. Finally, others coalesce with
neighboring warm eddies in the region. This includes some NBC rings.
The results of the data assimilating model-based NBC ring census presented here with comparisons to the earlier studies indicate that it is important to make a definitive determination of the origin of the warm eddies in this region in order to determine their relative contribution to each cross-gyre pathway.