Eulerian and Lagrangian-mean flows and waves in exp2_w through exp2_w6

I present the Eulerian and Lagrangian-mean flow and the wave field for 5 experiments. The experiments are identical except that the Laplacian viscosity is progressively reduced (Tab. 1).

	Laplacian viscosity (m2/s)
exp2_w	500
exp2_w3	250
exp2_w4	100
exp2_w5	50
exp2_w6	10

** Table 1:** Laplacian viscosity in the 5 experiments considered.

The Eulerian and Lagrangian-mean flow in the middle layer and over a wave cycle (days 2000 to 2100) are shown in Figs. 1 and 2, respectively. With reduced viscosity, both Eulerian and Lagrangian-mean flow increase. Although the increase seems progressive in the first 4 experiments, in exp2_w6, the increase seems to be suddenly much larger, suggesting that there is some sort of instability, numerical or hydrodynamical, that is going on. The time series of the zonal flow (U) at 5°E and 33°N (Fig. 3) shows that in the first 4 experiments, as expected, the wave field amplitude increases and the phase of the wave changes progressively with reduced viscosity. The motion stays also periodic with a period of 100 days. In exp2_w6, however, the increase is surprisingly much larger and the flow is no more periodic at this period. The addition of a sponge layer along the western boundary (exp2_w6-sponge) does not seem to change qualitatively the present conclusions.

Figure 1: Eulerian-mean flow in the middle layer over a wave cycle (days 2000 to 2100).

Figure 2: Lagrangian-mean flow in the middle layer over a wave cycle (days 2000 to 2100). The Lagrangian-mean flow has been computed by following water parcels over the wave cycle. The velocity is the net displacement divided by the wave period and is then plotted at the initial position of the parcels.

Figure 3: Zonal velocity (U) at 5°E and 33°N. The zonal velocity has been plotted twice to see if the flow was still periodic or not, with a period of 100 days.