instabilities, eddies, and topographically trapped currents in the Arctic Ocean.
Koji Shimada (JAMSTEC)
The Arctic Ocean circulation is mainly governed by mesoscale eddies and topographically trapped currents along shelf breaks and major sea mounts, The eddies are generally generated though instabilities of the unstable currents on shelf breaks, coastal boundary, and density fronts. The spatial behavior of the eddies may be an important process to establish the Arctic Ocean stratification, since the motion of eddies affects the water mass mixing and transportation. From the different point of view, the eddies can drive the mean currents though the interactions with seafloor topography. An evidence of such phenomena was observed by a drifting buoy observation.
1. OBSERVATION OF BAROCLINIC INSTABILITY
The ice-Ocean Environmental Buoy-2 (IOEB-2) consisting of meteorological, sea-ice, and ocean sensors was deployed at 85゜,50'N, 12゜ 03' W north of Greenland in 1994, The IOEB-2 was drifted southward along a front north of Fram Strait where the warm and salty Atlantic Water through the Fram encounters the cold and fresh East Greenland current (Fig. 1). The time series of CTD mounted on the line of IOEB-2 at 8m, 43m, 75m, and 110m showed several temperature and salinity events (Fig. 2). These events seemed to appear due to vertical convection accompanying brain ejection. For the present, however, it may be useful to examine more closely at some of the more important features of the events. Figure 3 shows the meridional section of the density anomaly from its average value between 81°15' N and 84°45' N. The pattern sifts northward with depth. This feature corresponds to the pattern of the unstable Eady waves. Therefore the temperature and salinity events can be explained by disturbances generated by baroclinic instabilities.
2. NONLINEAR EVOLUTION OF UNSTABLE BOUNDARY CURRENTS
Nonlinear evolution of linearly unstable barotropic and baroclinic boundary currents is investigated using the contour dynamics method, and we also give a physical interpretation on their nonlinear behaviors.
In the case of the barotropic currents, we consider boundary currents consisting of three piecewise constant vorticity regions which are divided by two vorticity fronts, located at y= L1(x,t) and L2(x,t). A straight coastal boundary lies at y=0 and periodic domain with length of λ, is considered in the x-direction (Fig. 4). The nonlinear behavior fall into three distinctive regimes referred to as breaking wave, vortex pair and boundary trapped vortex regime. The first regime, breaking wave, is characterized by the breaking of the wave crest at upstream side (Fig. 5). The strong mixing can occur around the strong vorticity pool within the width of the basic current, but the coastal water mass cannot be carried far away from the boundary. The second regime, vortex pair, is characterized by the formation of the dipole structure consisting of two vortices with opposite sign. The coastal water mass trapped by the vortex pair is advected far away from the boundary (Fig. 6). The third regime, boundary trapped vortex, is characterized by the meander of the current in which the vortex generated from the unstable currents keeps to be trapped by the coastal boundary (Fig. 7). Therefore the transport hardly occurs. The main difference among these three regimes is interpreted by the temporal change of the phase relationship between vorticity centers in the piecewise constant vorticity regions. In the breaking wave regime, the phase lag of the vorticity center in the offshore region, xc2, to that of the coastal region, xc1, increases. When the lag exceeds π, the phase relationship turn from the growing stage into the damping stage, since the periodic length is short enough. On the contrary, in the vortex pair regime, the periodic length is too large for the phase relationship to change from that in the growing stage to the damping stage. As a result, the unstable wave continues to grow and forms a vortex pair structure. From this explanation, we can easily understand that the boundary of the vortex pair and the breaking wave regime does not depend only on the basic current structure but also on the wavelength of the initial disturbance. An example showing this dependency on the wavelength is shown in Fig. 8 in which the basic current is chosen as the same as in Fig. 5, but we double the wavelength. In this case, the phase relationship between the two vorticity regions does not turn after the breaking and a dipole structure is organized after all. In the boundary trapped vortex regime, the change of the phase relationship can occur similarly to that in the breaking wave regime, but the manner is different. In the boundary trapped vortex, the phase difference between the vorticity centers decreases and the phase relationship changes when difference of the two vorticity centers is 0. After that, the positive vorticity pool in the offshore region precedes the negative vorricity pool in the coastal side, and the amplitude decreases.
Based on the importance of the temporal behavior between the voracity centers, we consider a point vortices model in which the piecewise constant vorticity regions are approximated by the point vortices with the same circulation (Fig. 9). Since the motion of point vortices in the present model is determined by the several restrictions of conserved quantities, the motion is examined by analytically using phase diagram method. In spite of the simplicity of the problem, the point vortices analysis captures the essence of the nonlinear behavior of the unstable wave, as for the classification of the motion of the point vortices and the temporal evolution of the phase relationship between the point vortices. For example, the phase diagrams are shown in Fig. 10(a-c) , in which the parameters are fixed, except the periodic length. As the periodic length increases, the contours, on which the maximum offshore distance occurs at ξ = π corresponding to the breaking wave disappears and the entire space is occupied by the vortex pair regimes except for the region in the vicinity of ξ-axis. From this transition of the motion of the point vortices, we can expect that the transition from the breaking wave into the vortex pair regime accompanies with catastrophic change.
In the case of baroclinic currents, we consider boundary currents consisting of two piecewise constant vorticity regions in each layer and the other geometry is
Corresponding author address: Koji Shimada, JAMSTEC, 2-15 Natsushima, Yokosuka 237-0061, Japan
e-mail shimadak@jamstec.go.jp
