D. Nof et Wk. Dewar, ALIGNMENT OF LENSES - LABORATORY AND NUMERICAL EXPERIMENTS, Deep-sea research. Part 1. Oceanographic research papers, 41(8), 1994, pp. 1207-1229
This paper addresses the question of what happens when two anticycloni
c lens-like eddies with different densities come in contact with each
other by, say, an advective current. We use laboratory and numerical e
xperiments to demonstrate that, in a similar fashion to lenses with id
entical densities which often merge, lenses with unequal densities oft
en align (i.e. the lenses ''strive'' toward a slate where one lens is
situated on top of the other). The two initially isolated eddies are f
ormed in the laboratory by continuously injecting salty water through
small filters situated in the interface separating two environmental l
ayers. Hence, there is a total of four different fluids with graduated
densities in the tank. The heaviest water is the lowest layer. One of
the lenses is slightly lighter than this heaviest water and the other
is still lighter than the first lens. Finally, the upper layer is the
lightest of the four. Since the lenses are formed by injection, weak
anticyclones which extend beyond the boundary of the lenses are formed
above and below the lenses. Due to the continuous injection of fluid
into the lenses they grow slowly so that after some time their edges m
eet. As in merging of lenses with identical densities, arms are then e
xtended from one vortex to the other. However, in the present case the
arms are situated above the heavier vortex and below the lighter vort
ex instead of being at the same level. At this point, the lenses are l
ocked together and the arms continue to propagate and grow until the l
enses ultimately align. This process is accompanied by filamentation w
here fluid is expelled from both lenses to accommodate the change in a
ngular momentum. Comparable numerical experiments also are discussed.
Here Gaussian lenses in a four-layer fluid are studied; the lenses occ
upy the two intermediate layers. The physical parameters of these expe
riments are chosen to resemble the now famous case of observed eddy me
rging off East Australia. In contrast to the laboratory study, the num
erical lenses have no initial anticyclones located in the layers surro
unding them. The results indicate alignment in a manner similar to the
laboratory experiments, provided the eddies are initially sufficientl
y close.