NBC Ring Census

      The number of NBC rings (20) and other warm eddies (11) found in the 1/16° global NLOM with data assimilation during the 6-year period from 1993 through 1998 is quite similar to the results obtained by Goni and Johns (2001, GJ hereafter) for the same time period (22 and 10, respectively). The average number of NBC rings per year in the model is 3.3+/-0.7 compared to 3.7+/-1 in the TOPEX/POSEIDON analysis by GJ. Other warm eddies in the NBC region that did not shed from the NBC retroflection occurred at a rate of nearly two per year in both studies. The annual mean of all warm eddies in the NBC region is 5.2+/-0.8 in the model and 5.3+/-1.3 reported by GJ.

     Both studies indicate the lack of NBC retroflection eddy shedding in the months of April and May. The highest concentration of other warm rings occurs during these months and the early summer when the NBC retroflection is not in place. NBC rings tend to form most often in the winter months. A few of the NBC rings observed by GJ (upper panel) were not seen in the model (lower panel) and vice versa. Some warm features that GJ could not discern to be from the NBC retroflection were revealed by the model to have indeed been pinched off from the retroflecting NBC. Other warm eddies observed by GJ clearly originated north of the retroflection in the model. Still others appeared along the NBC ring pathway apparently carrying South Atlantic water, but definitely not formed as a result of the NBC looping back on itself.

      The interannual variation in NBC ring formation (upper panel) is very similar between GJ and this study with a decrease from 1993 to 1995 then an increase in 1996 followed by a steady decrease over the next two years. The two studies also parallel each other in interannual variability of non-NBC ring formation (bottom panel). There is no variation from 1993 to 1995 (to 1996 in GJ), but then a sharp increase is seen during the last part of the time period examined.