Therefore, the interaction process between the Okhotsk Sea and the North Paciflc needs to be clarified in order to better understand the circulation and the water properties of the North Pacific intermediate layer.
Recent observations implied the importance of the water modification by strong vertical mixing at the Kurils in freshening the intermediate layer to supply low salinity required for production of the NPIW, and suggested that tidal mixing there is responsible for the intense vertical mixing (e.g., Talley 1991). However, the actual physical mechanism is still unknown, and hence this is an important problem for clarifying the boundary mixing processes in the North Pacific.
According to observations and our oreceding tidal flow simulation, the currents are predominated by the diurnal tidal components in the Kuril Straits, and the semidiurnal components are rather weak. Very swift K1 currents have been thought to cause intense vertical mixing by interactions with large-amplitude sills in the Kuril Straits. However, this situation is out of the range of previous theories for oceanic internal waves, because they assumed that an oscillating tidal flow over an obstacle excites only internal waves at its tidal frequency (internal tides). For example, Hibiya (1986)'s theory, which is often used for interpretation of wave growth, states that internal tides propagating upstream are trapped at the generation region and amplified when the barotropic flow is critical (i.e., when the Froude number Fn is unity where Fn is the ratio of the barotropic tidal flow speed to phase speed of nth mode, Fn = ｜U(t)/cn｜). But the diurnal tides are subinertial around this high latitude (〜 47°N) and hence internal tides at the K1 tidal frequency are not freely propagating waves. Further, the K1 currents are subcritical.
Thus, as a first step to clarifying the physics responsible for the vertical mixing in the Kuril Straits and its influence on the North Pacific intermediate layer, we have performed numerical simulations of tidally generated internal waves and their nonlinear evolution in the Kuril Straits, and estimated the vertical mixing induced by those waves.
2. Numerical model
The model topography is a representative of the sills in the northeastern part of the Kuril Straits, where tidal currents are so strong that could cause considerable mixing (Fig.1). The sill top is 550m deep, and the bottom on both sides of the sill is set to be flat with a maximum depth of 2000m. We can adopt this flat bottom simplification since the density core of the NPIW (〜 26.8δθ) is around a depth of 300m.
In order to reproduce vertical mixing by internal waves, we used a nonhydrostatic model with horizontal and vertical grid sizes of 500m and 10m, respectively. The horizontal and vertical eddy viscosity coefficients are assigned the relatively small values of 2 × 105cm2s-1 and 0.1cm2s-1, respectively, so that their effect on mixing is small enough to demonstrate the wave mixing clearly. As basic forcing terms for internal wave generation, barotropic K1 and M2 currents are given at both lateral boundaries. Their maximum speeds at the sill top are 0.5ms-1 for the K1 case and 0.2ms-1 for the M2 case, as determined from our preceding barotropic tidal simulations. We took account of the effect of rotation by adopting an f-plane approximation to distinguish the physics of waves generated by the subinertial K1 flow from that of the superinertial M2 flow in the Kuril Straits.
